ADC - Fetal & Neonatal - March2009

Egyptian_Pediatric yahoo group Egyptian_Pediatric yahoo group
http://health.groups.yahoo.com/group/ http://health.groups.yahoo.com/group/
egyptian_pediatric/ egyptian_pediatric/
Contents Volume 94 Number 2 | ADC Fetal and Neonatal Edition March 2009
F138 How common are rib fractures in extremely low

Fantoms birth weight preterm infants?
F79 M Ward Platt
D Smurthwaite, N B Wright, S Russell,
A J Emmerson, M Z Mughal
Original articles
F80 The effect of two levels of pressure support F140 Mortality of twin and singleton livebirths under
ventilation on tidal volume delivery and minute 30 weeks’ gestation: a population-based study
ventilation in preterm infants B Ray, M P Ward Platt
Journal of the Royal College of
S Gupta, S K Sinha, S M Donn
Paediatrics and Child Health F144 Neonatal infections in Asia
F84 Volume-guarantee ventilation: pressure may R Tiskumara, S H Fakharee, C Q Liu,
Editor-in-Chief decrease during obstructed flow P Nuntnarumit, K M Lui, M Hammoud,
Howard Bauchner K I Wheeler, C J Morley, C O F Kamlin, J K F Lee, C B Chow, A Shenoi, R Halliday,
Deputy Editors P G Davis D Isaacs, on behalf of APNIS (Asia-Pacific
Patrick Cartlidge Neonatal Infections Study)
F87 Oxygen saturation and heart rate during delivery
Imti Choonara Short reports
room resuscitation of infants ,30 weeks’
Martin Ward Platt
gestation with air or 100% oxygen F149 Neonatal blood pressure waves are associated
Commissioning Editor J A Dawson, C O F Kamlin, C Wong, A B te Pas, with surges of systemic noradrenaline
Patrick Cartlidge C P F O’Donnell, S M Donath, P G Davis, B Wefers, S Cunningham, R Stephen, N McIntosh
Perspectives Editor C J Morley
F152 Early individualised parenteral nutrition for
Ieuan Hughes F92 No change in developmental outcome with preterm infants
incubator covers and nesting for very preterm S Eleni-dit-Trolli, E Kermorvant-Duchemin,
Editorial Office
infants in a randomised controlled trial C Huon, M Mokthari, K Husseini, M-L Brunet,
Archives of Disease in Childhood
C M Maguire, F J Walther, P H T van Zwieten, C Dupont, A Lapillonne
BMJ Publishing Group Ltd
BMA House S Le Cessie, J M Wit, S Veen, On behalf of the
Leiden Developmental Care Project
PostScript
Tavistock Square F154 Impact of delayed screening for prolonged
London WC1H 9JR, UK F98 Psychological stress of parents of preterm infants jaundice in the newborn
T: +44 (0)20 7383 6331 enrolled in an early discharge programme from M Tyrell, S Hingley, C Giles, J O Menakaya
F: +44 (0)20 7383 6668 the neonatal intensive care unit: a prospective
E: archdischild@bmjgroup.com randomised trial F154 Hydrocortisone treatment for severe evolving
ISSN: 0003-9888 (print) P Sáenz, M Cerdá, J L Dı́az, P Yi, M Gorba, bronchopulmonary dysplasia and cerebral
ISSN: 1359-2998 (online) N Boronat, P Barreto, M Vento haemodynamics
G Cambonie, R Mesnage, C Milési, A Rideau,
Impact factor 2.342 F105 Risk of stillbirth and neonatal death linked with C Veyrac, J-C Picaud
maternal mental illness: a national cohort study
S King-Hele, R T Webb, P B Mortensen, F155 Differences between the amino acid concentra-
Disclaimer: ADC is published by BMJ Publishing tions of umbilical venous and arterial blood
Group Ltd (a wholly owned subsidiary of the British L Appleby, A Pickles, K M Abel
Medical Association) and the Royal College of H Tsuchiya, K Matsui, T Muramatsu, T Ando,
Paediatrics and Child Health. The owners grant F111 Impact of shielding parenteral nutrition from light F Endo
editorial freedom to the Editor of ADC. ADC follows
guidelines on editorial independence produced by on routine monitoring of blood glucose and
the World Association of Medical Editors and the triglyceride levels in preterm neonates F156 Comparison of peripheral and cerebral tissue
code on good publication practice of the Committee
M Khashu, A Harrison, V Lalari, J-C Lavoie,
oxygenation index in neonates
on Publication Ethics.
K Grossauer, G Pichler, G Schmölzer, H Zotter,
ADC is intended for medical professionals and is
P Chessex
W Mueller, B Urlesberger
provided without warranty, express or implied.
Statements in the journal are the responsibility of their F116 Random safety audits in the neonatal unit
authors and advertisers and not authors’ institutions, F156 Correction
L Lee, S Girish, E van den Berg, A Leaf
the BMJ Publishing Group, the Royal College of
Paediatrics and Child Health or the BMA unless Images in neonatal medicine
otherwise specified or determined by law.
Acceptance of advertising does not imply F120 Enteral feeding regimens and necrotising F104 The seasonal orchidometer
endorsement. C Durand, J Gibbs
enterocolitis in preterm infants: a multicentre
To the fullest extent permitted by law, the BMJ case–control study
Publishing Group shall not be liable for any loss, injury F137 Depressed skull fracture in a newborn baby
or damage resulting from the use of Heart or any G Henderson, S Craig, P Brocklehurst, W McGuire S T Dharmaraj, N D Embleton, A Jenkins, G Jones
information in it whether based on contract, tort, or
otherwise. Readers are advised to verify any F124 Cytokine gene polymorphisms in preterm infants
information they choose to rely on.
with necrotising enterocolitis: genetic association
Copyright E 2009 BMJ Publishing Group and Royal study
College of Paediatrics and Child Health. All rights This article has been chosen by the Editor to be of special interest
G Henderson, S Craig, R J Baier, N Helps, or importance and is freely available online.
reserved; no part of this publication may be
reproduced, stored in a retrieval system, or P Brocklehurst, W McGuire
transmitted in any form or by any means, electronic, Articles carrying the Unlocked Logo are freely available online
mechanical, photocopying, recording, or otherwise F129 Neonatal extracorporeal membrane oxygenation: under the BMJ Journals unlocked scheme.
without prior permission.
practice patterns and predictors of outcome in See http://adc.bmj.com/info/unlocked.dtl
ADC is published by BMJ Publishing Group Ltd, the UK
typeset by The Charlesworth Group, and printed in the
UK on acid-free paper from sustainable forests by A Karimova, K Brown, D Ridout, W Beierlein, C O P E COM M ITTE E ON P U B LICATION ETH ICS
Cambrian Printers Limited, Aberystwyth, UK. J Cassidy, J Smith, H Pandya, R Firmin,

ADC Fetal & Neonatal Edition (USPS No: 015-715) is M Liddell, C Davis, A Goldman This journal is a member of and subscribes to the principles of the
published bimonthly by BMJ Publishing Group and Committee on Publication Ethics
distributed in the USA by SPP, 75 Aberdeen Road, F133 Sleeping position, oxygenation and lung function in
Emigsville, PA 17318. Periodicals postage paid at www.publicationethics.org.uk
Emigsville, PA. POSTMASTER: send address changes prematurely born infants studied post term
to Archives of Disease in Childhood, PO Box 437, T Saiki, H Rao, F Landolfo, A P R Smith,
Emigsville, PA 17318-0437, USA.
S Hannam, G F Rafferty, A Greenough
Downloaded from fn.bmj.com on 5 March 2009
The seasonal orchidometer

C Durand and J Gibbs
Arch. Dis. Child. Fetal Neonatal Ed. 2009;94;F104

doi:10.1136/adc.2008.140921
Updated information and services can be found at:

http://fn.bmj.com/cgi/content/full/94/2/F104
These include:
Rapid responses You can respond to this article at:
http://fn.bmj.com/cgi/eletter-submit/94/2/F104
Email alerting Receive free email alerts when new articles cite this article - sign up in the box at
service the top right corner of the article
Notes
To order reprints of this article go to:

http://journals.bmj.com/cgi/reprintform
To subscribe to Archives of Disease in Childhood - Fetal and Neonatal Edition go to:

http://journals.bmj.com/subscriptions/
Original article
22. Forcada-Guex M, Pierrehumbert B, Borghini A, et al. Early dyadic patterns of mother-infant 29. Melnyk BM, Feinstein NF, Alpert-Gillis L, et al. Reducing premature infants’ length of
interactions and outcomes of prematurity at 18 months. Pediatrics 2006;118:e107–14. stay and improving parents’ mental health outcomes with the creating opportunities
23. Arnaud F. Discharge of very preterm infants from neonatology: check list. J Gynecol for parent empowerment (COPE) neonatal intensive care unit programme: a
Obstet Biol Reprod 2004:S108–10. randomized, controlled trial. Pediatrics 2006;118:e1414–27.
24. Sauve R, Lee SK. Neonatal follow-up programmes and follow-up studies: Historical 30. Allen EC, Manuel JC, Legault C, et al. Perception of child vulnerability among
and current perspectives. Paediatr Child Health 2006;11:267–70. mothers of former premature infants. Pediatrics 2004;113:267–73.
25. Executive Committee of the Connecticut Chapter of the AAP. Neonatal 31. Spear ML, Leef K, Epps S, et al. Family reactions during infants’ hospitalization in the
Intensive care Unit (NICU). Discharge Guidelines 2005. neonatal intensive care unit. Am J Perinat 2002;19:205–13.
26. Kotagal UR, Perlstein PH, Gamblian V, et al. Description and evaluation of a 32. Wigert H, Johansson R, Berg M, et al. Mothers’ experiences of having
programme for the early discharge of infants from a neonatal intensive care unit. their newborn child in a neonatal intensive care unit. Scand J Caring Sci
J Pediatr 1995;127:285–90. 2006;20:35–41.
27. Casiro OG, McKenzie Me, McFadyen L, et al. Earlier discharge with community- 33. Vento M, Saénz P, Valle S, et al. Early discharge programme from the NICU with co-
based intervention for low birth weight infants: a randomized trial. Pediatrics operation of the Primary Care Paediatrician. EPAS 2006;595535.204.
1993;92:128–34. 34. Garel M, Bahaud M, Blondel B. Consequences for the family of a very preterm birth
28. Raddish M, Merrit TA. Early discharge of premature infants. A critical analysis. Clin two months after discharge. Results of EPIPAGE qualitative study. Arch Pediatr
Perinatol 1998;25:499–520. 2004;11:1299–307.
Images in neonatal medicine

In springtime, as crocuses emerge, daffodils blossom and lambs
gambol in the fields, paediatricians’ thoughts naturally turn to
orchidometers. A seminal paper in 2001 elegantly demonstrated
how a Teaser sweet could be substituted for an 8 ml
orchidometer bead.1 This provided grateful paediatricians with
a readily available means to assess mid-puberty in adolescent
boys as well as a source of sustenance at the end of a busy clinic.
To the dismay of many paediatricians, by 2007 the testicular
Teaser had mutated into a flat-bottomed dome that retained its
edible qualities but was quite useless as an orchidometer
substitute.2 The manufacturer was urged to reinstate Teasers
to their former aesthetic and functional glory. Fortunately, as
shown in the illustration, the Teaser has been refashioned into
its celebrated orchidometer shape, although somewhat dimin-
ished as a 6 ml rather than 8 ml orchidometer bead, along with
Galaxy, Mars and Milky Way mini eggs. Furthermore, the
Eastertide appearance of the Cadbury mini creme egg provides
an excellent substitute for the 10 ml orchidometer bead with all
the tactile and edible qualities so well described in relation to
the Teaser. The standard Cadbury creme egg, although slightly
too large to substitute as an orchidometer bead, is a useful
indicator of a pathologically enlarged testis. Each spring,
paediatricians should grasp this seasonal opportunity of an
edible orchidometer with both hands. We’ve got the balls—let’s Figure 1 The seasonal orchidometer.
not be afraid to use them.
C Durand, J Gibbs Competing interests: None.
Department of Paediatrics, Countess of Chester Hospital, Chester, Cheshire, UK Arch Dis Child Fetal Neonatal Ed 2009;94:F104. doi:10.1136/adc.2008.140921
Correspondence to: J Gibbs, Department of Paediatrics, Countess of Chester REFERENCES

Hospital, Liverpool Road, Chester CH2 1UL, Cheshire, UK ; john.gibbs@coch.nhs.uk 1. Bhalla P, Sally, Pippa, et al. An inexpensive and edible aid for the diagnosis of puberty
in the male: multispecies evaluation of an alternative orchidometer. BMJ
Acknowledgements: Thanks to R Cooke for his photographic expertise and for 2001;323:1486.
refraining from eating the artwork. 2. Williams G, Dharmaraj P. Dissent of the testis. BMJ 2007;335:1287.
F104 Arch Dis Child Fetal Neonatal Ed March 2009 Vol 94 No 2

Risk of stillbirth and neonatal death linked with

S King-Hele, R T Webb, P B Mortensen, L Appleby, A Pickles and K M Abel
Arch. Dis. Child. Fetal Neonatal Ed. 2009;94;F105-F110; originally published

online 10 Nov 2008;
doi:10.1136/adc.2007.135459

These include:
References This article cites 27 articles, 13 of which can be accessed free at:
http://fn.bmj.com/cgi/content/full/94/2/F105#BIBL

Notes


Original article
Risk of stillbirth and neonatal death linked with

S King-Hele,1 R T Webb,1,3 P B Mortensen,2 L Appleby,1 A Pickles,3 K M Abel1
c Additional tables are ABSTRACT

published online only at http:// Background: Babies of mothers with psychotic disorders What is already known on this topic
adc.bmj.com/content/vol94/
issue2
are known to have higher rates of poor obstetric outcome,
1
including higher mortality rates. c A number of studies have linked raised risks of
Centre for Women’s Mental Objective: To estimate risks of stillbirth and neonatal
Health, University of perinatal death to serious mental illness in the
Manchester, Manchester, UK; death by specific causes in babies of mothers with mother; a few studies have not found raised
2
National Centre for Register- histories of severe mental illness, relative to the general risks in this population.
Based Research, University of population. c Maternal substance and alcohol abuse during
Aarhus, Aarhus, Denmark; Methods: A cohort of 1.45 million live births and 7021
3
Biostatistics/Health
pregnancy has been linked to raised risks of
Methodology Research Group, stillbirths during 1973–98 was identified from Danish babies having poor birth outcomes.
University of Manchester, national registers. These registers were linked to identify
Manchester, UK babies who were stillborn or died neonatally after
exposure to maternal psychiatric illness.
Correspondence to:
Results: Risks of stillbirth and neonatal death were raised What this study adds
Dr K M Abel, Centre for
Women’s Mental Health, for virtually all causes of death for all of the maternal
University of Manchester, 2nd psychiatric diagnostic categories. For most causes of c The degree of raised perinatal mortality risk
Floor East, University Place, linked with serious maternal mental disorder
Oxford Road, Manchester M13 death, offspring of women with schizophrenia and related
9PL, UK. kathryn.abel@ disorders had no greater risks of stillbirth or neonatal does not vary greatly or systematically by cause
manchester.ac.uk death than offspring of women with other maternal of death or by type of mental illness.
psychiatric disorders (eg, neonatal death (NND) due to c This lack of specificity suggests that multiple
Accepted 22 August 2008 immaturity: relative risks (95% CI) schizophrenia and causal mechanisms are involved and that
Published Online First complex approaches to intervention are
10 November 2008
related disorders: 1.1 (0.4 to 3.5), affective disorders: 2.0
(1.2 to 3.5)). There was a greater risk of fatal congenital therefore required.
malformation associated with a history of maternal
affective disorder (stillbirth 2.4 (1.1 to 5.1), NND 2.1 (1.4
to 3.3)) or schizophrenia and related disorders (stillbirth or affective psychosis adjusted for a range of
2.4 (0.8 to 7.6), NND 2.2 (1.1 to 4.1)) than with maternal factors including maternal smoking history.10
alcohol/drug-related disorders (stillbirth 1.2 (0.4 to 3.8), The psychiatric journals have tended to concen-
NND 1.1 (0.6 to 2.2)). trate on associations between psychiatric diagnoses
Conclusions: Higher risk of perinatal loss may be linked (especially schizophrenia) and poor birth out-
to factors associated with maternal psychiatric illness in comes: Bennedsen et al found raised risks of
general, such as insufficient attendance for antenatal care congenital malformations in babies of women
with schizophrenia in Denmark,11 but paediatric
and unhealthy lifestyles rather than the maternal mental
publications have focused on the effects of drug
illness itself.
and alcohol misuse on birth outcomes. Many
studies have identified maternal substance abuse
during pregnancy as a risk factor for a range of poor
Raised risks of perinatal death in babies of women birth outcomes such as congenital malformations
with psychiatric illness have been found in a of varying type and severity, prematurity and
number of studies.1–4 Gaining a greater under- perinatal death.12–15
standing of these risks is an important public In this study we aimed to examine a range of
health concern, especially since increasing numbers specific causes of perinatal death in babies of
of mentally ill women now become pregnant.5 women with psychiatric inpatient histories,
Previous research has found raised risks of perinatal including substance-related disorders. We exam-
death associated with maternal schizophrenia,1–3 ined the problem of rarity of exposure (severe
maternal psychotic disorder4 and parental schizo- maternal mental illness) and rarity of outcome
phrenia,2 6 7 and raised risks of stillbirth among (perinatal death) by using data from the large
women with affective or substance-related dis- population registers available in Denmark.
orders in the general US population.8 Our research
has suggested that higher risks of neonatal death METHODS
are associated with maternal affective or alcohol/ Study cohort
drug-related disorders than with maternal schizo- We identified all live and stillbirths between 1
phrenia and related disorders.2 Several studies have January 1973 and 31 December 1998 using data
found no evidence of raised risks for either stillbirth from the Danish Civil Registration System.16 Each
or neonatal death in the offspring of women with Danish resident is assigned a unique number which
either schizophrenia,9 bipolar or unipolar disorders9 may be used to link information about the
Arch Dis Child Fetal Neonatal Ed 2009;94:F105–F110. doi:10.1136/adc.2007.135459 F105

Original article
individual between different registers to a high degree of Cause-of-death classifications

completeness. We used the Central Population Register contain- We classified causes of death according to the ICD codes for the
ing the date of birth, death and emigration of each person living primary cause of stillbirth or neonatal death that were recorded
in Denmark, the Psychiatric Central Register recording diag- in the national registers. The ICD-8 and ICD-10 codes used in
noses for all psychiatric inpatient admissions since 1969,17 the each category of stillbirth and neonatal death are shown in the
Cause of Death Register18 and the Medical Births Register,19 Appendix, table A1. These codes were selected by an obste-
which record the cause of death or stillbirth, respectively, since trician (Dr Louise Kenny; see ‘‘Acknowledgements’’) on the
1973. The International Classification of Diseases (ICD) 8th basis of clinical relevance.
revision20 was used from 1 January 1973 to 31 December 1993
and the 10th revision21 thereafter. We used 1 January 1973 as the Modelling
study entry date because the Medical Births Register was We used Poisson regression (STATA, version 8.0) to estimate
computerised from that date. Causes of stillbirth were not relative risks of mortality in the offspring of mothers previously
recorded in the Medical Births Register after 31 December 1996. admitted with mental illness, compared with unexposed
A stillbirth is recorded in Denmark when the fetus is lost at offspring. We compared the cause-specific stillbirth rates and
28 weeks’ gestation or later.22 A neonatal death is one that neonatal mortality in exposed versus unexposed populations,
occurs during the first 28 days of life. with relative risks estimated as risk ratios. The risk of stillbirth
We restricted our cohort to singleton births since observations was calculated as the number of deaths divided by the number
from multiple birth sets are not statistically independent. We of live and stillbirths; risk of neonatal death was the number of
identified 7021 stillbirths between 1 January 1973 and 31 deaths over the number of live births. Using Poisson models
December 1996, and 1 450 329 live births between 1 January these risk ratios were adjusted for 5-year period bands. We
1973 and 31 December 1998. additionally adjusted for maternal age at birth for stillbirths,
and maternal age at birth, birth order and offspring sex for live
births (these additional adjustments made no material differ-
Exposure status and maternal psychiatric admission
ence to the effect and variance estimates, and so these results
Babies were ‘‘exposed’’ to maternal mental illness if their
are not included in this paper).
mother was admitted to hospital with any psychiatric illness (or
A previous study using the same dataset to 1993 estimated
specific diagnostic category) before the offspring date of birth
relative risks of stillbirth and neonatal death using generalising
(live or stillborn). Diagnoses of maternal mental illness were
estimating equation methods to take account of familial
categorised using ICD-8 and ICD-10 codes below; we included
clustering effects.11 Since most perinatal deaths occurred during
all admissions for psychiatric disorders in women aged 16 years
the early part of the period, and the previous study found little
and over. These groups of codes were selected for reasons of
difference between results using this method and those that
clinical relevance, and have been used in other Danish registry
made no such adjustment, we chose not to use methods which
studies.2
account for within-family correlation. This decision was further
c Schizophrenia and related disorders: (schizophrenia, schizo-
supported by prior validation work conducted by one of this
phrenia-like disorders and schizoaffective disorders) ICD-8:
study’s coauthors (RTW) using the same study cohort we
295, 296.8, 297, 298.39, 301.83; ICD-10: F20–F29.
report in this paper.23
c Affective disorders: (bipolar disorder and other affective
disorders) ICD-8: 296.09, 296.19, 296.29, 296.39, 296.99,
298.09, 298.19, 300.49, 301.19; ICD-10: F30–F39. RESULTS
c Alcohol/drug-related disorders: alcohol-related disorders: ICD- Stillbirths
8: 291, 303; ICD-10: F10 and drug-related disorders: ICD-8: Of the 7021 stillbirths, 188 were exposed to a history of any
294.3, 304, 980.09; ICD-10: F11–F16, F18–F19. maternal psychiatric admission before birth. Of these mothers,
Figure 1 Relative risks of stillbirth to

mothers admitted with psychiatric illness
compared with the Danish general
population, 1973–96. Maternal
psychiatric history and cause of stillbirth:
Affect, affective disorders; Alc/drug,
alcohol or drug-related disorders; Schizo,
schizophrenia and related disorders.
F106 Arch Dis Child Fetal Neonatal Ed 2009;94:F105–F110. doi:10.1136/adc.2007.135459

Original article
Figure 2 Neonatal deaths: relative risks

of mortality for offspring of mothers
admitted with psychiatric illness
compared with the Danish general
population, 1973–98. Maternal
psychiatric history and cause of neonatal
death: Affect, affective disorders; Alc/
drug, alcohol or drug-related disorders;
Schizo, schizophrenia and related
disorders.
19 had been admitted for schizophrenia and related disorders, 47 Compared with the general population, a maternal history of
for affective disorders and 38 for alcohol/drug-related disorders. alcohol/drug-related disorder was associated with a greater than
Figure 1 and supplementary table 1 (online only) show the threefold raised risk of both neonatal death due to anoxia and
relative risks of cause-specific stillbirth linked with maternal birth injury to the brain (RR = 3.4, 95% CI 2.1 to 5.6), and
admission for specific psychiatric disorders, compared with the ‘‘other conditions originating in the perinatal period’’ (RR = 3.8,
general population. 95% CI 2.1 to 6.9). We found greater than doubled risks of
The risks of stillbirth are raised for each of the causes of death neonatal death in babies of women with histories of affective
for all of the maternal psychiatric diagnostic categories. Some of disorders for death due to anoxia and birth injury to the brain
the relative risks are more than twofold but we found no (RR = 2.6, 95% CI 1.6 to 4.3), ‘‘other conditions originating in
pattern of raised risks by either cause of stillbirth or maternal the perinatal period’’ (RR = 2.3, 95% CI 1.2 to 4.7) and
diagnostic category. Maternal psychiatric history of alcohol/ congenital malformations (RR = 2.1, 95% CI 1.4 to 3.3). A
drug-related disorder is associated with a greater than twofold history of schizophrenia and related disorders is associated with
increased risk of stillbirth due to complications of delivery greater than doubled risks of neonatal death due to congenital
(relative risk (RR) = 2.3, 95% confidence interval (CI) 1.2 to 4.2) malformations compared with the general population
and similar raised risks for stillbirth due to congenital (RR = 2.2, 95% CI 1.1 to 4.1).
malformations of the fetus in women with histories of affective
disorders (RR = 2.4, 95% CI 1.1 to 5.1). Non-significant raised
Similarities between risks of stillbirth and neonatal death
risks were seen in women with schizophrenia and related
Across the maternal diagnostic categories, there were similar
disorders, where there were only three women in the exposed
patterns of relative risks between stillbirth and neonatal death
group (RR = 2.4, 95% CI 0.8 to 7.6, n = 3). ‘‘All other causes of
from congenital malformations, and between stillbirth due to
stillbirth’’ include those due to injury to the mother or due to
complications of delivery, which include anoxia and birth injury
maternal illness. We found sevenfold and twofold raised risks,
to the brain, and neonatal death as a result of anoxia and birth
respectively, for these causes of death (injury to mother:
injury to the brain.
RR = 7.5, 95% CI 2.9 to 19.0, n = 5; maternal illness:
RR = 2.5, 95% CI 1.3 to 4.8, n = 10).
DISCUSSION
Neonatal deaths Main findings
Of the 6646 neonatal deaths, 201 were of offspring of mothers We observed raised risks of stillbirth and neonatal death for each
who were previously admitted for any psychiatric illness (22 cause of death and for all maternal psychiatric histories, with
with maternal schizophrenia and related disorders, 66 with only one exception. In addition, we did not find higher risks of
maternal affective disorders and 55 with maternal alcohol/drug- stillbirth or neonatal death in babies of women with schizo-
related disorders). Figure 2 and supplementary table 2 (online phrenia and related disorders compared with the other
only) show the relative risks of cause-specific neonatal death psychiatric disorders. A markedly raised risk of stillbirth after
linked with history of maternal admission for specific psychia- injury to the mother was indicated amongst women with any
tric disorders, compared with the general population. psychiatric admission history, compared with the general
The relative risk of each cause of neonatal death for each type population, but this result was based on just five deaths in
of maternal psychiatric history was raised, with one exception. the exposed population. A history of maternal schizophrenia
This was neonatal death due to anoxia and birth injury to the and related disorder or maternal affective disorder was found to
infant brain in children of women with histories of schizo- be a greater risk factor for perinatal death due to congenital
phrenia and related disorders (RR = 0.9, 95% CI 0.2 to 3.7). malformation than maternal history of substance-related

Original article
disorders. Our results are consistent with previous studies and substance abuse as a risk factor for receiving inadequate
reporting raised risks of stillbirth and neonatal death in antenatal care30 and other research indicates that late booking
offspring of women with schizophrenia.1 and fewer antenatal visits increase the incidence of babies born
with low or very low birth weight or preterm birth.29
Strengths and limitations Conversely, perhaps greater care is taken with women with
Data were collected prospectively over a 26-year period and are schizophrenia and similar disorders, illnesses that are regarded
based on the whole population of Denmark, allowing us to as particularly serious. We had no information about the nature
examine cause-specific outcomes for a range of mental illnesses. of the injuries of the women who lost babies owing to injury to
A limitation is that we examined offspring outcomes for the mother and, in particular, whether the injuries were
mothers with a history of inpatient admission for severe sustained accidentally or by violent assault.
psychiatric illness, so our results possibly cannot be generalised
to those with less serious mental illness. For this investigation Congenital malformations
we had no data to examine whether mothers smoked, drank A maternal history of schizophrenia or affective disorder is a
alcohol or took illicit drugs during pregnancy, and no measures greater risk factor for perinatal death from congenital mal-
of socioeconomic status. Prescription of psychotropic drugs, formations than a maternal history of substance-related
mothers’ physical health status in pregnancy and antenatal care psychiatric disorders. The mechanism for this is unclear. One
attendance record were also not available from the study possible explanation might be the use of prescribed psychotropic
registers. drugs during pregnancy. Little research has been conducted in
Bennedsen et al found comparable point estimates of relative this field and our results require confirmation by replication
risk of all-cause stillbirth and neonatal death associated with using other datasets.
maternal schizophrenia using the same data from the Danish
registers over a shorter time period.11 Our study reports relative
risks associated with the wider diagnostic range of maternal CONCLUSION
schizophrenia and related disorders and over a longer time Future research is required to replicate our findings, particularly
period, increasing the power of the study. where we found high relative risks for small numbers of
Since psychiatric data begin from 1969, only the previous 4 stillbirths or neonatal deaths in the exposed population. The
years of psychiatric history were known for women who gave raised risks related to psychiatric disorders are probably the
birth in 1973, whereas the psychiatric history of women giving result of a range of factors including social problems which can
birth in later years was known for a much longer time period. be hard to deal with. However, women who have contact with
However, any underascertainment of maternal mental illness in mental health services are a readily identifiable group for whom
the earlier years would tend to underestimate relative risks in extra care during and after pregnancy may reduce preventable
this period and overall. The number of beds available for perinatal deaths. Our results indicate that obstetric, paediatric
psychiatric patients has decreased since the 1970s.24 The and mental health workers should be aware that women with
perinatal mortality rate has also fallen; in our birth cohort mental health problems are in particular need of good
there was an approximate 20% fall in the stillbirth rate between reproductive health planning and antenatal care, ideally
the early 1970s and late 1990s, while the neonatal death rate fell provided by multidisciplinary teams that can care for their
by more than 50%. This study allowed for these cohort effects complex range of physical and mental needs during pregnancy.
in its design by adjusting for period as a covariate.
Acknowledgements: We thank Dr Louise Kenny for classifying the stillbirths and
neonatal deaths into appropriate cause-of-death categories. We also acknowledge the
Timing of psychiatric illness and perinatal death contribution of Heine Gøtzsche and Thomas Munk Laursen for linking registers,
This study limited stillbirths and neonatal deaths in the exposed supplying data and answering queries. In accordance with Danish legislation this
population to those that occurred at some time after an project was approved by the Danish Data Protection Agency and the relevant Register
authorities.
admission for psychiatric illness, in order to prevent reverse
causality bias.25 We thereby aimed to estimate risks associated Funding: The study was funded by project grant 073935 from the Wellcome Trust,
with known serious mental illness before death, and specifically England and by the Stanley Medical Research Institute, Chevy Chase, Maryland.
to exclude mental illness that may in part have been Competing interests: None.
precipitated by the loss of a child. Using national Danish
registry data, Li et al reported raised risks of maternal psychiatric REFERENCES
hospitalisation in these circumstances.26 1. Nilsson E, Lichtenstein P, Cnattingius S, et al. Women with schizophrenia:
pregnancy outcome and infant death among their offspring. Schizophr Res
2002;58:221–9.
Poor lifestyle choices and antenatal care 2. Webb RT, Abel KM, Pickles AR, et al. Mortality risk among offspring of psychiatric
The general nature of the raised risks seen across all maternal inpatients: a population-based follow-up to early adulthood. Am J Psychiatry
2006;163:2170–7.
disorder categories suggests the involvement of risk factors
3. Sobel D. Infant mortality and malformations in children of schizophrenic women:
related to the psychiatric population itself, rather than to preliminary data and suggested research. Psychiatr Q 1961;35:60–4.
specific diagnostic groups. Likely contributory factors to these 4. Howard LM, Goss C, Leese M, et al. Medical outcome of pregnancy in women with
raised risks in this population include poor lifestyle choices such psychotic disorders and their infants in the first year after birth. Br J Psychiatry
2003;182:63–7.
as antenatal smoking and drinking,27 28 and the relatively low
5. Oates M. Patients as parents: The risk to children. Br J Psychiatry Suppl
socioeconomic status of many people with mental health 1997;170:22–7.
problems.29 6. Rieder RO, Rosenthal D, Wender P, et al. The offspring of schizophrenics. Fetal and
Raised risks due to complications of delivery, anoxia and birth neonatal deaths. Arch Gen Psychiatry 1975;32:200–11.
7. Modrzewska K. The offspring of schizophrenic parents in a North Swedish isolate.
injury to the infant brain may be related to poor uptake of Clin Genet 1980;17:191–201.
prenatal care in some sections of the psychiatric population. 8. McCabe JH, Martinsson L, Lichtenstein P, et al. Adverse pregnancy outcomes in
Previous research in the USA has identified psychiatric illness mothers with affective psychosis. Bipolar Disord 2007;9:305–9.

Original article
9. Jablensky AV, Morgan V, Zubrick SR, et al. Pregnancy, delivery, and neonatal 22. Kesmodel U, Wisborg K, Olsen SF, et al. Moderate alcohol intake during pregnancy
complications in a population cohort of women with schizophrenia and major affective and the risk of stillbirth and death in the first year of life. Am J Epidemiol
disorders. Am J Psychiatry 2005;162:79–91. 2002;155:305–12.
10. Gold KJ, Dalton VK, Schwenk TL, et al. What causes pregnancy loss? Pre-existing 23. Webb R. Mortality in the offspring of people admitted for psychiatric
mental illness as in independent factor. Gen Hosp Psychiatry 2007;29:207–13. treatment: a national cohort study. PhD thesis, University of Manchester, 2007.
11. Bennedsen BE, Mortensen PB, Olesen AV, et al. Congenital malformations, 24. Søgaard HJ, Godt HH, Blinkenberg S. Trends in psychiatric hospitalization and
stillbirths, and infant deaths among children of women with schizophrenia. Arch Gen changes in admission patterns in two counties in Denmark from 1977 to 1989. Soc
Psychiatry 2001;58:674–9. Psychiatry Psychiatr Epidemiol 1992;27:263–9.
12. Jones KL, Smith DW, Ulleland CN, et al. Pattern of malformation in offspring of 25. Marquis GS, Habicht JP, Lanata CF, et al. Association of breastfeeding and stunting
chronic alcoholic mothers. Lancet 1973;1:1267–71. in Peruvian toddlers: an example of reverse causality bias. Int J Epidemiol
13. Handler A, Kistin N, Davis F, et al. Cocaine use during pregnancy: perinatal 1997;26:349–56.
outcomes.[see comment]. Am J Epidemiol 1991;133:818–25. 26. Li J, Laursen TM, Precht DH, et al. Hospitalization for mental illness among parents
14. Ostrea EM, Jr., Brady M, Gause S, et al. Drug screening of newborns by meconium after the death of a child. N Engl J Med 2005;352:1190–6
analysis: a large-scale, prospective, epidemiologic study. Pediatrics 1992;89:107–13.
27. Murray JL, Bernfield M. The differential effect of prenatal care on the incidence of
15. Gillogley KM, Evans AT, Hansen RL, et al. The perinatal impact of cocaine,
low birth weight among blacks and whites in a prepaid health care plan. N Engl J Med
amphetamine, and opiate use detected by universal intrapartum screening.
1988;319:1385–91.
Am J Obstet Gynecol 1990;163:1535–42.
28. Lasser K, Boyd JW, Woolhandler S, et al. Smoking and mental illness: a population-
16. Pedersen CB, Gotzsche H, Moller JO, et al. The Danish Civil Registration System: a
cohort of eight million persons. Dan Med Bull 2006;53:441–9. based prevalence study. JAMA 2000;284:2606–10.
17. Munk-Jorgensen P, Mortensen PB. The Danish Psychiatric Central Register. Dan 29. Hudson CG. Socioeconomic status and mental illness: tests of the social causation
Med Bull 1997;44:82–4. and selection hypotheses. Am J Orthopsychiatry 2005;75:3–18.
18. Juel K, Helweg-Larsen K. The Danish registers of causes of death. Dan Med Bull 30. Kelly RH, Danielsen BH, Golding JM, et al. Adequacy of prenatal care among women
1999; 46:354–7. with psychiatric diagnoses giving birth in California in 1994 and 1995. Psychiatr Serv
19. Knudsen LB, Olsen J. The Danish Medical Birth Registry. Dan Med Bull 1999;50:1584–90.
1998;45:320–3.
20. WHO. ICD-8 international classification of diseases. Geneva: World Health
Organization, 1965. APPENDIX
21. WHO. ICD-10 international classification of diseases. Geneva: World Health Table A1 shows the classification of causes of stillbirth and neonatal death.
Organization, 1992.
Table A1 Classification of causes of stillbirth and neonatal death*

Cause of death Description and comments ICD-10 1994-8{ ICD-8 1973–93{
Stillbirths
Antenatal complications Includes: P01.1-P01.5, P01.8, P02.0-P02.7, 762.1–762.3, 762.9, 769.0, 769.1, 770.0–770.2,
Placental abruption and infarction P03.0, P03.1, P03.8, P05.0-P05.2, 770.8–771.1, 771.9, 776.1, 776.2, 777.0
Poor growth of placenta P05.9, P07.3
Interuterine growth restriction
Pre-eclampsia
Prematurity
Complications of delivery Includes: O69.8, P20.1, P20.9, P24.0 634.3, 764.4, 764.9, 765.0, 765.1, 765.4, 765.9,
Anoxia 766.0, 766.3, 766.4, 766.9, 767.0, 767.4, 767.9,
Hypoxia 768.0,768.3–768.5, 768.9, 769.2, 769.4, 769.5,
Prolapse of umbilical cord 769.9, 772.0, 772.8, 772.9, 776.0, 776.3, 776.4,
Difficult labour with various problems such as 776.9
malposition of fetus and birth injury without mention
of cause
Congenital malformations of Includes: Q00.0–Q99.9 740.0–759.9
the fetus Spina bifida
Congenital anomalies of the heart, digestive system,
urinary system, etc
Maternal illness Includes: P00.0–P00.4, P00.8, P00.9 760.0–760.5, 761.1– 761.3, 761.6, 761.7, 761.9
Congenital heart disease of the mother
Chronic hypertension of the mother
Maternal diabetes
Rubella
An operation
Injury to mother Neither ICD-8 nor ICD-10 codes specify the type or P00.5 761.5
cause of the injury to the mother which leads to the
stillbirth
All other causes of stillbirth Includes: E88.9, P04.3, P23.9, P35.9, P37.9, 775.0, 775.9, 778.0, 778.2, 779.0, 779.9, 795.0,
Unknown causes of death P39.2, P39.9, P50.3, P52.5, P52.8, 170.6, 192.2, 199.1, 228.0, 320.9, 593.2, 778.0,
Cancers P55.0, P54.9, P55.9, P56.0, P70.1, 778.1, 778.2, 778.9
Postmaturity P70.2, P83.2, P95.9
Haemorrhages
Neonatal deaths
Immaturity related conditions Includes: P01.0, P01.1, P07.2, P07.3, P220, 769.0, 769.1, 776.1, 776.2, 777.0
Incompetent cervix P228, P22.9
Premature rupture of membranes
Respiratory distress of newborn
Anoxia and birth injury to the Includes: P10.3, P20.0, P20.9, P21.0, P21.9 765.4, 766.0, 766.4, 767.0, 767.4, 768.0, 772.0,
brain Difficult labour with malposition of fetus with 776.3, 776.9
asphyxia, anoxia or hypoxia
Birth asphyxia
Continued

Original article
Table A1 Continued
Cause of death Description and comments ICD-10 1994-8{ ICD-8 1973–93{
Other conditions originating Includes: All other codes within P00.0–P99.9 All other codes within 760.0–779.9 not included in
in the perinatal period Other maternal conditions unrelated to pregnancy not included in either immaturity or either immaturity or anoxia and birth injury to the
Conditions of placenta, placental infarction anoxia and birth injury to the brain brain
Conditions of umbilical cord
Aspiration of contents of birth canal
Cardiovascular disorders originating in the perinatal
period
Congenital malformations Includes: Q00.0–Q99.9 740.0–759.9
Spina bifida
Congenital anomalies of the heart, digestive system,
urinary system, etc
All other causes of death Includes: C76.7, D33.1, G71.1, X990 009.2, 038.9, 258.9, 273.9, 280.0, 285.9, 320.8,
(exposed population only{) Unknown causes of death and cancers 320.9, 466.9, 486.0, 567.0, 795.0, 795.8, 796.2,
910.5, 913.9, 962.0
*These were the codes used for any maternal psychiatric illness. Owing to small numbers, we grouped together the maternal illnesses, injury to mother and all other causes of
stillbirth categories for specific psychiatric diagnoses; {the codes are only those reported in the medical births or cause-of-death registers as the primary cause of stillbirth or death
in Denmark 1973–98; {the codes in the unexposed population are too numerous to be included in this table. They may be obtained from the corresponding author.
Drug and Therapeutics Bulletin (DTB)
Your key source of unbiased, independent advice

For over 45 years DTB has been an independent, indispensable part of evidence-based clinical practice.
DTB offers healthcare professionals detailed assessment of, and practical advice on, individual
medicines and other treatments, groups of treatment and the overall management of disease.
DTB is now also available online at http://dtb.bmj.com:

c browse or search all DTB content from the latest issue back to 1994
c email alerting, sophisticated searching, RSS feeds and full text links from cited references
c interactive services such as My Folders for quick access to articles that you have viewed previously
and My Searches to save and re-use useful searches
c comment online on any DTB article
To subscribe, or for further information, please visit http://dtb.bmj.com

Impact of shielding parenteral nutrition from light

on routine monitoring of blood glucose and
triglyceride levels in preterm neonates
Minesh Khashu, Adele Harrison, Vikki Lalari, Jean-Claude Lavoie and Philippe
Chessex

online 23 Jul 2008;
doi:10.1136/adc.2007.135327

These include:

Notes


Original article
Impact of shielding parenteral nutrition from light on

routine monitoring of blood glucose and triglyceride
levels in preterm neonates
Minesh Khashu,1 Adele Harrison,1 Vikki Lalari,1 Jean-Claude Lavoie,2 Philippe Chessex1
1
Division of Neonatology, ABSTRACT
Children’s and Women’s Health Background: Premature infants are vulnerable to What is already known on this topic
Centre of BC, University of
British Columbia, Vancouver, BC,
complications related to oxidative stress. Exposure to light
Canada; 2 Department of increases oxidation products in solutions of total c Light exposure increases the products of
Pediatrics, CHU Sainte–Justine, parenteral nutrition (TPN) such as lipid peroxides and oxidation in total parenteral nutrition.
University of Montreal, QC, hydrogen peroxide. Oxidative stress impairs glucose c Oxidative stress affects glucose and lipid
Canada
uptake and affects lipid metabolism. Hypothesis: products metabolism.
Correspondence to: of photo-oxidation contaminating TPN affect lipid meta-
Philippe Chessex, Division of bolism.
Neonatology, Children’s and Objective: Evaluate the effect of photoprotection of TPN
Women’s Health Centre of B.C., What this study adds
4480 Oak St, Vancouver, BC, in preterm infants on plasma glucose and triglyceride (TG)
Canada, V6H 3V4; pchessex@ concentrations.
cw.bc.ca Design: Secondary analysis of a prospective study c Failure to photoprotect total parenteral nutrition
allocating preterm infants to light-exposed (LE, n = 32) or (TPN) causes alterations in routine monitoring of
Accepted 3 July 2008 light-protected (LP, n = 27) TPN. blood glucose and plasma triglycerides.
Published Online First
23 July 2008 Setting: Level III NICU referral centre for patients of c Shielding TPN from light decreases the
British Columbia. accumulation of triglycerides allowing for
Patients: Preterm infants requiring TPN. increased parenteral delivery of energy in the
Interventions and outcome measures: TG and blood form of lipid emulsion.
glucose measured during routine monitoring while on full
TPN were compared between LE and LP.
Results: Clinical characteristics were similar between the solution from the generation of peroxides.9 10 The
two groups (gestational age 28¡1 wk; birth weight: infusion of light-exposed (LE) MVP or H2O2
1.0¡0.1 kg). Nutrient intakes from TPN and from induces comparable oxidative responses in lungs
minimal enteral nutrition were comparable between LE of guinea pig pups.11 12 In the liver13 and plasma14 of
and LP. Blood glucose was higher in preterm infants these animals, the infusion of LE MVP is associated
receiving LE (p,0.001). The accumulation of TG with with increased triglyceride (TG) concentration
increasing lipid intake was twice as high with LE suggesting that peroxides infused with TPN might
accounting for significantly higher TG levels on days 8 and interfere with lipid metabolism. Furthermore,
9 (p,0.05). oxidative stress impairs glucose uptake in muscle
Conclusions: Failure to photoprotect TPN may cause and fat15; therefore, we questioned whether the
alterations in intermediary metabolism. Shielding TPN generation of oxidants in TPN could influence
from light provides a potential benefit for preterm infants parameters frequently used to monitor the ade-
by avoiding hypertriglyceridaemia allowing for increased quacy of the metabolic response.
substrate delivery. The aforementioned studies and results from our
trial to evaluate effects of photoprotection of TPN
on clinical and biochemical endpoints16–18 prompted
Early nutrition is associated with marked long- us to test in a post hoc analysis whether photo-
term benefits for preterm infants.1 2 In view of the protection of TPN in preterm infants affects
small gastric capacity and functional immaturity of intermediary metabolism resulting in alterations
the gastrointestinal tract, very low birth weight in plasma glucose and TG concentrations.
premature infants require total parenteral nutri-
tion (TPN) to help achieve adequate nutritional METHODS
intakes until full enteral feeds can be established. We conducted a prospective study in which
Exposure of TPN to ambient light generates preterm infants were allocated to LE and light-
organic peroxides3 4 and hydrogen peroxide protected (LP) TPN to compare the effects of
(H2O2)5 6 that represent an oxidative load7 which shielding TPN on clinical16 17 and on biochemical
could be of significance in preterm infants who endpoints.18 Infants admitted to the NICU
have immature antioxidant defences.8 Photo-sensi- between 2001 and 2004 and requiring TPN were
tised riboflavin present in parenteral multivitamin eligible for the study. Infants with multiple
preparations (MVP) catalyses electron transfer congenital anomalies, and those who were
between electron donors such as vitamin C, amino unstable (requiring vasopressors, inhaled nitric
acids or lipids6 and dissolved oxygen producing oxide or paralysis) or had sepsis were excluded.
H2O2. Shielding TPN from light protects the Allocation to TPN regimens was carried out in the

Original article
pharmacy prior to commencement of the multivitamin/amino RESULTS

acid/dextrose parenteral nutrition (PN) solution using a compu- Sixty-two neonates allocated to LP and 66 randomised to LE
ter-generated randomisation sequence. PN was initiated from were recruited to a study aimed at determining the effects of
birth with amino acid/dextrose solution providing 2.25 g/kg/d photoprotecting TPN on nutrient handing.16 17 Of that initial
of amino acids. Following introduction of micronutrients, group, 32 LP subjects and 27 LE subjects received full TPN and
minerals, vitamins and lipids on day 2, TPN was increased daily were therefore included in the present secondary analysis. Mean
to achieve full intakes of amino acids, glucose and fat. Lipid was gestational age (LE: 28¡1 vs LP: 28¡1 wk), birth weight
initiated at 0.5 g/kg/d in infants with a birth weight ,1000 g, and (1.0¡0.1 vs 1.0¡0.1 kg), initial severity of illness index SNAP II
at 1 g/kg/d in infants .1000 g. PN was increased daily until score19 (LE: 24¡4 vs LP: 18¡2), did not differ between infants
achieving full TPN intake: 130 ml/kg/day of amino acid/dextrose receiving full TPN. Figure 1 documents that there was no
solution + 1.5 ml/kg/day of multivitamins (MVP: Multi-12 difference between LE and LP in the progression over time of the
Pediatric, Sandoz, Montréal, QC, Canada) and 20 ml/kg/day of intravenous macronutrient intakes. Upon reaching full TPN
20% lipid solution (Intralipid 20%, Pharmacia Upjohn, LIEU) mean (SD) intravenous intakes for LE vs LP on day 7 of life were
introduced into the venous line close to the site of infusion in the as follows: glucose: 13.9 (0.5) vs 13.2 (0.4) g/kg/d; amino acids:
baby. Full TPN consisted of amino acids: 3.5–4 g/kg/d, glucose: 2.9 (0.1) vs 2.9 (0.1) g/kg/d; lipids: 2.5 (0.2) vs 2.7 (0.1) g/kg/d;
12–15 g/kg/d and fat: 3–4 g/kg/d. Light protection of TPN was MVP: 1.2 (0.1) vs 1.3 (0.1) ml/kg/d. Figure 2 shows that the
started with the initiation of MVP and lipids in the regimen. volume of enteral feeds was ,5 ml/kg/d and similar between
Photoprotection was started from the moment of preparation in both groups.
the pharmacy and continued throughout the delivery of the
solution to the infant. Shielding from light was achieved using Effect of light exposure on plasma triglyceride concentration
protective covering for the TPN bags and syringes and amber Plasma TG was compared for the first 9 days of life, while on
tubing (Codan Santa Ana, California, USA); therefore, in the LP full TPN. After that age the increasing volume of enteral feeds
group both the amino acid–dextrose solution as well as the lipid
emulsion were shielded from light. This procedure decreases the
amount of infused peroxides.10 As part of monitoring of TPN, TGs
were routinely measured with each increase in lipid intake until
reaching full TPN; blood glucose was measured daily when blood
glucose was in normal range (.2.6 or ,10) or with every change
in glucose infusion when blood glucose was outside this range.
Enteral nutrition was initiated within 48 h of birth in stable
preterm infants according to a protocol described previously.17
Infants who were not tolerating advancement of early enteral
nutrition progressed to full TPN. Daily nutrient intake,
parenteral and enteral, was monitored in LE and LP until full
TPN was achieved, around days 7–9 after birth.
To test whether photoprotection of TPN in preterm infants
affects plasma TGs, a post hoc analysis of TG concentrations
was performed in a subgroup of preterm infants receiving LP or
LE. In view of the influence of enteral nutrition on plasma TG,
only infants on minimal intakes of enteral feeds (,5 ml/kg/d)
were included in this analysis. Plasma accumulation of TG was
obtained by calculating the slope of plasma TG concentrations
(TG, mM) as a function of the amount of lipids received (g/kg/d)
during the preceding 24 h. A similar post hoc analysis of blood
glucose was performed.
Analytical procedures
TG concentrations were determined on 10 ml of plasma using a
colorimetric commercial kit (VITROS Chemical Products TRIG
kit, Ortho-Clinical Diagnostics Inc, Rochester, New York, USA).
Blood glucose was obtained by point of care testing using the
Sure StepFlexH (Life Scan, Canada Ltd, Burnaby, BC, Canada).
In these subjects point of care testing correlated with laboratory
glucose oxidase technique (y = 0.88x+0.1, r2 = 0.95, n = 200).
Statistical analysis
Plasma TG, blood glucose and nutrient intakes (daily averages)
were analysed by factorial ANOVA (days 6 light exposure). Figure 1 Macronutrient intakes as a function of postnatal age. There was
Student’s t test was used to compare slopes. Chi2 was used to no difference in macronutrient intakes between LE and LP. Data expressed
compare proportions of hypoglycaemia and hyperglycaemia as mean ¡ SEM. LE, infants receiving light-exposed parenteral nutrition
determinations. Data are presented as mean ¡ SEM and the (open circle; sample size: 27–32). LP, infants receiving light-protected
threshold of significance was set at p,0.05. parenteral nutrition (dark circle; sample size: 20–27).

Original article
higher in the group receiving LE TPN (1.5 (0.3) vs 0.9 (0.1),

p,0.01 and 1.4 (0.2) vs 0.9 (0.1), p,0.05, respectively).
Effect of light exposure on blood glucose concentration

Overall, mean (SEM) blood glucose measured over the first
9 days of life was significantly higher (p,0.001) in LE (6.6 (0.2)
mM, n = 215) compared to LP (6.0 (0.1) mM, n = 240).
Significant logarithmic relationships between mean daily blood
glucose and postnatal age were found for both LE (y = 1.14Ln(x)
+ 4.75, r2 = 0.62, p,0.05) and LP (y = 0.82Ln(x) + 4.86, r2 = 0.64,
p,0.05) (fig 4) documenting that blood glucose continued to
rise with age in spite of the levelling off of glucose intake after
day 5 of life (fig 1). There was no interaction between light
exposure and days. The number of subjects presenting with
hyperglycaemia (.10 mM) was not significantly different
Figure 2 Enteral feeding volumes as a function of postnatal age. Each between LE (11/32) and LP (6/27). The number of data points
circle represents a sample size of 14–20 infants. There was no difference in the hyperglycaemia range was significantly higher (p,0.05)
between LE and LP. Data expressed as mean ¡ SEM. LE, infants in LE: 103/570 compared to LP: 45/512; however, the number of
receiving light-exposed parenteral nutrition (open circle). LP, infants subjects requiring insulin because of hyperglycaemia was not
receiving light-protected parenteral nutrition (dark circle). significantly different (LE: n = 6/32 vs LP: n = 4/27). When
subjects who required insulin were excluded from this analysis,
(.5 ml/kg/d) given every 1–2 h would interfere with TG in order to isolate the effect of photoprotection on blood glucose
concentrations. The linearity of the relationship between mean from any further outside influence, there was no statistically
plasma TG and postnatal age (y = 0.11x + 0.47) was significant significant difference in mean (SEM) blood glucose between LE
only in infants receiving LE (r2 = 0.78, p,0.01) (fig 3). (6.1 (0.2) mM, n = 161) and LP (5.9 (0.1) mM, n = 212) over the
The influence of iv lipid intake on plasma TG was evaluated first 9 days of life.
by calculating the individual slope of plasma TG (mM) on the
amount of lipids received (g/kg) during the preceding 24 h. The DISCUSSION
mean (SEM) slope for each group was statistically different Failure to shield TPN from light contributed to high blood
from zero (p,0.01) (LE = 0.238 (0.062); LP = 0.095 (0.032)); glucose and plasma TG compared to newborn infants receiving
the mean slope for the LE group was 2.5 times higher than for photoprotected TPN. The same phenomenon observed in
the LP group. animals receiving TPN,14 supports the hypothesis that light
The significant (p,0.05) interaction between light exposure exposure of TPN has an effect on intermediary metabolism. In
and postnatal age allowed us to compare postnatal days, view of concerns of TPN-induced lipid dysregulation, plasma
individually. On days 8 and 9 of life, mean (SD) plasma TG was TG concentration is routinely monitored. An upper limit of
1.7 mM is considered safe in neonates.20 Once this threshold is
reached, further increase in energy provided by the fat emulsion
Figure 4 Blood glucose concentrations as a function of postnatal age.

Figure 3 Plasma triglyceride concentrations as a function of postnatal The factorial ANOVA demonstrated that overall blood glucose
age. The linear relation between TG and postnatal age was significant in concentrations were higher (p,0.05) in LE (y = 1.14 Ln(x) + 4.75,
the LE group (y = 0.11x + 0.47, r2 = 0.78, p,0.01) but not in the LP r2 = 0.62, p,0.05) compared to LP (y = 0.82 Ln(x) + 4.86, r2 = 0.64,
group (y = 0.02x + 0.76, r2 = 0.26). The factorial ANOVA demonstrated p,0.05) and that there was no interaction with days of life. The
a significant (*p,0.05) interaction between LE and day of life. Plasma correlations between average daily blood glucose and postnatal age
triglyceride concentrations were higher in LE group on days 8 were significant (p,0.05). No interaction was found between light
(**: p,0.01) and 9 (*: p,0.05). Data expressed as mean ¡ SEM exposure and days. Data expressed as mean ¡ SEM (sample size =
(sample size = 8–16). LE, infants receiving light-exposed parenteral 9–15). LE, infants receiving light-exposed parenteral nutrition (open
nutrition (open circle). LP, infants receiving light-protected parenteral circle). LP, infants receiving light-protected parenteral nutrition (dark
nutrition (dark circle). circle).

Original article
is curtailed until lipid clearance improves; therefore, shielding enteral nutrition on plasma TG levels in patients who are fed
TPN from light provides a potential benefit for preterm infants. every 2 h, we opted to test the effect of photoprotection on TG
By avoiding hypertriglyceridaemia it will allow for increased in the present subgroup of infants (fig 3) because they received
intake of the lipid emulsion at a time when optimising the full TPN with minimal enteral support (,5 ml/kg/d). It is
provision of energy is important to initiate and sustain reassuring however that we found for the original trial group
growth.21 (LE: n = 58; LP: n = 56) an effect of photoprotection on plasma
The two groups of infants (LE and LP) were similar in terms TG that was in the same direction as in the subgroup on days 8
of gestational age, birth weight, and severity of illness, and 9 (data not shown). On the other hand, there was no
intravenous and enteral feeds; thus, the differences noted in difference in blood glucose between LE and LP in the original
the TG and blood glucose concentrations appear to be related to group.
the degree of light exposure of TPN. Exposure of TPN to light Our study adds to the list of potential complications
generates oxidants in the form of peroxides and results in loss of associated with failure to protect TPN from light exposure16 17
antioxidant vitamins.4 Peroxide concentrations range from 255 and adds credence to the positive impact that photoprotection
to 400 mM in LE TPN solutions, while they range from 100 to of TPN may have on clinical endpoints, especially in this fragile
175 mM3 6 10 16 in photoprotected preparations. The differences population. Ethical limitations related to blood sampling in such
in intermediary metabolism might be associated with differ- a vulnerable population, favour investigation of the mechanisms
ences in peroxides or to other light-induced byproducts underlying these findings in an animal model14; however, the
generated in the TPN solution.22 23 aforementioned complications associated with failure of light
The difference in blood glucose observed between LE and LP is protection of TPN and the impact of being able to increase
statistically significant, but it is questionable whether this is of energy intake early in life beg for further research in the form of
clinical relevance since there was no difference in the need for a multicentre randomised trial to confirm the effects of
insulin or the number of subjects treated for hypoglycaemia; photoprotection of TPN on clinical outcomes in preterm
however, the higher blood glucose observed during LE fits with newborns.
the observations that oxidative stress impairs glucose uptake in
muscle and fat.15 Since this post hoc analysis involves a fairly Funding: This work was supported by the Canadian Institutes of Health Research
(grant: MOP 53270).
small number of subjects it should be viewed as hypothesis
generating. Although the actual mechanisms involved in Competing interests: None.
producing these results are currently unclear, several specula- Ethics approval: The study was approved by the Clinical Research Ethics Board of
tions can be entertained. H2O2 inhibits insulin receptor binding the University of British Columbia, and by the Clinical Research Committee of the
Children’s and Women’s Health Centre of BC.
and insulin receptor autophosphorylation24 proving that
increased reactive oxygen species secretion into peripheral blood Patient consent: Parental written informed consent was obtained prior to enrolment.
is involved in induction of insulin resistance25; therefore, LE
TPN, which is associated with higher H2O2 content, might REFERENCES
induce a state of insulin resistance as blood glucose continues to 1. Hack M, Schlechter M, Cartar L, et al. Growth of very low birth infants to age 20
rise after day 5 of life although glucose intake levels off (fig 4). years. Pediatrics 2003;112:e30–8.
2. Ehrenkranz RA, Dusick AM, Vohr BR, et al. Growth in the neonatal intensive care
The liver plays a central role in lipid homeostasis. It unit influences neurodevelopmental and growth outcomes of extremely low birth
constitutes the major site of lipogenesis and very-low-density weight infants. Pediatrics 2006;117:1253–61.
lipoprotein (VLDL) production involved in lipid transport.12 13 3. Neuzil J, Darlow BA, Inder TE, et al. Oxidation of parenteral lipid emulsion by
The light-induced hepatic triglyceride accumulation14 could ambient and phototherapy lights: potential toxicity of routine parenteral feeding.
J Pediatr 1995;126:785–90.
result from enhanced lipogenesis and/or decreased mitochon- 4. Silvers KM, Sluis KB, Darlow BA, et al. Limiting light-induced lipid peroxidation and
drial beta oxidation of fatty acids26 27 and/or inhibition of vitamin loss in infant parenteral nutrition by adding multivitamin preparations to
triglyceride hydrolase and/or diminished VLDL secretion. The Intralipid. Acta Pediatr 2001;90:242–9.
observed accumulation of plasma TG (fig 3) could be explained 5. Lavoie JC, Bélanger S, Spalinger M, et al. Admixture of a multivitamin preparation to
parenteral nutrition: The major contributor to in vitro generation of peroxides.
by a stimulation of lipogenesis, and/or a lower hepatic uptake. Pediatrics 1997;99:E61–70.
This is supported first by the fact that the difference in plasma 6. Laborie S, Lavoie JC, Chessex P. Paradoxical role of ascorbic acid and riboflavin in
TG was observed only after 1 week on TPN and second by the solutions of total parenteral nutrition: Implication in photoinduced peroxide
generation. Pediatr Res 1998;43:601–6.
observation that the plasma TG/lipid intake ratio increased
7. Laborie S, Lavoie JC, Chessex P. Increased urinary peroxides in newborn infants
faster with LE. The absence of effect of light/peroxides on receiving parenteral nutrition exposed to light. J Pediatr 2000;136:628–32.
triglyceride hydrolase14 suggests that the initial step of transport 8. Lavoie JC, Chessex P. Gender and maturation affect glutathione status in human
of TG through the endoplasmic reticulum to the VLDL is not neonatal tissues. Free Radic Biol Med 1997;23:648–57.
9. Knafo L, Chessex P, Rouleau T, et al. Association between hydrogen peroxyde-
involved, which is clearly supported by elevated plasma TG dependent byproducts of ascorbic acid and increased hepatic acetyl-CoA carboxylase
levels. Plasma cholesterol would have informed on the impact of activity. Clin Chem 2005;51:1462–71.
light on the clearance of lipoproteins in which case both TG and 10. Laborie S, Lavoie JC, Pineault M, et al. Protecting solutions of parenteral nutrition
cholesterol would have been affected. The higher blood glucose from peroxidation. J Parenter Enteral Nutr 1999;23:104–8.
11. Lavoie JC, Rouleau T, Gagnon C, et al. Photoprotection prevents TPN-induced lung
associated with elevated TG is more characteristic of the procollagen mRNA in newborn guinea pigs. Free Radic Biol Med 2002;33:512–20.
metabolic syndrome, which is associated with insulin resistance. 12. Lavoie JC, Laborie S, Rouleau T, et al. Peroxide-like oxidant response in lungs of
Conversely, with lipogenesis from glucose one would have newborn guinea pigs following the parenteral infusion of a multivitamin preparation.
expected to find a decrease in circulating glucose rather than the Biochem Pharmacol 2000;60:1297–303.
13. Chessex P, Lavoie JC, Rouleau T, et al. Photooxidation of parenteral multivitamins
significantly higher blood glucose observed with LE. induces hepatic steatosis in a neonatal guinea pig model of intravenous nutrition.
There is a risk of selection bias when performing a non-pre- Pediatr Res 2002;52:958–63.
specified post hoc analysis on a subgroup. Therefore, a 14. Lavoie JC, Rouleau T, Khashu M, et al. Peroxide contamination of TPN solutions
leads to perturbation of endogenous lipid metabolism and plasma triglyceride
limitation of this report might stem from the fact that results
accumulation in preterm infants. PAS 2005;57:1523 (Abstract).
are derived from a subgroup that represents close to half of the 15. Rudich A, Tirosh A, Potashnik R, et al. Prolonged oxidative stress impairs insulin
original trial group.16 In view of the confounding influence of induced GLLJT 4 translocation in 3T3-L1 adipocytes. Diabetes 1998;47:1562–9.

Original article
16. Khashu M, Harrison A, Lalari V, et al. Photoprotection of parenteral nutrition 22. Lavoie JC, Chessex P, Rouleau T, et al. Light-induced byproducts of vitamin C in
enhances advancement of minimal enteral nutrition in preterm infants. Semin multivitamin solutions. Clin Chem 2004;50:135–40.
Perinatol 2006;30:139–45. 23. Lavoie JC, Rouleau T, Chessex P. Interaction between ascorbate and light-exposed
17. Chessex P, Harrison A, Khashu M, et al. In preterm neonates, is the risk of riboflavin induces lung remodeling. J Pharamcol Exp Ther 2004;311:634–9.
developing bronchopulmonary dysplasia influenced by the failure to protect total 24. Gardner CD, Eguchi S, Reynolds CM, et al. Hydrogen peroxide inhibits insulin
parenteral nutrition from exposure to ambient light? J Pediatr 2007;151:213–14. signaling in vascular smooth muscle cells. Exp Biol Med 2003;228:836–42.
18. Harrison A, Khashu M, Friel J, et al. Variations in metabolic response to TPN are influenced 25. Takeda E, Arai H, Yamamoto H, et al. Control of oxidative stress and metabolic
more by gender than by light exposure. J Pediatr Gastroenterol Nutr 2007;45:577–81. homeostasis by the suppression of postprandial hyperglycemia. J Med Invest
19. Richarson DK, Corcoran JD, Escobar GJ, et al. SNAP-II and SNAPPE-II: Simplified 2005;52:259–65.
newborn illness severity and mortality risk scores. J Pediatr 2001;138:92–100. 26. Romeo C, Eaton S, Quant PA, et al. Neonatal oxidative liver metabolism: Effects of
20. Putet G. Lipid metabolism of the micropremie. Clin Perinatol 2000;27:57–69. hydrogen peroxide, a putative mediator of septic damage. J Pediatr Surg
21. Chessex P, Reichman BL, Verellen GJ, et al. Influence of postnatal age, energy 1999;34:1107–11.
intake, and weight gain on energy metabolism in the very low-birth-weight infant. 27. Fukumoto K, Pierro A, Spitz L, et al. Cardiac and renal mitochondria respond
J Pediatr 1981;99:761–6. differently to hydrogen peroxide in suckling rats. J Surg Res 2003;113:146–50.
Quality & Safety in Health Care

Quality & Safety in Health Care is a leading international peer-review journal in the growing area of
quality and safety improvement. It provides essential information for those wanting to reduce harm and
improve patient safety and the quality of care. The journal reports and reflects research, improvement
initiatives and viewpoints and other discursive papers relevant to these crucial aims with contributions
from researchers, clinical professionals and managers and experts in organisational development and
behaviour.
qshc.bmj.com

Random safety audits in the neonatal unit

L Lee, S Girish, E van den Berg and A Leaf

online 1 Aug 2008;
doi:10.1136/adc.2007.131052

These include:

Topic collections Articles on similar topics can be found in the following collections
Editor's choice (371 articles)
Notes


Original article
Random safety audits in the neonatal unit

L Lee, S Girish, E van den Berg, A Leaf
Neonatal Unit, Southmead ABSTRACT

Hospital, Bristol, UK Background: Random safety audits have been shown to What is already know on this topic
Correspondence to:
be effective in improving standards of practice in high-risk
LLeona Lee, Neonatal Unit, industries. They are process audits rapidly performed Audit is an essential part of clinical governance
University Hospital of North during real-time clinical activity, with immediate feedback, strategy to improve practice.
Staffordshire, Stoke-on-Trent allowing for immediate change of practice.
ST4 5QG, UK; lleonalee@
hotmail.com Aim: Based on a concept described by the Vermont-
Oxford Network, we aimed to introduce random safety
Accepted 18 June 2008 audits to our unit to improve infection control and routine
Published Online First neonatal care. What this study adds
1 August 2008 Method: We designed simple data collection tables to
audit 11 infection control and four routine care standards. c A random safety audit can be performed, and
Audits were undertaken during the weekly grand round. feedback given, in one morning.
Immediate feedback was given. c Use of random safety audits enables high audit
Results: In 6 months we completed three cycles of 15 turnover and loop closure.
audits each. Complete results were available for 14 c Random safety audits can effectively improve
audits. The compliance with the infection control practice.
standards improved from a median of 70% (range 20%–
100%) to 95% (range 66%–100%). The results of the
routine care standards were more variable.
Conclusion: We have shown that this innovative method A formal action plan can be made and circulated
of random safety audits is effective in quickly improving and the standard can be re-audited within a short
practice. We believe this to be due to the instant timeframe — typically weeks. Ursprung and
feedback, continued emphasis on infection control and colleagues have shown that it is feasible to adapt
good clinical practice, and improved teamwork. this process for neonatal intensive care unit
(NICU) practice.4
Our aim was to adapt random safety audits for
Audit is one of the key components of any clinical use in our own neonatal unit and to analyse the
governance strategy1 and is used throughout the effects on practice, focusing particularly on issues
NHS for ensuring maintenance of standards. The related to nosocomial infection.
traditional model of medical audit is often unsa-
tisfactory for junior staff in 6-month posts as there METHOD
is insufficient time to see changes implemented, a We chose 15 audit standards: 11 that were part of
re-audit carried out and ‘‘closure of the audit loop’’. our infection control strategy and four routine care
They thus get little ownership of, or satisfaction standards that were considered suitable for this
from, their audit, which has been shown to be method of audit (table 1).
important2 if we are to avoid the danger of audit The standards were chosen by the medical staff
becoming a chore instead of a useful tool for in consultation with the nursing team to ensure
improving practice. that the audit standards were correct reflections of
We describe the use of ‘‘random safety audits’’ what was recognised to be best practice.
which overcome many of the negative aspects of Southmead is a Level 3 NICU with seven
the traditional audit. They are adapted from intensive care cots, five high-dependency and 14
industry3 where they have been very effective special care cots. The audits were performed by
process audits: checklists are compiled for each of a medical and nursing staff working within the unit.
number of pre-identified error-prone activities. To In order to ease the process and reduce the time
perform the audit, a checklist is chosen at random taken to perform and report the audits, we
and the auditors then go to that point in the designed clear, simple data collection tables and
process to directly engage staff in an immediate found that even complicated guidelines could be
review of the work in progress relative to the translated into a simple-to-use table. Figure 1
checklist endpoints. The idea in industry is that shows the lipid prescription guideline and the table
this will identify error and error-prone situations designed for the audit, as an example.
and increase safety awareness of workers ‘‘on the Two standards were audited each week during
shop floor’’. the grand round when the greatest number of staff
This method of auditing is attractive in clinical would be present. Since this was always the same
practice for many reasons. First, the audits are in day of the week, staff were aware when audits
‘‘real time’’, assessing actual practice. Second, the would be taking place but not which topics were
immediacy of feedback allows for immediate being performed. Standards to be audited were
awareness and change in practice where necessary. chosen in a non-random manner from the 15

Original article
topics. Each was audited once before repeating the cycle of 15 RESULTS
audits. After the first cycle of 15 audits, staff were aware of the In 6 months we completed three cycles of auditing 15 standards
topics to be audited, but not which audits were being carried making a total of 45 audits. Unfortunately, one set of results for
out on a particular day. the leaning topic were mislaid and therefore we report complete
If there was non-compliance with standards this was results for 14 topics.
discussed at the time with the relevant staff member. The results of the audits of infection control standards
Feedback was given in a non-judgemental way and designed showed compliance in the first cycle ranging from 20% to 100%
to encourage compliance with the standard rather than (see table 2). These figures improved or remained the same in all
resentment. All staff were treated in the same way. We also but one of the standards (fig 2). The overall improvement in
tried to engage the staff in discussing the reasons for non- performance expressed as a median and range is shown in fig 3.
compliance to encourage greater ownership of the process. The results of the audits of routine care standards were more
As well as the immediate feedback, at the end of the ward varied and are presented in table 3.
round, we summarised the results and disseminated them by
personally informing the staff, and by use of a template poster DISCUSSION
designed for displaying results on a designated audit notice We have shown that random safety audits can improve
board. For each topic, we had a laminated eye-catching compliance with unit guidelines and protocols. We had greatest
photograph which could be stuck onto the poster to attract success with our infection control standards. The only infection
attention of those walking past. Results were also summarised control standard that appeared to show a marked fall in
in the staff communication book by the end of the day to try compliance was lipid prescribing. However the fall from 100%
and ensure wide dissemination of information. compliance in the second audit to 66% compliance in the third
Any changes to policy were discussed at the monthly audit represented small numbers of 3/3 and 2/3. At the time of
neonatal unit audit meeting. the audits, our unit particularly focused on infection control,
The audits were approved by the North Bristol Trust Audit and the introduction of random safety audits was part of this
Department. campaign. We had chosen to focus on infection control as a
Figure 1 Transformation of a complicated lipid prescription guideline (inset) into an easy-to-use audit table. DIC, disseminated intravascular
coagulation; TPN, total parenteral nutrition; VLBW, very low birth weight.

Original article
Table 1 Audit standards

Infection control standards
There should be no soft toys in intensive care cots
All sleeves should be rolled above the elbows
No wrist watches are to be worn
No rings with stones are to be worn
There should be no leaning on incubators
There should be no foil bowls (used for warming milk) left at the sinks
Lipid should be prescribed as per protocol
There should be a documented antibiotic plan
All stethoscope bells should be kept in the incubator/cot
Each cot should have its own bottle of alcohol gel in its own bracket
Each baby should have its own tube of paraffin (used for heel prick blood tests)
Routine care standards Figure 3 Median and range of results for percentage compliance with
Central venous access should be documented clearly in the notes infection control standards.
First-day checks should be complete
Vitamins and supplements should be prescribed according to protocol
Oxygen saturation limits should be set according to protocol process of review. These discussions led to greater under-
standing of our practice and what our guidelines should be.
Another possible reason for poor compliance in the routine
result of benchmarking in the Vermont Oxford Network and standards was that they were more focused on junior medical
the idea for random safety audits came from the VON 5th staff work — compared to the infection control standards that
Annual Quality Congress in 2004. This pre-existing emphasis were multidisciplinary — and there was a change-over of junior
may partly explain the greater improvement in compliance as staff between the first and second audits. We feel that the quick
compared with the routine care standards. In the context of turn-around and re-auditing by random safety audits had an
infection control, Kilbride et al5 have described the importance important role to play here in highlighting areas for improve-
of a ‘‘habit for change’’ and a ‘‘habit for systems thinking’’, that ment and in stressing our commitment to maintenance of
is, considering the small processes that contribute to the overall standards especially for those who are new to the unit and in
care. We already had a habit for change as part of our infection short-term jobs.
control campaign and the random safety audits were part of the There are possible limitations to using random safety audits.
‘‘habit for systems thinking’’. One is that it requires dedicated staff to maintain the
Compliance with routine care standards did not show the momentum. However, our experience is that these audits are
sustained improvement of the infection control standards. well received and considered to be a very effective form of audit
Although the actual numbers in the results seemed more and therefore, we hope that they would be embraced by other
disappointing, the process still had many positive outcomes. units or departments. Although we concentrated on neonatal
Guidelines were discussed and plans made to clarify and topics, Ursprung et al4 showed that they could be adapted for
formalise changes, such as in oxygen saturation limits and more general measures including the processes of ordering and
first-day checks. For example, the reason for incomplete first- obtaining results of investigations, communication problems
day checks was often that the infant was too unstable for and equipment problems. A second concern may be that they
complete examination on admission. Therefore we planned to are time-consuming but we found that if well prepared, the
review first-day checks on a weekly basis. Confusion in the process of doing the audits was very quick. Dissemination of
oxygen saturation guideline was highlighted and this was in the results also took little time once the results poster was designed
Figure 2 The results of random safety

audits on infection control standards.

Original article
Table 2 Results of the infection control standards Table 3 Results of the routine care audits
Results (compliant/total observations) Results
Audit standard 1st audit 2nd audit 3rd audit Audit standard 1st audit 2nd audit 3rd audit
Soft toys 15/24 17/17 15/15 All babies Fully complete 12/20 9/20 6/17
Sleeves and watches 18/20 20/21 23/24 should have Partially 5/20 9/20 10/17
documented complete
Stethoscope bells 7/7 6/7 12/13
first-day
Hand washing 10/13 14/16 11/13 Not done 3/20 2/20 1/17
checks
Sterilium 21/25 16/17 21/22 Long-line tip 2/3 2/4 1/2
Paraffin 4/7 6/8 6/7 positions
Foil bowls 3/14 17/17 12/12 should be
Rings with stones 17/20 20/21 12/12 documented in
the notes
Lipid protocol 3/5 3/3 2/3
Oxygen Upper limits 2/9 8/8 3/6
Antibiotic use 1/5 4/4 4/4 saturation Lower limits 3/9 5/8 4/6
limits should be
set according
and photographs of the audits were obtained. The most time- to guidelines
consuming area will be collating the results to show trends over Vitamins and 11/13 6/10 20/20
nutritional
time but this is a worthwhile investment for improving patient supplements to
safety. be prescribed
Another potential limitation is that staff may have looked to as per protocol
see what was being audited and then tried to improve their
performance prior to being audited. To minimise this when
We did not directly involve families in these audits but they
performing audits of the environment, we tried to complete
were aware of the processes, particularly regarding hand
audits of all of the relevant areas prior to giving feedback.
washing and jewellery, and reacted positively to the overall
Although we may not have been able to completely rule out any
aim of improving quality of care.
quick alterations prior to our audit, at the very least, the process
In summary the benefits to our unit were: increased
of performing the audit reminded the staff of what the
awareness of infection control measures, improvement in
guidelines should have been and produced a positive change in infection control practices, continued emphasis on infection
behaviour. control and general good clinical practice, clarification of
Perhaps the greatest concern is that random safety audits guidelines and improved team working. We would strongly
may be punitive. This should not be the case if they are done in recommend this form of audit to other units.
the correct spirit of working together to improve the care of
patients. The NHS aims to foster a positive environment to Competing interests: None.
enhance working practices rather than one where blame is
apportioned.6 We did not experience adverse reactions from the REFERENCES
staff. Indeed our experience was that the medical and nursing 1. Halligan A, Donaldson L. Implementing clinical governance: turning vision into reality.
Br Med J 2001;322:1413–17.
staff were very supportive of this style of audit and that they 2. Cook S, Spreadbury P. Audit for a purpose. Physiotherapy 1995;81:182–4.
generated informal conversation both before and after the 3. Juran JM, Gyrna FM. Juran’s quality control handbook. 4th edn. New York: McGraw
audits. Concerns regarding the audits were generally regarding Hill, 1988.
4. Ursprung R, Gray JE, Edwards JD, et al. Real time patient safety audits: improving
the clarity of guidelines and so the audits provided a forum for safety every day. Qual Saf Health Care 2005;14:284–9.
re-writing of guidelines where appropriate. Because of the 5. Kilbride HW, Powers R, Wirtschafter DD, et al. Evaluation and development of
immediate feedback and the demonstration of improvement in potentially better practices to prevent neonatal nosocomial bacteremia. Pediatrics
2003;111:504–18.
adherence to guidelines, medical staff in particular found this 6. Scally G, Donaldson LJ. Clinical governance and the drive for quality improvement in
form of audit to be logical and appealing. the new NHS in England. Br Med J 1998;317:61–5 (4 July).

Enteral feeding regimens and necrotising

casecontrol study
G Henderson, S Craig, P Brocklehurst and W McGuire

online 3 Sep 2007;
doi:10.1136/adc.2007.119560

These include:

Notes


Original article
Enteral feeding regimens and necrotising

case–control study
G Henderson,1 S Craig,2 P Brocklehurst,3 W McGuire4
1
Griffith University, Brisbane, ABSTRACT
Australia; 2 Royal Jubilee Background: Most preterm infants who develop What is already known on this topic
Maternity Hospital, Belfast,
Northern Ireland, UK; 3 National
necrotising enterocolitis (NEC) have received enteral
Perinatal Epidemiology Unit, feeds. Uncertainty exists about which aspects of the c Inter-unit variation in the incidence if NEC in
Oxford, UK; 4 Australian National feeding regimen affect the risk of NEC. preterm infants is not fully explained by case
University, Canberra, Australia Aim: To examine associations between various enteral mix.
feeding practices and the development of NEC in preterm c Enteral feeding regimens for preterm infants
Correspondence to:
W McGuire, Centre for Newborn infants. affect the risk of development of NEC but current
Care, The Canberra Hospital, Methods: Multicentre case–control study. 53 preterm data are insufficient to guide clinical practice.
ACT 2606, Australia; william. infants with NEC were enrolled together with a
mcguire@act.gov.au
gestational age frequency-matched control without NEC
Accepted 8 August 2007
from a randomly selected neonatal unit. Clinical and
Published Online First feeding data were extracted and compared between the What this study adds
3 September 2007 groups.
Results: Significantly fewer cases than controls had c The duration of trophic feeding and rate of
received human breast milk (75% vs 91%; OR 0.32, 95% advancement of enteral feeding may be
CI 0.11 to 0.98). The day on which enteral feeding was modifiable risk factors for the development of
started did not differ significantly (mean (SD) days after NEC in preterm infants.
birth: cases 2.9 (2.8) and controls 2.8 (1.8)). The mean
(SD) duration of trophic feeding (,1 ml/kg/h) was
significantly shorter in the cases (3.3 (3.1) days) than feeding regimens affect the risk of NEC. Studies
controls (6.2 (6.7) days) (mean difference (MD) 22.9, from larger neonatal networks generally use
95% CI 24.9 to 20.9) days. Cases were fully fed routinely collected data but these datasets contain
significantly earlier than controls (mean (SD) days after only limited information on infants’ enteral feed-
birth: cases 9.9 (4.2) and controls 14.3 (9.8); MD 24.4, ing regimens.8 9
95% CI 27.3 to 21.5). This case–control study was undertaken in 10
Conclusions: These data suggest that the duration of independent neonatal centres, where there was
trophic feeding and rate of advancement of feed volumes evidence of marked variation in feeding practices
may be modifiable risk factors for NEC in preterm infants. with respect to type of milk used, timing of
Further randomised controlled trials are warranted to introduction of enteral feeds, and the rate of feed
assess the effect of different rates of feed advancement advancement.10 Controls from a separate unit to
on the incidence of NEC, as well as other outcomes. each individual case were randomly selected to
allow determination of any associations between
the different feeding practices and the development
Most preterm infants who develop necrotising of NEC.
enterocolitis (NEC) will have received enteral feeds.
The timing of the introduction of milk feeds, and METHODS
their rate of advancement, may be important We conducted a case–control study in 10 neonatal
determinants of the risk of NEC.1–3 However, units (appendix 1) in the north of Britain between
prospective studies have so far provided only January 2004 and December 2005. The Northern
limited evidence about the effect of different and Yorkshire multicentre research ethics commit-
enteral feeding strategies on the risk of NEC in tee approved the study.
preterm infants.4–7 Thus there is a need for further Cases were preterm infants (,37 completed
randomised controlled trials to determine how weeks’ gestation) with NEC diagnosed using
different feeding regimens for preterm infants modified Bell criteria or at laparotomy or autopsy
affects their risk of developing NEC. Data from examination (table 1).11
observational studies may inform the development When a case was reported, a participating
of such trials, and identify and prioritise specific neonatal centre was randomly selected and a
interventions for assessment. However, examina- control infant who had not developed NEC was
tion of associations between different feeding identified. Controls were more than 34 weeks’
practices and NEC has been limited for two postmenstrual age at recruitment and therefore
reasons. Studies based in centres or networks unlikely to develop NEC subsequently. Cases and
where feeding protocols for preterm infants are controls were frequency matched for gestational
standardised cannot examine whether different age at birth in one of three bands (,28 weeks,

Original article
Table 1 Case definition of necrotising enterocolitis pressure for more than 4 h), surfactant replacement, or use of
non-steroidal anti-inflammatory (NSAID) therapy for patent
Stage I Abdominal distension or abdominal x ray showing gaseous distension or
frothy appearance of bowel lumen (or both); blood in stool; hypotonia,
ductus arteriosus (PDA) closure were not significantly different
apnoea, or bradycardia (or combination of these) between the groups (table 3). Restricting analyses to the cases
Stage II Abdominal tenderness or rigidity; absent bowel sounds; tissue in stool; with stage II/III NEC (n = 40) did not alter any of these
abdominal x ray showing gas in the bowel wall or portal tree; abnormal findings.
bleeding with trauma; thrombocytopenia; lymphocytopenia
Stage III Marked abdominal distension or rigidity; free gas in the peritoneum;
spontaneous bleeding; coagulopathy; severe metabolic acidosis Feeding practices
All cases had commenced enteral milk feeding prior to diagnosis.
Significantly fewer cases received expressed breast milk (40/53
28–32 weeks, .32 weeks). We obtained parental consent to vs 48/53; odds ratio (OR) 0.32, 95% CI 0.11 to 0.98; p,0.05). Of
access details of the infants’ clinical history, and compared the cases with stage II/III NEC (n = 40), 28 received breast milk
antenatal, perinatal and postnatal clinical risk factors between versus 37 of matched controls (OR 0.19, 95% CI 0.05 to 0.73;
the groups. p,0.05).
The mean (SD) day on which enteral feeding was commenced
RESULTS did not differ significantly between the groups (cases 2.9 (2.8)
We enrolled a total of 53 cases (32 male, 21 female). Thirteen and controls 2.8 (1.8) days after birth). The mean (SD) duration
infants fulfilled the case definition for stage I NEC and 40 for of trophic feeding (,1 ml/kg/h) was significantly shorter in the
stage II/III NEC; 18 cases were confirmed at laparotomy, and 3 cases (excluding seven infants diagnosed while still receiving
at autopsy. The cases were diagnosed at a median postnatal age trophic feeds) (cases 3.3 (3.1) and controls 6.2 (6.7) days; mean
of 15 days (range 2–71 days). Controls matched by gestational difference (MD) 222.9, 95% CI 224.9 to 220.9; p,0.05). Forty-
age band were recruited at a median postnatal age of 54 days two cases achieved full enteral feeds before diagnosis of NEC.
(range 12–144 days). These infants were fully fed significantly earlier than controls
The two groups did not differ significantly with regard to (cases 9.9 (4.2) and controls 14.3 (SD 9.8) days after birth; MD
mean (SD) gestational age at birth (cases 27.9 (3.1) weeks vs 224.4, 95% CI 227.3 to 221.5; p,0.05). The significant
controls 28.0 (2.7) weeks) or birth weight (cases 1114 (427) g vs differences remained when analyses were restricted to cases with
controls 1179 (478) g). Also, there were no significant differ- stage II/III NEC (mean duration of trophic feeding: 2.9 (3.3) days;
ences between the groups with regard to rates of maternal pre- time to full enteral feeds: 9.5 (4.7) days after birth; p,0.05).
eclampsia, diabetes mellitus (including gestational), documen-
ted umbilical arterial absent or reversed end-diastolic flow Effect of type of milk feeding
velocity (AREDFV), history of maternal smoking during The findings were not altered when analyses were stratified by type
pregnancy and exposure to antenatal corticosteroids (more of milk feeding (formula fed versus partially or exclusively breast
than 24 h before delivery), tocolytics or antibiotics within milk fed) (Mantel–Haenszel weighted MD in duration of trophic
1 week before delivery. Only three mothers (two controls and feeding: 22.9 days (95% CI 24.9 to 20.9); and weighted MD in
one case) received co-amoxiclav in the week prior to delivery. time to full enteral feeding: 24.4 days (95% CI 27.3 to 21.5)).
There were no significant differences between the groups with
regard to the incidence of prolonged preterm rupture of the
Effect of AREDFV
membranes (more than 24 h before delivery) or maternal fever
Seven cases and five controls had an antenatal finding of
in labour (table 2).
AREDFV. There were no significant differences between infants
The mean (SD) Apgar scores were not significantly different
with or without documented AREDFV in the time of feed
at 1 min (cases 5.74 (2.47) vs controls 5.73 (2.45)) or 5 min after
commencement, duration of trophic feeding, or time to full
birth (cases 8.09 (1.56) vs controls 7.92 (1.74)). Nor was there
enteral feeding in either the case or the control group. Stratified
any significant difference in the mean (SD) umbilical arterial pH
analyses (AREDFV detected versus undetected) did not alter the
levels (available for 19 cases and controls) of the two groups
significant differences between cases and controls in duration of
(cases 7.25 (0.16) vs controls 7.27 (0.13)).
trophic feeding or time to achieve full feeds.
Postnatal management
The rates of umbilical artery catheter use, mechanical ventila- DISCUSSION
It has long been postulated that differences in enteral feeding
tion (positive pressure ventilation or continuous positive airway
regimens contribute to inter-unit variation in the incidence of
NEC in preterm infants. Multicentre benchmarking studies
Table 2 Antenatal characteristics have suggested that those units which introduce enteral feeding
Cases Controls earlier, and advance feeding volumes more quickly, tend to have
n (%) n (%) OR (95% CI) a higher incidence of NEC.9 However, such studies, using
Maternal pre-eclampsia 10 (19) 14 (26) 0.65 (0.26 to 1.63) routinely collected data, have been unable to examine whether
Diabetes mellitus 2 (4) 3 (4) 0.65 (0.10 to 4.08) the feeding regimens of individual infants are associated with
Documented AREDFV 7 (13) 5 (9) 1.46 (0.43 to 4.93) the risk of developing NEC.
Maternal smoking 19 (36) 19 (36) 1.00 (0.45 to 2.21) This case–control study was undertaken within an informal
Antenatal corticosteroids 34 (64) 38 (72) 0.71 (0.31 to 1.60) collaborative network of neonatal units with a relatively
Tocolytic therapy 5 (9) 4 (8) 1.28 (0.32 to 5.04) homogeneous population but without a cross-unit standardised
Maternal antibiotics 20 (38) 26 (49) 0.63 (0.29 to 1.36) policy for enteral feeding of preterm infants.10 This allowed us
Membranes ruptured .24 h 15 (28) 12 (23) 1.35 (0.56 to 3.25) to examine whether the regimens used to feed individual
Maternal fever (.38 uC) in labour 5 (9) 3 (6) 1.74 (0.39 to 7.67) preterm infants were associated with NEC. Another strength of
AREDFV, arterial absent or reversed end-diastolic flow velocity. this study is that we removed the confounding effect of

Original article
Table 3 Postnatal management inconsistent findings.21 This may be related to differences in

study design, especially with regard to management of
Cases Controls
n (%) n (%) OR (95% CI) confounding variables that are risk factors for NEC. Notably,
none of the studies that have used a study design that
Mechanical ventilation 47 (89) 46 (87) 1.19 (0.37 to 3.82)
accounted for birth weight and gestational age found a
Surfactant replacement 41 (77) 42 (79) 0.89 (0.36 to 2.26)
significant association between AREDFV and NEC.22–24 This
Umbilical artery catheter 25 (47) 22 (42) 1.26 (0.58 to 2.71)
uncertainty may be resolved, and clinical practice better
NSAID treatment for PDA 13 (25) 22 (42) 0.46 (0.20 to 1.05)
informed, when the findings of an ongoing multicentre
NSAID, non-steroidal anti-inflammatory drug; PDA, patent ductus arteriosus. randomised controlled trial comparing early versus delayed
enteral feeding for infants with AREDFV become available (see
gestational age by frequency matching. Since short gestation Abnormal Enteral Doppler Prescription Trial: http://www.
remains the single most important determinant of the risk of npeu.ox.ac.uk/adept/).
NEC in preterm infants, older studies that matched solely for Although these and other data suggest that the more rapid
birth weight may have been subject to bias.12 advancement of enteral feeding volumes beyond trophic feeds is
We found that feeding with human breast milk was associated with a higher risk of developing NEC, a firm practice
associated with lower risk of NEC, consistent with findings recommendation can only be made when sufficient data from
from previous observational studies.13 14 Although this associa- randomised controlled trials are available. The currently
tion may, in part, be due to other confounding variables, recent available trial data indicate that compared with enteral fasting,
meta-analyses of randomised trials have also indicated that trophic feeding reduces the time to full feeding and the length of
feeding preterm infants with breast milk reduces their risk of hospital stay without increasing in the risk of NEC.6 In addition,
developing NEC.15–17 These findings endorse the current practice infants of mothers who express breast milk for early trophic
of encouraging mothers to express breast milk for their preterm feeding are more likely to receive breast milk as their ongoing
infants, and of supporting them to do so with evidence-based principal form of nutrition.25 However, the only trial that has
interventions.18–20 The question of whether donor breast milk is compared trophic feeding with progressive advancement of
the best alternative when maternal milk is not available requires enteral feeds in preterm infants was stopped early because of a
consideration of feasibility, costs, acceptability and the effect on borderline significant higher incidence of NEC in the advanced
other important outcomes, principally nutrient intake, growth feeding group.7 There is a high chance that this represents a
and development. spurious result.26 Furthermore the findings of this single-centre
We did not find any evidence that commencing enteral feeds study are unlikely to be widely generalisable—the trial excluded
within the first few days after birth was associated with the risk small for gestational age infants, enteral feeds were not
of NEC. However, we did find that the subsequent feeding introduced at all until about 10 days after birth in both cases
experience of these infants differed significantly between cases and control groups, and fewer than a third of the study
and controls. Cases received about three days of trophic feeds on participants received breast milk.
average compared with six days in controls. The rate of feeds A large multicentre trial of progressive advancement of
advancement was faster in cases, and infants achieved full enteral feeds versus prolonged trophic feeding appears to be a
enteral feeding on average about 5 days earlier than controls. research priority. Because of the potential for feeding interven-
These findings should be interpreted cautiously. We have tions to affect other competing outcomes (such as duration of
accounted for the major confounding variable, gestational age at use of parenteral nutrition, the risk of nosocomial infection,
birth, by frequency matching our cases with controls. However, length of hospital stay),27 as well as the problems inherent in
although we did not find any significant differences in other minimising bias in (unblinded) feeding studies, it is recom-
potential antenatal and perinatal risk factors between the mended that any future trials should also aim to assess the
groups, unknown confounding variables may have affected the effect on objective outcomes including mortality and longer-
study results. In common with all unblinded studies of enteral term neurological disability.6
feeding in preterm infants, two other sources of bias exist.
First, clinicians may have been more likely to investigate and Acknowledgements: We thank the local investigators in the participating centres: S
Ainsworth, S Bali, B Holland, S Kinmond, N Matta, M Schwager, C Skeoch, B Stenson
diagnose NEC in infants they considered to be at higher risk, for and G Stewart.
example infants fed only formula milk (surveillance bias). A
Funding: The study was funded by Tenovus (Scotland). The funder had no role in the
second major potential source of bias in feeding studies is the collection, analysis and interpretation of data, or in the writing of the report and the
‘‘substrate effect’’. Since the generation of gas in the bowel wall decision to submit the paper for publication. The National Perinatal Epidemiology Unit
(pneumatosis intestinalis) or portal tract requires the presence receives funding from the Department of Health. The views expressed in this
of milk substrate, there may be a tendency to diagnose stage II/ publication are those of the authors and not necessarily those of the Department of
III NEC more often in infants who have received more enteral Health.
feeds.4–6 Our primary analyses therefore included infants with Competing interests: None.
all stages of NEC rather than only those where the diagnosis Ethics approval: The Northern and Yorkshire multicentre research ethics committee
was ‘‘confirmed’’ radiologically. However, the differences we approved the study.
detected in duration of trophic feeding and time to full feeding
persisted when analyses were restricted to infants with stage II/ REFERENCES
III NEC.
We did not find any evidence that an antenatal finding of 1. Cooke RJ, Embleton ND. Feeding issues in preterm infants. Arch Dis Child Fetal
AREDFV was associated with the risk of NEC in this study. Neonatal Ed 2000;83:F215–18.
However, the 95% CI for the effect size was wide because few 2. Williams AF. Early enteral feeding of the preterm infant. Arch Dis Child Fetal
participants in either group had documented AREDFV. Previous Neonatal Ed 2000;83:F219–20.
3. Patole SK, de Klerk N. Impact of standardised feeding regimens on incidence of
observational studies that have examined associations between neonatal necrotising enterocolitis: a systematic review and meta-analysis of
AREDFV and the risk of developing NEC have reported observational studies. Arch Dis Child Fetal Neonatal Ed 2005;90:F147–51.

Original article
4. Bombell S, McGuire W. Delayed introduction of progressive enteral feeds to prevent 19. Fewtrell MS, Loh KL, Blake A, et al. Randomised, double blind trial of oxytocin nasal
necrotising enterocolitis in very low birth weight infants. Cochrane Database Syst Rev spray in mothers expressing breast milk for preterm infants. Arch Dis Child Fetal
2008;(2):CD001970. Neonatal Ed 2006;91:F169–74.
5. McGuire W, Bombell S. Slow advancement of enteral feed volumes to prevent 20. Jones E, Spencer SA. Optimising the provision of human milk for preterm infants.
necrotising enterocolitis in very low birth weight infants. Cochrane Database Syst Rev Arch Dis Child Fetal Neonatal Ed 2007;92:F236–8.
2008;(2):CD001241. 21. Dorling J, Kempley S, Leaf A. Feeding growth restricted preterm infants with
6. Tyson JE, Kennedy KA. Trophic feedings for parenterally fed infants. Cochrane abnormal antenatal Doppler results. Arch Dis Child Fetal Neonatal Ed
Database Syst Rev 2005;(3):CD000504. 2005;90:F359–63.
7. Berseth CL, Bisquera JA, Paje VU. Prolonging small feeding volumes early in life 22. Malcolm G, Ellwood D, Devonald K, et al. Absent or reversed end diastolic flow
decreases the incidence of necrotizing enterocolitis in very low birth weight infants. velocity in the umbilical artery and necrotising enterocolitis. Arch Dis Child
Pediatrics 2003;111:529–34. 1991;66:805–7.
8. Guthrie SO, Gordon PV, Thomas V, et al. Necrotizing enterocolitis among neonates in 23. Wilson DC, Harper A, McClure G. Absent or reversed end-diastolic velocity in the
the United States. J Perinatol 2003;23:278–85. umbilical artery and necrotising enterocolitis. Arch Dis Child 1991;66:1467.
9. Uauy RD, Fanaroff AA, Korones SB, et al. Necrotizing enterocolitis in very low birth 24. Adiotomre PN, Johnstone FD, Laing IA. Effect of absent end diastolic flow velocity
in the fetal umbilical artery on subsequent outcome. Arch Dis Child 1997;76:F35–8.
weight infants: biodemographic and clinical correlates. National Institute of Child Health
25. Schanler RJ, Shulman RJ, Lau C, et al. Feeding strategies for premature infants:
and Human Development Neonatal Research Network. J Pediatr 1991;119:630–8.
randomised trial of gastrointestinal priming and tube-feeding method. Pediatrics
10. Boyle EM, Menon G, Elton R, et al. Variation in feeding practice in preterm and low
1999;103:434–9.
birth weight infants in Scotland. Early Hum Dev 2004;77:125–6.
26. Montori VM, Devereaux PJ, Adhikari NK, et al. Randomized trials stopped early for
11. Walsh MC, Kliegman RM. Necrotizing enterocolitis: treatment based on staging benefit: a systematic review. JAMA 2005;294:2203–9.
criteria. Pediatr Clin North Am 1986;33:179–201. 27. Flidel-Rimon O, Friedman S, Lev E, et al. Early enteral feeding and nosocomial sepsis
12. McKeown RE, Marsh TD, Amarnath U, et al. Role of delayed feeding and of feeding in very low birth weight infants. Arch Dis Child Fetal Neonatal Ed 2004;89:F289–92.
increments in necrotizing enterocolitis. J Pediatr 1992;121:764–70.
13. Beeby PJ, Jeffery H. Risk factors for necrotising enterocolitis: the influence of
gestational age. Arch Dis Child 1992;67:432–5. APPENDIX 1
14. Lucas A, Cole TJ. Breast milk and neonatal necrotising enterocolitis. Lancet
1990;336:1519–23.
15. Chauhan M, Henderson G, McGuire W. Enteral feeding for very low birthweight infants: Participating centres (local investigators)
reducing the risk of enterocolitis. Arch Dis Child Fetal Neonatal Ed 2008;93:F62–6. c Royal Jubilee Maternity, Belfast (Dr S Craig)
16. Boyd CA, Quigley MA, Brocklehurst P. Donor breast milk versus infant formula for c Princess Royal Maternity Hospital, Glasgow (Dr N Matta)
preterm infants: a systematic review and meta-analysis. Arch Dis Child Fetal Neonatal c Queen Mother’s Hospital, Glasgow (Dr B Holland)
Ed 2006;92:F169–75. c Ninewells Hospital, Dundee (Dr M Schwager)
17. Quigley MA, Henderson G, Anthony MY, et al. Formula milk versus donor breast milk c Royal Infirmary, Edinburgh (Dr B Stenson)
for feeding preterm or low birth weight infants. Cochrane Database Syst Rev c Forth Park Hospital, Kirkcaldy (Dr S Ainsworth)
2007:(4):CD0029714. c Antrim Hospital, Antrim (Dr S Bali)
18. Jones E, Dimmock PW, Spencer SA. A randomised controlled trial to compare c Ayrshire Central Hospital, Irvine (Dr S Kinmond)
methods of milk expression after preterm delivery. Arch Dis Child Fetal Neonatal Ed c Royal Victoria Infirmary, Newcastle (Dr S Oddie)
2001;85:F91–5. c Royal Alexandra Hospital, Paisley (Dr G Stewart)

Cytokine gene polymorphisms in preterm infants

study
G Henderson, S Craig, R J Baier, N Helps, P Brocklehurst and W McGuire

online 3 Sep 2007;
doi:10.1136/adc.2007.119933

These include:

Notes


Original article
Cytokine gene polymorphisms in preterm infants

study
G Henderson,1 S Craig,2 R J Baier,3 N Helps,4 P Brocklehurst,5 W McGuire6
c Additional data are published ABSTRACT

online only at http://adc.bmj. Background: The inflammatory cytokine cascade is What is already known on this topic
com/content/vol94/issue2
implicated in the pathogenesis of necrotising enterocolitis
1
Department of Health (NEC). Genetic association studies of cytokine poly- c Inflammatory cytokine cascades may play a key
Sciences, Griffith University, morphisms may help to detect molecular mechanisms
Brisbane, Australia; 2 Regional role in the pathogenesis of necrotising
Neonatal Unit, Royal Jubilee that are causally related to the disease process. enterocolitis (NEC) in preterm infants.
Maternity Hospital, Belfast, UK; Aim: To examine associations between the common c Genetic association studies may help to
3
Department of Paediatrics, genetic variants in candidate inflammatory cytokine genes distinguish causal molecular mechanisms from
University of Manitoba, Canada; and NEC in preterm infants.
4
The Sequencing Service,
epiphenomena.
College of Life Sciences, Methods: Multi-centre case–control and genetic asso-
University of Dundee, UK; ciation study. DNA samples were collected from 50
5
National Perinatal Epidemiology preterm infants with NEC and 50 controls matched for
Unit, University of Oxford, UK;
6 gestational age and ethnic group recruited to a multi- What this study adds
Centre for Newborn Care,
Australian National University, centre case–control study. Ten candidate single-nucleo-
Canberra, Australia tide polymorphisms in cytokines previously associated None of the candidate cytokine polymorphisms so
with infectious or inflammatory diseases were genotyped. far studied is significantly associated with NEC.
Correspondence to: The findings were included in random-effects meta-
Dr W McGuire, Centre for
Newborn Care, The Canberra analyses with data from previous genetic association IL10.5 6 Studies using animal models have sug-
Hospital, ACT 2606, Australia; studies. gested that modulating the inflammatory cytokine
william.mcguire@act.gov.au Results: All allele distributions were in Hardy–Weinberg cascade may be a beneficial adjunctive strategy for
equilibrium. None of the studied cytokine polymorphisms treating preterm infants with NEC.7–9
Accepted 13 August 2007 was significantly associated with NEC. Four previous In investigating the contribution of these com-
8 November 2007
genetic association studies of cytokine polymorphisms plex cytokine cascades to the pathogenesis of NEC,
and NEC in preterm infants were found. Meta-analyses it is difficult to distinguish molecular mechanisms
were possible for several single-nucleotide polymorph- that are causal from those that are epiphenomena
isms. These increased the precision of the estimates of of the disease process. Raised plasma and tissue
effect size but did not reveal any significant associations. concentrations of inflammatory mediators may be
Conclusions: The available data are not consistent with a consequence, rather than the cause, of the
more than modest associations between these candidate disease. An alternative strategy that obviates these
cytokine variant alleles and NEC in preterm infants. Data problems is to examine whether the risk of NEC is
from future association studies of these polymorphisms associated with genetic factors that regulate
may be added to the meta-analyses to obtain more cytokine production or function. This approach
precise estimates of effects sizes. has been used to study the role of inflammatory
mediators in a number of diseases associated with
preterm birth, including NEC.10–12
Necrotising enterocolitis (NEC) is a major cause of Here we have examined associations between
death and disability in preterm infants but its variants in candidate cytokine genes with suscept-
pathogenesis is incompletely understood. Gut ibility to NEC in preterm infants recruited to a
immaturity and sub-optimal gut perfusion, exacer- multi-centre case–control study. We focused on
bated by enteral feeding, may be important. common single-nucleotide polymorphisms (SNPs)
Additional factors, including enteric or systemic that are of likely or proved functional significance,
infection, may then precipitate a cascade of or have plausible disease associations described
inflammatory events leading to the clinical and previously. As genetic association studies in this
pathological end point of NEC.1 field are often limited by small sample sizes and are
Evidence exists for the involvement of several underpowered to exclude modest associations,
host inflammatory mediators in the final common where possible, we have quantitatively synthesised
pathway leading to NEC. Plasma concentrations the study findings with data from other studies in
and tissue expression of the proinflammatory order to provide a more precise estimate of the
cytokines, tumour necrosis factor (TNF), interleu- effect size.13
kin (IL)-1b, IL6, IL8 and IL18, are increased in
infants with NEC.2–4 Proinflammatory effects are METHODS
modulated by receptor antagonists and regulation A case–control study of NEC in preterm infants
of receptors as well as by the effects of anti- was conducted in 10 neonatal units in the north of
inflammatory cytokine cascades, principally Britain between January 2004 and December 2005.

Original article
Table 1 Location of candidate single-nucleotide polymorphism (SNP) Table 2 Variant cytokine genotypes in cases and controls
relative to transcription start site Allele Cases Controls OR (95% CI)
SNP position
Cytokine (major/minor allele) Putative effect of minor allele TNF (2308A) 19 (38%) 17 (34%) 1.19 (0.53 to 2.69)
TNF (2238A) 6 (12%) 9 (18%) 0.62 (0.20 to 1.90)
TNF 2238 (G/A) Increased transcription15* IL1 (231G) 28 (56%) 33 (66%) 0.66 (0.29 to 1.47)
TNF 2308 (G/A) Increased transcription15 16* IL1 (2511T) 30 (60%) 33 (66%) 0.77 (0.24 to 1.74)
IL1b 231 (T/C) Decreased transcription17 IL4R (+1902G) 15 (30%) 13 (26%) 1.22 (0.51 to 2.93)
IL1b 2511 (C/T) Increased transcription17 IL6 (2174C) 37 (74%) 34 (68%) 1.34 (0.56 to 3.19)
IL4 receptor a +1902 (G/A) Enhanced signalling18 IL8 (2251A) 36 (72%) 40 (80%) 0.64 (0.25 to 1.63)
IL6 2174 (G/C) Increased production19 IL10 (21082G) 39 (78%) 38 (76%) 1.12 (0.44 to 2.84)
IL8 2251 (T/A) Increased production20 IL18 (2137C) 20 (40%) 25 (50%) 0.67 (0.30 to 1.47)
IL10 21082 (G/A) Reduced transcription21 IL18 (2607A) 31 (62%) 30 (60%) 1.09 (0.49 to 2.43)
IL18 2137 (G/C) Reduced transcription22
IL, interleukin; IL4R, interleukin 4 receptor; TNF, tumour necrosis factor.
IL18 2607 (C/A) Reduced transcription22
*Cell/stimulus specific.
IL, interleukin; TNF, tumour necrosis factor. were examined. Abstracts presented at the Society for Pediatric
Research and European Society for Pediatric Research between
1995 and 2006 were searched.
Briefly, cases were preterm infants with NEC diagnosed using For each potentially eligible study, information on setting,
modified Bell criteria or at laparotomy or autopsy examination. design, inclusion/exclusion criteria and genotyping method was
Controls were infants who had not developed NEC by extracted. Study investigators were contacted to obtain addi-
34 weeks’ postmenstrual age. Cases and controls were fre- tional information if necessary. Case–control and cohort studies
quency-matched for gestational age at birth. More details of the were eligible for inclusion provided that (a) NEC was defined
methods are published elsewhere.14 The Northern & Yorkshire using standard criteria (Bell staging or modifications), (b) the
multi-centre research ethics committee approved the study. enrolment of participants was not made on the basis of prior
A dried blood sample from each infant was collected on filter knowledge of genotype, (c) genotyping had been blinded to
paper (Whatman FTA, Whatman International, Maidstone, clinical status, (d) the study reported the ethnic ancestry of
Kent, UK). DNA was extracted using the REDExtract-N-Amp participants, (e) the reported genotype distributions were in
Blood PCR kit (Sigma Chemicals, Poole, Dorset, UK). The Hardy–Weinberg equilibrium, and (f) the report provided data
relevant sequences were amplified by PCR, and the products sufficient to calculate an odds ratio (OR). Included data were
sequenced in an automated sequencer using a standard synthesised in random-effects meta-analyses—that is, with no a
SNaPshot run setup (see online supplement for PCR SNP priori assumption of effect homogeneity—using RevMan soft-
primer sequences and PCR run conditions). Alleles were ware (version 4.2). Heterogeneity was assessed using the x2 test
assigned with the ABI GeneMapper software and rechecked (p,0.1 considered significant).
manually. Data were analysed combining individuals homo-
zygous and heterozygous for the variant allele into a single RESULTS
exposure class.
Genetic association study
Ten polymorphisms were genotyped in 50 cases and 50 matched
Meta-analysis controls (table 1). Allele distribution was in Hardy–Weinberg
Medline (1966–2007) and EMBASE (1980–2007) were searched equilibrium for all polymorphisms. The proportion of infants
for genetic association studies using the following text words with the variant allele did not differ significantly between cases
and MeSH terms: [Enterocolitis, Necrotizing OR necrotising and controls for any of the comparisons (table 2). The findings
enterocolitis OR NEC] AND [Polymorphism, Genetic OR did not change when analysis was restricted to infants with
Cytokines/genetics]. References in previous reviews and studies stage 2/3 NEC (n = 38).
Table 3 Cytokine genetic association studies in NEC

Country and
predominant
Reference ethnic groups Participants Design (cases/controls) Alleles studied
23–25 Hungary/Caucasian BW,1500 g Case–control (46/90) TNF (2308A)

NEC = Bell stage 1–3 TNF (2238A)
IL4R (+1902G)
IL6 (2174C)
IL10 (21082G)
IL18 (2607A)
26 USA/Caucasian and BW,1250 g Prospective cohort (39/102) TNF (2308A)
African–American NEC = Bell stage 1–3 TNF (2238A)
27,28 USA/Caucasian and BW,1500 g Retrospective cohort (26/262) TNF (2308A)
African–American Mechanically ventilated IL6 (2174C)
NEC = Bell stage 1–3 IL10 (21082G)
29 Germany/Caucasian GA,32 weeks Retrospective cohort (9/64) TNF (2308A)
NEC = Bell stage 1–3 IL10 (21082G)
BW, birth weight; GA, gestational age; IL, interleukin; IL4R, interleukin 4 receptor; NEC necrotising enterocolitis; TNF, tumour
necrosis factor.

Original article
Figure 1 Meta-analyses (random

effects) of genetic association studies. IL,
interleukin; TNF, tumour necrosis factor.
a, African–American; b, Caucasian.
Systematic literature search was significantly less common in cases than controls (cases 10/
Four genetic association studies of cytokine polymorphisms and 46; controls 37/90; OR 0.40 (95% CI 0.18 to 0.90)).25 We did not
NEC in preterm infants were found (table 3).23–29 find any significant difference in the present study (cases 15/50;
controls 13/50; OR 1.22 (95% CI 0.51 to 2.93)). Meta-analysis of
Meta-analyses combined data did not find a significant association (pooled OR
TNF (2308A), TNF (2238A), IL6 (2174C) and IL10 (21082A) 0.66 (95% CI 0.37 to 1.18)).
Random-effects meta-analyses of data from this study with
those from previous studies did not reveal any significant IL18 (2607A)
differences (fig 1). None of the meta-analyses revealed One study (in a post hoc analysis) reported that the proportion
significant heterogeneity. of infants homozygous for the IL18 (2607A) allele (AA
genotype) was significantly higher in infants with stage 3
IL4 receptor (+1902G) NEC (cases 4/8; controls 7/90; OR 11.9 (95% CI 2.4 to 57.9)).24
The only previous study to have examined the association of the We did not find a significant difference in the present study
IL4 receptor (+1902G) with NEC reported that the variant allele (cases 1/7; controls 7/50; OR 0.77 (95% CI 0.08 to 7.12)). When

Original article
these data were meta-analysed, the association was not surveillance and intervention to prevent NEC in at-risk infants.
statistically significant (pooled OR 3.36 (95% CI 0.21 to 53.6)). In the present study, and on meta-analysis of data from both
studies, we did not find a significant association, suggesting that
DISCUSSION such an approach is not justified at present.
The available data suggest that these common cytokine genetic The same investigators reported an association between
polymorphisms are not strongly associated with the risk of NEC homozygosity of the IL18 (2607A) allele with stage 3 NEC.24
in preterm infants. In most cases, the estimates of effect size However, this association appears to have been the result of a
suggest that modest effects have not been missed. For some post hoc subgroup analysis, and therefore may have been
associations, use of meta-analytical techniques commonly used spurious. We did not detect a similar effect, and meta-analysing
for synthesising the findings of controlled trials allowed us to the data suggests there to be no significant association with
increase the precision of the estimates. Meta-analyses were NEC. These findings support the need to confirm reports of
possible because the studies recruited broadly similar popula- genetic associations in independent populations, particularly if
tions of infants using standard case definitions of NEC as the first reported effect size is modest, derived from a post hoc
inclusion criteria. Controls were selected from ethnically similar or subgroup analysis, and was detected as part of a larger
populations, or data provided to allow stratified analysis by association study where multiple comparisons were made.13
ethnic background to account for the potential confounding Genetic association studies may also be useful in examining
effect of population admixture.13 Although data were pooled the role of other inflammatory mediators in the pathogenesis of
from different study designs, we did not find statistical evidence NEC. Candidates for further investigation include platelet-
of heterogeneity in the meta-analyses, suggesting that these activating factor (a phospholipid with complex biological
estimates are reliable. functions that may be a key mediator in the process leading
The TNF promoter polymorphisms (2308A and 2238A) have to NEC), cyclo-oxygenase-2, and various forms of nitric oxide
been the most commonly studied cytokine gene variants. Our synthase that mediate the downstream vascular effects of the
findings are consistent with those of three previous studies, inflammatory cascade.1 Such studies should aim to study similar
which did not detect any significant associations with populations of infants (using standard case definitions) and use
NEC.23 24 29 Although the estimates of effect size from each measures to avoid confounding, particularly ethnic heterogene-
individual study were wide, meta-analysis of all of these data ity. The use of family-based studies, which allow multiple
suggests that, if any associations do exist, they are likely to be genetic markers to be examined without the possibility of
very modest. This lack of association does not rule out the confounding due to population admixture, may be particularly
possibility that TNF and other proinflammatory cytokines are suitable for investigating complex diseases of preterm infants.
important in the pathogenesis of NEC in preterm infants. True
Acknowledgements: We thank the principal investigators of the cited genetic
associations may exist but with very modest effect sizes too association studies for providing further data for inclusion in the meta-analyses.
small to be excluded by the available data. Alternatively, it may
Funding: The study was funded by Tenovus (Scotland). The funder had no role in the
be that the candidate polymorphisms that we have studied are collection, analysis and interpretation of data, or in the writing of the report and the
not directly involved in gene regulation, or that any functional decision to submit the paper for publication. The National Perinatal Epidemiology Unit
effect is dependent on the haplotypic background for which receives funding from the Department of Health. The views expressed in this
neither we, nor others, have data. publication are those of the authors and not necessarily those of the Department of
Meta-analysis of association studies also suggests that the Health.
common IL6 (2174C) SNP is very unlikely to increase suscept- Competing interests: None.
ibility to NEC. However, the lower bound of the 95% CI for the
OR was consistent with a modest protective effect (halving of REFERENCES
odds). Evidence exists that this polymorphism increases IL6 1. Martin CR, Walker WA. Intestinal immune defences and the inflammatory response
in necrotising enterocolitis. Semin Fetal Neonatal Med 2006;11:369–77.
production in neonatal lymphocytes stimulated with lipopoly- 2. Harris MC, Costarino AT Jr, Sullivan JS, et al. Cytokine elevations in critically ill
saccharide.17 Carriage of the variant is associated with protection infants with sepsis and necrotizing enterocolitis. J Pediatr 1994;124:105–11.
against invasive infection in preterm infants, presumably by 3. Viscardi RM, Lyon NH, Sun CC, et al. Inflammatory cytokine mRNAs in surgical
enhancing the immune response.30 Although it is biologically specimens of necrotizing enterocolitis and normal newborn intestine. Pediatr Pathol
Lab Med 1997;17:547–59.
plausible that a polymorphism that enhances the inflammatory 4. Halpern MD, Holubec H, Dominguez JA, et al. Up-regulation of IL-18 and IL-12 in the
response may increase the risk (or severity) of NEC, another ileum of neonatal rats with necrotizing enterocolitis. Pediatr Res 2002;51:733–9.
possibility is that such a polymorphism may reduce the risk of 5. Edelson MB, Bagwell CE, Rozycki HJ. Circulating pro- and counterinflammatory
cytokine levels and severity in necrotizing enterocolitis. Pediatrics 1999;103:766–71.
NEC by preventing enteric or systemic infections that trigger the 6. Ng PC, Li K, Wong RP, et al. Proinflammatory and anti-inflammatory cytokine
inflammatory cascade. Similarly, the IL10 (21082A) variant has responses in preterm infants with systemic infections. Arch Dis Child Fetal Neonatal
been associated with an increase in susceptibility to late-onset Ed 2003;88:F209–13.
7. Travadi J, Patole S, Charles A, et al. Pentoxifylline reduces the incidence and
invasive infection in preterm infants.28 Meta-analysis of data from severity of necrotizing enterocolitis in a neonatal rat model. Pediatr Res
five studies did not reveal a significant association with NEC, but 2006;60:185–9.
the upper bound of the 95% CI of the OR (2.6) does not exclude a 8. Halpern MD, Clark JA, Saunders TA, et al. Reduction of experimental necrotizing
modest effect size. Data from future association studies may be enterocolitis with anti-TNF-alpha. Am J Physiol Gastrointest Liver Physiol
2006;290:G757–64.
added to this meta-analysis to increase the precision of these 9. Ozturk H, Dokucu AI, Ogun C, et al. Protective effects of recombinant human
estimates of effect size. interleukin-10 on intestines of hypoxia-induced necrotizing enterocolitis in immature
Only two studies have previously reported significant rats. J Pediatr Surg 2002;37:1330–3.
10. Lin HC, Tsai FJ, Tsai CH, et al. Cytokine polymorphisms and chronic lung disease in
associations between cytokine polymorphisms and NEC. small preterm infants. Arch Dis Child Fetal Neonatal Ed 2005;90:F93–4.
Treszl and colleagues found that genetic variation in the IL4 11. Bokodi G, Derzbach L, Banyasz I, et al. Association of interferon gamma T+874A and
receptor (+1902G) was associated with a lower risk of NEC.25 interleukin 12 p40 promoter CTCTAA/GC polymorphism with the need for respiratory
The investigators suggested that the enhanced production of IL4 support and perinatal complications in low birthweight neonates. Arch Dis Child Fetal
Neonatal Ed 2007;92:F25–9.
protected the immature gastrointestinal tract from inflamma- 12. Harding D. Impact of common genetic variation on neonatal disease and outcome.
tion and that screening for this allele would allow targeted Arch Dis Child Fetal Neonatal Ed 2007;92:F408–13.

Original article
13. Hanchard NA. Genetic susceptibility and single-nucleotide polymorphisms. Semin 22. Giedraitis V, He B, Huang WX, et al. Cloning and mutation analysis of the human IL-
Fetal Neonatal Med 2005;10:283–9. 18 promoter: a possible role of polymorphisms in expression regulation.
14. Henderson G, Craig S, Brocklehurst P, et al. Enteral feeding regimens and J Neuroimmunol 2001:112:146–52.
necrotising enterocolitis in preterm infants: multi-centre case-control study. Arch Dis 23. Treszl A, Kocsis I, Szathmari M, et al. Genetic variants of the tumour necrosis factor-
Child Fetal Neonatal Ed 2008;93:in press. alpha promoter gene do not influence the development of necrotizing enterocolitis.
15. Bayley JP, de Rooij H, van den Elsen PJ, et al. Functional analysis of linker-scan Acta Paediatr 2001;90:1182–5.
mutants spanning the 2376, 2308, 2244, and 2238 polymorphic sites of the TNF- 24. Heninger E, Treszl A, Kocsis I, et al. Genetic variants of the interleukin-18 promoter
alpha promoter. Cytokine 2001;14:316–23. region (2607) influence the course of necrotising enterocolitis in very low birth
16. Wilson AG, Symons JA, McDowell TL, et al. Effects of a polymorphism in the human weight neonates. Eur J Pediatr 2002;161:410–11.
tumor necrosis factor alpha promoter on transcriptional activation. Proc Natl Acad Sci 25. Treszl A, Heninger E, Kalman A, et al. Lower prevalence of IL-4 receptor alpha-chain
USA 1997;94:3195–9. gene G variant in very-low-birth-weight infants with necrotizing enterocolitis. J Pediatr
17. Chen H, Wilkins LM, Aziz N, et al. Single nucleotide polymorphisms in the human Surg 2003;38:1374–8.
interleukin-1B gene affect transcription according to haplotype context. Hum Mol 26. Kazzi SNJ, Kim O, Quasney MH. Do alleles of tumour necrosis factor gene affect the
Genet 2006;15:519–29. susceptibility and/or severity of necrotizing enterocolitis? Pediatr Res 2004;55:2754.
18. Kruse S, Japha T, Tedner M, et al. The polymorphisms S503P and Q576R in the 27. Hedberg CL, Adcock K, Martin J, et al. Tumor necrosis factor alpha- 2308
interleukin-4 receptor alpha gene are associated with atopy and influence the signal polymorphism associated with increased sepsis mortality in ventilated very low birth
transduction. Immunology 1999;96:365–71. weight infants. Pediatr Infect Dis J 2004;23:424–8.
19. Kilpinen S, Hulkkonen J, Wang XY, et al. The promoter polymorphism of the 28. Baier RJ, Loggins J, Yanamandra K. IL-10, IL-6 and CD14 polymorphisms and sepsis
interleukin-6 gene regulates interleukin-6 production in neonates but not in adults. Eur outcome in ventilated very low birth weight infants. BMC Med 2006;4:10.
Cytokine Netw 2001;12:62–8. 29. Dordelmann M, Kerk J, Dressler F, et al. Interleukin-10 high producer allele and
20. Hull J, Thomson A, Kwiatkowski D. Association of respiratory syncytial virus ultrasound-defined periventricular white matter abnormalities in preterm infants: a
bronchiolitis with the interleukin 8 gene region in UK families. Thorax preliminary study. Neuropediatrics 2006;37:130–6.
2000;55:1023–7. 30. Harding D, Dhamrait S, Millar A, et al. Is interleukin-6 2174 genotype
21. Turner DM, Williams DM, Sankaran D, et al. An investigation of polymorphism in the associated with the development of septicemia in preterm infants? Pediatrics
interleukin-10 gene promoter. Eur J Immunogenet 1997;24:1–8. 2003;112:800–3.
Stay a step ahead with Online First
We publish all our original articles online before they appear in a print issue. This means that the latest
clinical research papers go straight from acceptance to your browser, keeping you at the cutting edge
of medicine. We update the site weekly so that it remains as topical as possible. Follow the Online First
link on the home page and read the latest research.

Neonatal extracorporeal membrane oxygenation:

practice patterns and predictors of outcome in the
UK
A Karimova, K Brown, D Ridout, W Beierlein, J Cassidy, J Smith, H Pandya, R
Firmin, M Liddell, C Davis and A Goldman

online 1 Oct 2008;
doi:10.1136/adc.2008.141051

These include:

Notes


Original article
Neonatal extracorporeal membrane oxygenation:

practice patterns and predictors of outcome in the UK
A Karimova,1 K Brown,1 D Ridout,2 W Beierlein,1 J Cassidy,3 J Smith,3 H Pandya,4
R Firmin,4 M Liddell,5 C Davis,5 A Goldman1
1
Cardiac Critical Care and ECMO ABSTRACT
unit, Great Ormond Street Objective: To review the UK neonatal extracorporeal
What is already known on this topic
Hospital for Children, London,
UK; 2 Centre for Paediatric
membrane oxygenation (ECMO) service and identify
Epidemiology and Biostatistics, predictors of outcome. c ECMO is a lifesaving therapy for term neonates
Institute of Child Health, London, Design: Retrospective review of the national cohort. with severe acute hypoxic respiratory failure and
UK; 3 Department of PICU and Patients and interventions: 718 neonates received was introduced to the UK in 1989.
ECMO, Freeman Hospital, c Alternate treatments including inhaled nitric
Newcastle upon Tyne, UK;
ECMO for respiratory failure between 1993 and 2005.
4 Measurements and results: Diagnoses were: 48.0% oxide, high frequency oscillation and surfactant
Department of ECMO, Glenfield
Hospital, Leicester, UK; meconium aspiration syndrome (97.1% survivors), 15.9% have been introduced in the past decade, leading
5
Department of Paediatric congenital diaphragmatic hernia (CDH; 57.9% survivors), to reduction of extracorporeal membrane
Surgery and ECMO, Royal 15.9% sepsis (62.3% survivors), 9.5% persistent pul- oxygenation (ECMO) cases worldwide.
Hospital for Sick Children,
Yorkhill, Glasgow, UK monary hypertension (79.4% survivors), 5.6% respiratory
distress syndrome (92.5% survivors) and 5.1% congenital
Correspondence to: lung abnormalities (24.3% survivors). The overall survival
A Karimova, Cardiothoracic Unit, rate of 79.7% compared favourably with the worldwide
Great Ormond Street Hospital for
Children, Great Ormond Street, Extracorporeal Life Support Organization (ELSO) Registry. What this study adds
London WC1N 3JH, UK; Over the period of review, pre-ECMO use of advanced
KarimA@gosh.nhs.uk respiratory therapies increased (p,0.001), but ECMO c Survival rates for neonatal ECMO in the UK
initiation was not delayed (p = 0.61). The use of veno- compare favourably with International Registry
Accepted 9 August 2008 venous (VV) ECMO increased (p,0.001) and average run
Published Online First results.
1 October 2008 time fell (p = 0.004). Patients treated with VV ECMO had c In non-CDH patients, higher OI and older age at
a survival rate of 87.7% compared with 73.4% in the ECMO initiation are identified as two significant
veno-arterial (VA) ECMO group; only 42.4% of those risk factors for death, reinforcing importance of a
needing conversion from VV to VA ECMO survived. In non- timely ECMO referral.
CDH neonates, lower birth weight, lower gestational age,
older age at ECMO and higher oxygenation index (OI)
were associated with increased risk of death. In CDH
neonates, lower birth weight and younger age at ECMO
reduction in the use of ECMO for neonatal ARHF,
were identified as risk factors for death.
reported by the ELSO registry.7–10
Conclusion: The UK neonatal ECMO service achieves
This paper aims to describe and evaluate the use
good outcomes and with overall survival rate reaching
of ECMO for neonatal AHRF in the UK since the
80% compares favourably with international results.
Advanced respiratory therapies are used widely in UK service was set up in 1993. Furthermore, we set out
ECMO patients. Identification of higher OI and older age at to establish the risk factors for death in these
ECMO as risk factors in non-CDH neonates reinforces the patients.
importance of timely referral for ECMO.
METHODS
We included all neonates treated with ECMO for
Extracorporeal membrane oxygenation (ECMO) AHRF in the UK between January 1993 and
was introduced in the UK in 1989 for the December 2005. Neonates with congenital heart
treatment of neonates with acute hypoxic respira- disease were excluded.
tory failure (AHRF). Between 1993 and 1995 the The following data were collected for each
UK Collaborative ECMO Trial was conducted,1 neonate:
showing a marked survival, and survival without c Pre-ECMO characteristics: diagnosis, sex, age
disability, benefit of ECMO over conventional (0–28 days), gestational age, maximum oxyge-
treatment, which has been maintained to 7 years’ nation index (OI), use of iNO, surfactant and
follow up.1 2 As a result of the UK ECMO Trial, the HFOV. Diagnosis was categorised as one of:
Department of Health established a national
– meconium aspiration syndrome (MAS);
service for neonatal ECMO with four centres:
– persistent pulmonary hypertension of the
Glasgow, Newcastle, Leicester and London.
newborn (PPHN);
Over the past 15 years, advanced respiratory
therapies for neonates with AHRF have evolved, in – sepsis (bacterial or viral);
particular high-frequency oscillatory ventilation – respiratory distress syndrome (RDS, pre-
(HFOV), inhaled nitric oxide (iNO) and surfac- viously known as hyaline membrane disease);
tant.3–6 Related to this, there has been worldwide – congenital diaphragmatic hernia (CDH);

Original article
Table 1 Characteristics of the cohort

Veno-
venous
Survivors OI Run time ECMO
Diagnosis No (%) No (%) Median (IQR) median (IQR) No (%)
Meconium aspiration syndrome 345 (48.0) 335 (97.1) 48 (35, 65) 100 (78, 138) 260 (75.4)
Congenital diaphragmatic hernia 114 (15.9) 66 (57.9) 53 (39, 68) 196 (120, 341) 41 (36.0)
Sepsis 114 (15.9) 71 (62.3) 44 (31, 69) 135 (84, 187) 44 (38.6)
Persistent pulmonary hypertension of the 68 (9.5) 54 (79.4) 44 (31, 63) 115 (84, 149) 31 (45.6)
newborn
Respiratory distress syndrome 40 (5.6) 37 (92.5) 42 (35, 52) 108 (86, 161) 18 (45)
Other 37 (5.1) 9 (24.3) 54 (41, 90) 210 (156, 336) 19 (51.4)
ECMO, extracorporeal membrane oxygenation; OI, oxygenation index; IQR, interquartile range.
– ‘‘others’’ (including congenital lung dysplasias such as (IQR 35, 65) and RDS having a median OI of 42 (IQR 35, 52)
surfactant protein B deficiency, alveolar-capillary dyspla- (p = 0.03).
sia and congenital lymphangiectasia). With respect to ECMO support, in 413 (57.6%) patients
c ECMO characteristics: type of ECMO used: veno-venous managed on VV ECMO the survival rate was 362/413 (87.7%)
(VV) or veno-arterial (VA), and length of ECMO run. compared with a survival rate of 193/263 (73.4%) among the
c Outcome measure: survival to hospital discharge. 263 (36.8%) patients who were managed on VA ECMO. Forty
(5.6%) patients needed conversion from VV to VA ECMO; the
survival rate for these was only 17/40 (42.5%). The proportion
of VV ECMO was the highest in MAS patients (75.4% cases)
Summaries are expressed as median values with interquartile
and the lowest in sepsis (38.6%) and CDH (36.3%) patients
ranges (IQRs). For investigation of variation within the cohort,
(table 1). Notably, of those needing conversion from VV to VA,
the Mann–Whitney test was used to compare skewed data
18% were patients from the ‘‘others’’ diagnostic group. The
between groups, and the Kruskal–Wallis test was used where median length of the ECMO run was 119 h (IQR 84, 180). The
there were more than two groups. For frequency data, a x2 test length of ECMO run related to both diagnosis and type of
was used to test for association between two factors and a x2 ECMO (VV or VA) support: VV runs were significantly shorter
test for trend was performed where one of the factors was (109 h (IQR 77, 156)) than VA runs (136 h (IQR 94, 203))
ordered, for example the year of treatment with ECMO. (p,0.001); the differences between the diagnoses are shown in
Spearman rank correlation coefficient was used to test for a table 1.
trend over time for continuous factors. Investigation of risk
factors for death was performed using logistic regression
Risk factors for mortality
analysis and fractional polynomials were used to obtain the
Pre-ECMO risk factors
best fitting model for total length of time on ECMO. Since the
In the non-CDH group of patients (MAS, PPHN, sepsis, RDS
CDH group differed from the rest in terms of ECMO benefit in
and others), lower birth weight, lower gestational age, older age
the UK ECMO trial1 and CDH patients are distinguishable at
at ECMO initiation and higher OI were all associated with
the time of ECMO referral, risk factors for death were analysed increased risk of death. In the CDH patients, only lower birth
separately in the CDH and non-CDH patients. weight and younger age at ECMO initiation were identified as
risk factors for death. The age at which ECMO was initiated
RESULTS proved to be a strong risk factor for death in both non-CDH and
CDH patients, but of note, in opposite directions (table 2).
Cohort characteristics
In total, 718 neonates (399 boys, 319 girls) were treated with
ECMO for AHRF between 1993 and 2005. The median birth ECMO-related risk factors
weight was 3.3 kg (IQR 2.9, 3.7), gestational age 40 weeks (IQR Both the type of ECMO and the length of ECMO run were
38, 41) and patient age when ECMO was initiated was 24 h significantly associated with outcome. In the non-CDH group,
(IQR 20, 50). patients managed on VV ECMO had a better survival rate
The most common diagnosis leading to ECMO was MAS (90.9%) compared with those that had VA ECMO (77.8%)
(n = 345; 48.0% patients), followed by CDH (n = 114; 15.9%), (p,0.001). Unsurprisingly, neonates treated on VA ECMO were
sicker, for example in MAS patients on VA ECMO the OI was
sepsis (n = 114; 15.9%), PPHN (n = 68; 9.5%) RDS (n = 40;
higher (57 (IQR 44, 80)) than in the VV group (45 (IQR 32, 61))
6.6%) and others (n = 37; 5.1%). The spectrum of diagnoses
(p,0.001). In the CDH patients the survival rate was similar
remained fairly constant over the study period except for RDS,
where the relative proportion decreased from 9.4% in 1996 to
5.9% in 2005 (p,0.01). Overall survival to discharge was 79.7%, Table 2 Univariate analysis of pre-ECMO risk factors
with significant differences in survival rates for different Non-CDH CDH
diagnoses (see table 1). (n = 604) Non-CDH (n = 114) CDH
Parameter OR (95% CI) p value OR (95% CI) p value
The degree of respiratory failure was determined by highest
pre-ECMO OI: the median highest OI was 48 (IQR 35, 65); the Birth weight (g) 0.49 (0.33 to 0.74) 0.001 0.43 (0.20 to 0.94) 0.03
highest OI was noted to decrease over the study period Gestational age 0.80 (0.72 to 0.89) ,0.001 1.03 (0.85 to 1.25) 0.75
(weeks)
(p,0.001). There was a difference in the pre-ECMO OI of the
Age (days) 1.17 (1.11 to 1.24) ,0.001 0.73 (0.59 to 0.91) 0.005
different diagnostic groups, with CDH patients having a median
OI (per 5 units) 1.03 (1.00 to 1.07) 0.06 1.01 (0.93 to 1.10) 0.84
OI of 53 (IQR 39, 68), MAS patients having a median OI of 48

Original article
Figure 1 Probability of death plotted against the duration of ECMO run

in hours.
Figure 2 Number of ECMO runs by calendar year: 1989–2005.
regardless of type of ECMO: survival was 58.5% in the VV

ECMO group and 60.9% in the VA ECMO group (p = 0.81). DISCUSSION
Conversion from VV to VA ECMO carried a significant As far as we are aware this is the first detailed review of a
mortality risk for both CDH and non-CDH patients national ECMO service. Our data show that outcome for UK
(p,0.001). The length of ECMO run had significant association patients considered by diagnosis is overall better than that
with outcome in all groups of patients, showing bimodal reported by the international ELSO Registry apart from sepsis
distribution: there was a higher mortality risk with very short (ELSO 75% vs UK 62.3%) and the ‘‘others’’ group (64% vs
runs and with runs of extended length. This reflects early deaths 24.3%).7 The ELSO Registry reports sepsis and pneumonia
in neonates with severe end organ injury and late deaths in (survival 59%) separately, whereas these were combined in our
those who failed to recover despite a long ECMO run (fig 1). dataset. The narrow case definition for the ‘‘others’’ group in
our cohort meant that this group contained only neonates with
Changes in ECMO practice over time irreversible congenital lung dysplasias that have a very poor
Whilst the ECMO trial was in progress (1993–5), approximately outcome.11
30 neonates with AHRF were treated with ECMO per year: this Several US studies have reported the widespread use of
low number reflects the fact that only half were randomised to advanced respiratory therapies prior to ECMO initiation and an
ECMO. Following the trial, the number of runs per year associated reduction in neonatal ECMO cases.8–10 Our study
stabilised from 1996. During the last 2 years of the study, 2004–5, confirms that these treatments are commonly used in UK
the numbers decreased slightly to 53 and 51 ECMO cases per ECMO candidates, although we have not seen a clear downturn
annum, respectively (fig 2). in ECMO cases. There has been concern that the more
With respect to pre-ECMO management, the use of advanced widespread use of iNO and HFOV in the UK might lead to a
respiratory therapies (iNO, HFOV and surfactant) significantly delay in initiation of ECMO support. Given the observed
increased over the study period. In contrast to the beginning of relationship between older age at ECMO initiation in non-CDH
the study period (1993–5), when the use of iNO, HFOV and neonates and higher mortality risk, which was also reported by
surfactant pre-ECMO was minimal, in 2005, iNO use rose to Gill et al12 from the US, this would have been an important
92.2%, HFOV to 60.8% and surfactant to 62.8% (p,0.001). drawback. We were pleased to note that the increased use of
When the potential effects of this were reviewed, it was noted advanced respiratory therapies prior to ECMO in the UK was
that the highest pre-ECMO OI decreased over the study period, associated with improved oxygenation (OI), but no trend
from a median OI of 57 (49, 86) in 1993 to a median OI of 45 towards later initiation of ECMO over time. This is in keeping
(IQR 36, 62) in 2005 (p,0.001). The age when ECMO was with studies in term non-CDH neonates that indicate advanced
initiated did not alter over the study period (p = 0.85). respiratory therapies, particularly iNO, are associated with an
When ECMO practice and outcome were evaluated, it was improvement in OI and a reduction in need for ECMO, but have
noted that ECMO runs became significantly shorter over the no effect on mortality.5 13
study period (p = 0.004) and the use of VV ECMO increased Given the significant changes in pre-ECMO management
significantly (p,0.001). While VV ECMO was used in only over the study period, the population treated during the UK
36.7% of patients during the ECMO trial from 1993 to 1995, VV ECMO trial (1994–5) is likely to have differed to the population
ECMO use increased to 74% between 2003 and 2005 (p,0.001). treated with ECMO in 2005, and this must be considered when
Despite this increase in the use of VV ECMO, the conversion interpreting our data. For instance, the observed reduction in OI
rate from VV to VA ECMO did not change over the study over time could be perceived as a lowering of the threshold for
period (p = 0.59). There was a trend towards increased ECMO and inclusion of ‘‘less sick patients’’. We consider it
mortality over the study period (p = 0.06) and notable annual more likely that the reduction in OI reflects the increased use of
fluctuations: while the survival rate was 82–88% between 1993 iNO and HFOV, and that the babies needing ECMO support in
and 1998, it fell to a minimum of 71.2% in 1999, and then the current era, were in fact sicker than those needing ECMO 10
improved to 78.8% and 80.4% in 2004 and 2005, respectively. years ago. Indeed some authors have suggested that the
We analysed data from 1999 looking for reasons for the worse improvement in oxygenation observed with iNO should be
outcome that year and failed to identify any risk factor. used as a window for safer ECMO cannulation and that

Original article
neonates treated with HFOV and iNO should be considered for our experience the failure to wean from ECMO after several
ECMO at a lower OI than those treated conventionally.9 14 weeks of support usually signifies irreversible lung pathology.
A trend towards increased mortality in neonatal ECMO has While the benefit of continuing ECMO support has to be
been observed by ELSO10 and was also suggested by our data. In carefully judged on individual basis, in those patients who fail to
the ELSO registry, this trend in mortality has been attributed to come off ECMO in an expected time frame every attempt
an increasing proportion of CDH babies treated with ECMO, should be made to reach the correct diagnosis and aggressively
but in the UK this was not the case since the diagnostic case mix treat the underlying pathology.
was stable. We therefore speculate that the mortality trend in
the UK is another reflection of the fact that the neonates who
CONCLUSION
required ECMO despite the current aggressive pre-ECMO
The UK neonatal ECMO service achieves good outcomes when
management strategies had a greater severity of illness.
compared with international results, with the overall survival to
With respect to the support approach, in the UK there has
discharge rate reaching 80%. Advanced respiratory therapies are
been a switch to VV ECMO in recent years. The proportion of
used widely in UK ECMO patients. Identification of higher OI
VV ECMO in the neonatal population differs from the
and older age at ECMO initiation as two risk factors in the non-
worldwide experience and the US experience reported by the
CDH patients reinforces the importance of timely referral for
ELSO registry, which indicate that VA ECMO still remains the
ECMO.
predominant mode of support.7 15 Although VV ECMO has
some theoretical advantages over VA (such as avoidance of Competing interests: The authors have no financial relationship or commercial
instrumentation of a carotid artery) and is associated with an association that might pose a conflict of interest in connection with this article.
improved survival and reduced neurological morbidity com- The study was registered with and granted approval by the Research and
pared with VA, the improved survival for VV ECMO is biased Development Office at the Institute of Child Health, London, UK.
by the selection of less sick neonates for VV compared with VA
ECMO.16 In the UK, VA ECMO is essentially reserved for REFERENCES
infants who cannot be cannulated for VV ECMO for technical 1. UK Collaborative ECMO Trail Group. UK collaborative randomised trial of neonatal
reasons or who have severe myocardial dysfunction and extracorporeal membrane oxygenation. Lancet 1996;348:75–82.
2. McNally H, Bennett CC, Elbourne D, et al. United Kingdom collaborative randomized
cardiovascular instability. It is worth noting that the patients trial of neonatal extracorporeal membrane oxygenation: follow-up to age 7 years.
who require VV to VA conversion have the worst outcome of all Pediatrics 2006;117:e845–54.
groups and it is important to note that the rate of conversion 3. Konduri GG, Solimano A, Sokol GM, et al. A randomized trial of early versus standard
inhaled nitric oxide therapy in term and near-term newborn infants with hypoxic
has not changed over the study period, indicating that selection respiratory failure. Pediatrics 2004;113(3 Pt 1):559–64.
of candidates for VV ECMO over VA ECMO in the UK is fairly 4. Barefield ES, Karle VA, Phillips JB 3rd, et al. Inhaled nitric oxide in term infants with
robust. hypoxemic respiratory failure. J Pediatr 1996;129:279–86.
Our data regarding risk factors for outcome is informative to 5. Clark RH, Kueser TJ, Walker MW, et al. Low-dose nitric oxide therapy for persistent
pulmonary hypertension of the newborn. Clinical Inhaled Nitric Oxide Research Group.
clinicians considering inclusion – exclusion criteria for neonatal N Engl J Med 2000;342:469–74.
ECMO. For example, the higher mortality in CDH neonates 6. Clark RH, Yoder BA, Sell MS. Prospective, randomized comparison of high-frequency
needing very early ECMO initiation likely reflects deaths in oscillation and conventional ventilation in candidates for extracorporeal membrane
oxygenation. J Pediatr 1994;124:447–54.
neonates with severe pulmonary hypoplasia and no tolerance 7. Extracorporeal Life Support Organization. ECMO registry report of the
for mechanical ventilation from birth. These data may help the Extracorporeal Life Support Organization, International summary. Ann Arbor, MI, 2006.
bedside clinician to recognise the CDH neonates with irrever- 8. Hintz SR, Suttner DM, Sheehan AM, et al. Decreased use of neonatal extracorporeal
sible forms of lung hypoplasia and therefore avoid futile ECMO membrane oxygenation (ECMO): how new treatment modalities have affected ECMO
utilization. Pediatrics 2000;106:1339–43.
runs. Following the UK ECMO trial, an OI of 40 has been used 9. Fliman PJ, deRegnier RA, Kinsella JP, et al. Neonatal extracorporeal life support:
as a key threshold for ECMO referral.1 The relationship between impact of new therapies on survival. J Pediatr 2006;148:595–9.
higher OI and mortality in non-CDH neonates reinforces the 10. Roy BJ, Rycus P, Conrad SA, et al. The changing demographics of neonatal
extracorporeal membrane oxygenation patients reported to the Extracorporeal Life
continued relevance of this parameter in the current era: in our Support Organization (ELSO) Registry. Pediatrics 2000;106:1334–8.
study every 5-point increase in OI raised the risk of death by 5%. 11. Cassidy J, Smith J, Goldman A, et al. The incidence and characteristics of neonatal
Furthermore, older age at ECMO initiation proved to be irreversible lung dysplasia. J Pediatr 2002;141:426–8.
strongly associated with poor outcome in non-CDH patients. 12. Gill BS, Neville HL, Khan AM, et al. Delayed institution of extracorporeal membrane
oxygenation is associated with increased mortality rate and prolonged hospital stay.
Similar findings have been noted in other studies particularly in J Pediatr Surg 2002;37:7–10.
patients ventilated for more than a week before ECMO.12 17 18 13. The Neonatal Inhaled Nitric Oxide Study Group (NINOS). Inhaled nitric oxide
This underlines the importance of timely ECMO referral in the and hypoxic respiratory failure in infants with congenital diaphragmatic hernia.
Pediatrics 1997;99:838–45.
non-CDH patients. 14. Kossel H, Bauer K, Kewitz G, et al. Do we need new indications for ECMO in
The length of ECMO run had important association with neonates pretreated with high-frequency ventilation and/or inhaled nitric oxide?
outcome in both CDH and non-CDH patients, showing Intensive Care Med 2000;26:1489–95.
bimodal distribution. We observed a similar relationship 15. Extracorporeal Life Support Organization. ECMO registry report of the
Extracorporeal Life Support Organization, United States trends. Ann Arbor, MI, 2007.
between ECMO run duration and outcome in cardiac ECMO 16. Khambekar K, Nichani S, Luyt DK, et al. Developmental outcome in newborn infants
cases, where early deaths were due to severe end organ damage treated for acute respiratory failure with extracorporeal membrane oxygenation:
and later deaths were predominantly due to failure to recover.19 present experience. Arch Dis Child Fetal Neonatal Ed 2006;91:F21–5.
17. Kugelman A, Gangitano E, Taschuk R, et al. Extracorporeal membrane oxygenation in
In this neonatal cohort, most of the early deaths within 24 h of infants with meconium aspiration syndrome: a decade of experience with venovenous
ECMO initiation were due to withdrawal of ECMO support in ECMO. J Pediatr Surg 2005;40:1082–9.
patients in whom severe brain injury became apparent. Longer 18. Lewis DA, Gauger P, Delosh TN, et al. The effect of pre-ECLS ventilation time on
ECMO runs were also associated with lower survival rates, for survival and respiratory morbidity in the neonatal population. J Pediatr Surg
1996;31:1110–14; discussion 1114–15.
example, the estimated survival rate after 2 weeks on ECMO 19. Chaturvedi RR, Macrae D, Brown KL, et al. Cardiac ECMO for biventricular hearts
dropped to 57% for non-CDH and 42% for CDH patients. In after paediatric open heart surgery. Heart 2004;90:545–51.

Sleeping position, oxygenation and lung function

in prematurely born infants studied post term
T Saiki, H Rao, F Landolfo, A P R Smith, S Hannam, G F Rafferty and A
Greenough
Arch. Dis. Child. Fetal Neonatal Ed. 2009;94;F133-F137

doi:10.1136/adc.2008.141374

These include:

Notes


Original article
Sleeping position, oxygenation and lung function in

prematurely born infants studied post term
T Saiki, H Rao, F Landolfo, A P R Smith, S Hannam, G F Rafferty, A Greenough
MRC-Asthma Centre, Division of ABSTRACT

Asthma, Allergy and Lung Objective: To determine the effect of sleeping position What is already known on this topic
Biology, King’s College London,
UK
on the lung function of prematurely born infants when
post term, whether any effect was similar to that before c Prematurely born infants, particularly if they
Correspondence to: discharge from the neonatal unit, and if it differed sleep prone, have a much higher risk of sudden
Professor A Greenough, Regional according to bronchopulmonary (BPD) status.
Neonatal Intensive Care Centre,
infant death syndrome (SIDS) than infants born
4th Floor Golden Jubilee Wing, Design: Prospective study. at term.
King’s College Hospital, Denmark Setting: Tertiary neonatal unit. c At 36 weeks postmenstrual age (PMA),
Hill, London SE5 9RS, UK; anne. Patients: Twenty infants, median gestational age 30 prematurely born infants have superior lung
greenough@kcl.ac.uk
weeks (range 25–32); 10 had BPD. function (higher oxygen saturation and lung
Interventions: Before neonatal unit discharge (median volume) in the prone compared with the supine
age 36 weeks postmenstrual age (PMA)) and when post position.
term, infants were studied prone and supine, each c There are, however, no data on the effect of
position maintained for 3 h. position on either oxygenation or lung volume in
Main outcome measures: Oxygen saturation was prematurely born infants at the high risk age for
monitored continuously and, at the end of each 3 h period, SIDS (6 weeks post term).
functional residual capacity (FRC) and compliance (CRS) and
resistance (RRS) of the respiratory system were measured.
Results: At a median of 36 weeks PMA and 6 weeks later
What this study adds
(post term), respectively, oxygen saturation (98% vs 96%,
p = 0.001; 98% vs 97%, p = 0.011), FRC (26 vs
24 ml/kg, p,0.0001; 35 vs 31 ml/kg, p = 0.001) and CRS c Oxygenation, lung volume and compliance of the
(3.0 vs 2.4 ml/cm H2O, p = 0.034; 3.7 vs 2.5 ml/cm H2O, respiratory system have been shown to be
p = 0.015) were higher in the prone than the supine significantly higher both before neonatal unit
position. In the prone position, both BPD and non-BPD discharge (at a median of 36 weeks PMA) and
infants had significantly greater FRCs on both occasions and 6 weeks later (post term) in prematurely born
oxygen saturation at 36 weeks PMA, but oxygen saturation infants when nursed prone.
was significantly better post term only in non-BPD infants. c There were, however, no significant correlations
Twelve infants had superior oxygen saturation and 17 between the changes in functional residual
superior FRCs in the prone compared with the supine capacity and changes in oxygen saturation
position at both 36 weeks PMA and post term. levels between positions at either age,
Conclusions: These results suggest that lung function suggesting that the improvement in oxygenation
impairment does not explain why prematurely born infants in the prone position is not due to improved lung
are at increased risk of sudden infant death syndrome in volume; it may be due to enhanced ventilation/
the prone compared with the supine position. perfusion ratios.
c These results suggest that impairment of lung
function does not explain why prematurely born
The association of prone compared with supine infants are at increased risk of SIDS in the prone
sleeping and a higher rate of sudden infant death compared with the supine position.
syndrome (SIDS) is well documented, the odds
ratio for prone and side sleeping compared with
supine sleeping for the last sleep being 13.9.1 In the refute that hypothesis. As a consequence, an aim of
era since the reduce-the-risk campaigns, the this study was to assess the effect of sleeping
adverse effect of the prone sleeping position has position on lung function of very prematurely born
been much greater in prematurely born infants, the infants when post term. By making paired mea-
odds ratio for SIDS being 48.4.1 Yet, we have surements, we also aimed to determine whether
the influence of positioning on lung function post
shown that, at 36 weeks postmenstrual age
term was similar to that before neonatal unit
(PMA), prematurely born infants have significantly
discharge. In addition, we wished to determine if
higher oxygen saturation and functional residual
the influence of positioning differed according to
capacity (FRC) in the prone compared with the
bronchopulmonary dysplasia (BPD) status.
supine position.2 These results suggest that lung
function impairment in the prone position may
not contribute to the risk of SIDS. There are, PATIENTS AND METHODS
however, no data on the effect of sleeping position Protocol
on lung function in prematurely born infants post Infants born before 33 weeks of gestational age
term at the high risk age for SIDS,3 to confirm or who were being prepared for discharge home were

Original article
eligible for entry into the study. Parents were approached and consumption (assumed to be 7 ml/kg/min)4 and for body
the study design explained to them once their infant was temperature, pressure and water vapour-saturated conditions.
tolerating three-hourly feeds. If informed, written parental FRC was measured twice in each position, and the results of the
consent was obtained, the infant was included in the study. The paired measurements were meaned and related to body weight.
study was approved by the King’s College Hospital Foundation The coefficient of repeatability of FRC in non-ventilated infants
Trust Research Ethics Committee. is 3.9 ml/kg.5
Infants were studied on the neonatal unit and subsequently CRS and RRS were assessed using a single occlusion
at follow-up 6 weeks later (post term) in an infant sleep and technique. Air flow was recorded using a pneumotachograph
lung function laboratory. They were studied on both occasions (Mercury F10L; GM Engineering, Kilwinning, UK) inserted into
both supine and prone, each position being maintained for 3 h. the facemask and differential pressure transducer (range:
None of the infants were sedated. The order in which the ¡2 cm H2O, MP45; Validyne Corporation, Northridge,
positions were examined was randomised between infants. California, USA). Airway pressure was measured from a side
Oxygen saturation was continuously monitored using a pulse port on the pneumotachograph using a differential pressure
oximeter (Ohmeda Biox 3740; BOC Health Care, Louisville, transducer (range: ¡100 cm H2O, MP45; Validyne
Colorado, USA) and a reusable infant saturation probe (Flex II). Corporation). The signals were amplified (CD280; Validyne
The accuracy of the Ohmeda Biox oximeter was ¡2.1% Corporation) and displayed in real time on a computer running
between oxygen saturation levels of 80% and 89.9% and Labview software (version 4.0; National Instruments, Austin,
¡1.5% between 90% and 100%. Saturations were recorded Texas, USA) with 100 Hz analog-to-digital sampling (DAQ
continuously over a 3 h period in each position, and mean 16XE-50; National Instruments). Tidal volume was obtained by
saturation was obtained using the Alice 4 sleep study system integration of the flow signal using the Labview software.
(Alice Recording System, Host version 1.8.52; Respironics, Occlusions were performed at end inspiration, which was
Carlsbad, California, USA). As we wished to compare the effect identified from the flow signal. The distal end of the
of posture on oxygen saturation for each 3 h study period, the pneumotachograph was briefly occluded, and only breaths with
averaging time for the pulse oximeter was set to 12 s to a pressure plateau of at least 100 ms were used in the
minimise the influence of motion artefacts and give a better calculation of CRS and RRS. The mean CRS and RRS in each
presentation of oxygen saturation across the study period. The posture were calculated from at least five technically acceptable
neonatal unit’s policy was to keep the oxygen saturation of occlusions, and the results for a particular posture then meaned.
oxygen-dependent infants at a minimum of 92%. To achieve The mean intrasubject coefficients of variation of the CRS and
this, the nurses adjusted the amount of supplementary oxygen RRS measurements were 14% and 19%, respectively.
that the infant received through their nasal cannulae.
At the end of each 3 h period, lung volume and compliance
(CRS) and resistance (RRS) of the respiratory system were
Differences were assessed for statistical significance using the
measured. Lung volume was assessed by measurement of FRC
paired Wilcoxon signed sum rank test or Mann–Whitney U test
using a helium gas dilution technique and a specially designed
as appropriate. Spearman’s correlation coefficients were calcu-
infant circuit (total volume 95 ml).2 The FRC system (Series
lated to determine the strength of any relationships between
7700; Equilibrated Biosystems Inc, Melville, New York, USA)
any differences in FRC and oxygen saturation results between
contained a rebreathing bag as the system reservoir. A facemask
the supine and prone positions at the two study occasions.
(Rendell Baker, Laerdal, Norway) was held snugly over the
infant’s nose and mouth; silicone putty was used around the
mask to achieve a tight seal. The facemask was connected to the Sample size
rebreathing bag via a three-way valve. The FRC system Prematurely born, convalescent infants previously cared for on
contained a helium analyser (Equilibrated Biosystems Inc, the neonatal unit had a mean (SD) oxygen saturation of 93.8
Series 7700) with a digital display. During the measurement, (3.8)% and a mean (SD) FRC of 27.2 (7.3) ml/kg. Recruitment
if there was no change in the helium concentration over a 15 s
period, equilibration was deemed to have occurred. The initial
Table 2 Oxygen saturation and lung volume according to position and
and equilibration helium concentrations were used in bronchopulmonary dysplasia (BPD) status
the calculation of FRC, which was corrected for oxygen
Prone Supine p Value
With BPD
Table 1 Lung function according to position in the study population
SaO2 (%)
Prone Supine p Value Before neonatal unit discharge 94 (92–98) 94 (92–97) 0.032
Before neonatal unit Post term 98 (92–99) 97 (90–99) 0.281
discharge FRC (ml/kg)
SaO2 (%) 98 (92–100) 96 (92–98) 0.001 Before neonatal unit discharge 25 (23–32) 23 (20–25) 0.005
FRC (ml/kg) 26 (23–32) 24 (20–27) ,0.0001 Post term 32 (28–43) 30 (22–35) 0.022
CRS (ml/cm H2O) 3.0 (1.1–5.7) 2.4 (1.4–4.7) 0.034 Without BPD
RRS (cm H2O/l/s) 87.2 (13.5–274.2) 73.3 (19.7–237) 0.698 SaO2 (%)
Post term Before neonatal unit discharge 99 (94–100) 96 (93–98) 0.007
SaO2 (%) 98 (92–100) 97 (90–99) 0.011 Post term 99 (97–100) 97 (96–99) 0.009
FRC (ml/kg) 35 (28–43) 31 (22–31) 0.001 FRC (ml/kg)
CRS (ml/cm H2O) 3.7 (1.8–12.1) 2.5 (1.4–12.8) 0.015 Before neonatal unit discharge 28 (26–32) 25 (22–27) 0.005
RRS (cm H2O/l/s) 71.7 (12.2–152.3) 78.7 (43.2–154) 0.201 Post term 35 (28–38) 32 (27–35) 0.005
Data are median (range). Data are median (range).
CRS, compliance of the respiratory system; FRC, functional residual capacity; RRS, CRS, compliance of the respiratory system; FRC, functional residual capacity; RRS,
resistance of the respiratory system; SaO2, oxygen saturation. resistance of the respiratory system; SaO2, oxygen saturation.

Original article
of 20 patients allowed, with at least 80% power at the 5% level, (table 2). Seventeen infants had higher oxygen saturations and
detection of differences between positions equivalent to at least all 20 had higher FRCs in the prone compared with the supine
one standard deviation of the measurements. position at 36 weeks PMA, compared with 14 infants having
higher oxygen saturations and 19 superior FRCs at 6 weeks post
term (table 3). Differences between the proportion of infants
Patients
who had higher oxygenation (p = 0.257) or FRC (p = 0.317) at
Twenty infants, median gestational age 28.5 weeks (range 25–
36 weeks PMA or post term did not reach statistical signifi-
32) and birth weight 1090 g (range 740–2030), were examined.
cance. Twelve infants had higher oxygen saturations and 17
All had had respiratory distress syndrome and, at initial
infants superior FRCs in the prone compared with the supine
examination, had a median PMA of 36 weeks (range 35–40).
position at both 36 weeks PMA and 6 weeks post term. There
Ten had BPD, being oxygen-dependent beyond 28 days6; they
were, however, no significant correlations between the differ-
were all oxygen-dependent at 36 weeks PMA, but not when
ences in the FRC and oxygen saturation levels between the
studied post term. The infants with BPD were of lower birth
supine and prone positions at a median of 36 weeks PMA
weight than the non-BPD infants (median 909 g (range 740–
(r = 0.393, p = 0.086) or at post term (r = 0.092, p = 0.7).
1590) vs 1285 g (range 744–1992); p = 0.031) and were born at
an earlier gestational age (27 weeks (range 25–30) vs 30 weeks
(range 26–32); p = 0.003). The neonatal unit’s policy was to DISCUSSION
increase the intervals between feeds more slowly in oxygen- We have shown that, post term, oxygenation, lung volume and
dependent infants, and hence the infants with BPD achieved compliance of the respiratory system were significantly greater
three-hourly feeds later and were studied at an older PMA than in the prone compared with the supine position in most of the
the non-BPD infants (median 37 weeks (range 36–40) vs median prematurely born infants examined. In contrast, in healthy term
36 weeks (range 35–37); p = 0.002). There was no significant born neonates studied in the first 72 h after birth, lung volumes
difference between the PMAs at the follow-up (post-term) were similar in the prone and supine positions,7 and, among
measurements of the non-BPD and BPD infants: median infants studied at a median age of 2.4 months, no significant
43 weeks (range 40–52) for infants with BPD vs 43 weeks effect of positioning on oxygenation saturation was noted.8 The
(range 42–49) for non-BPD infants (p = 0.677). The infants with latter study population, however, included infants born at term
BPD were ventilated for longer (median 9.5 days (range 0–40) vs as well as those born prematurely, and, although the infants
0.5 days (range 0–7); p = 0.027), were supported for longer with were all at increased risk of SIDS (previously experienced an
continuous positive airway pressure (median 21.5 days (range acute life-threatening event, inability to fall asleep in their
9–55) vs 1 day (range 0–43); p = 0.007), and received supple- normal position, very premature birth, upper airway obstruc-
mentary oxygen for longer (median 77 days (range 45–131) vs tion, or had a sibling who had died from SIDS), none had
4.5 days (range 0–62); p,0.001) than those who did not develop respiratory distress.8 In another study,9 no significant effect of
BPD. None of the infants were receiving any medication on positioning on transcutaneous oxygen levels was noted overall
either study occasion or were oxygen-dependent when exam- in infants born at term and studied at 2–11 months of age, but
ined post term. There was no significant difference in the there was a small advantage to using the prone position in those
haemoglobin concentrations between infants with and without with a lower respiratory tract infection. None of the infants
BPD at the time of the first examination (median 11.3 g/dl currently studied were oxygen-dependent when examined post
(range 8.7–12.5) vs 10.45 g/dl (range 8.3–12.1); p = 0.449). term, but all had had respiratory distress syndrome and half the
population had had BPD. Thus, a possible explanation for the
differences in position-related effects on oxygen saturation
RESULTS between studies7–9 is the presence or absence of previous or
At a median of 36 weeks PMA and 6 weeks later (post term), current lung disease. The impact of positioning on both lung
respectively, overall the median oxygen saturation (p = 0.001, volume and oxygenation has rarely been examined. In one study
p = 0.011), FRC (p,0.0001, p = 0.001) and CRS (p = 0.034, of ventilated infants and children, no significant effects were
p = 0.015) were higher in the prone than the supine position. seen except for improvements in oxygenation in the infants
There were no significant effects of positioning on RRS at either with obstructive disease.10 The patient population,10 however,
age (table 1). Differences in FRC with respect to position were was older (age range 3–7.6 years) than the infants currently
statistically significant at both ages in both sets of infants (with studied, and all of them were receiving neuromuscular blocking
and without BPD) (table 2). The differences in oxygen agents.
saturation according to position were significantly different In addition to lung volume and oxygenation, CRS and RRS
before neonatal unit discharge in both groups. Post term, were also measured, as we wished to assess postural-related
however, the difference in oxygen saturation according to effects on lung function in prematurely born infants as fully as
position was only statistically significant in the non-BPD group possible. We saw significant changes in CRS, but not RRS, and
at both ages. The likely explanation for the higher compliance
Table 3 Effect of positioning on oxygen saturation and values in the prone compared with the supine position is the
functional residual capacity (FRC) higher lung volume in the prone position. The lack of significant
Before neonatal
difference in the resistance results may reflect that this
unit discharge Post term measurement is less reproducible than the measurement of
CRS, but our sample size was sufficient to detect a 20%
Oxygen saturation 17 (8) 14 (6)
difference in RRS results. In addition, whereas RRS tended to be
FRC 20 (10) 19 (9)
higher (non-significant) in the prone position before neonatal
Oxygen saturation and FRC 17 (8) 13 (5)
unit discharge, the reverse was found post term.
Data are the number of infants with BPD showing higher oxygen
We did not measure the sleep states of the infants. Oxygenation
saturation and/or FRC levels in the prone position. The numbers of
infants with BPD showing higher oxygen saturation and/or FRC in is relatively stable during active and quiet sleep, but indeterminate
the prone position are shown in parentheses. sleep is associated with hypoxaemic episodes.11 The percentage of

Original article
time spent in indeterminate sleep, however, becomes less with function of prematurely born infants persists merits investiga-
increasing maturity and occupies a relatively small amount of tion. Our results are clinically important, as they emphasise to
sleeping time at 36 weeks PMA and post term.11 We have practitioners that prematurely born infants, when studied post
previously shown that indeterminate sleep is significantly more term, may have superior oxygenation in the prone position.
common in the supine than the prone position, but, even in the Infants with acute respiratory insufficiency who can be care-
supine position, indeterminate sleep made up only 3.7% of fully monitored in the hospital environment may benefit from
sleeping time.12 In this study, we examined each infant for 3 h in prone positioning.
both the supine and prone position on each study occasion, and In conclusion, we have shown that lung function and
thus it seems unlikely that the occurrence of indeterminate sleep oxygenation are superior in the prone compared with the
had an important influence on our results. supine position post term in the majority of prematurely born
Our finding that oxygenation and FRC were greater in the infants. These results suggest that impairment of lung function
prone than in the supine position both post term and at does not explain why prematurely born infants are at increased
36 weeks PMA suggests that the higher lung volumes were risk of SIDS in the prone position. An alternative explanation
responsible for the superior oxygenation in the prone position. for the vulnerability of infants in the prone position is
Yet, there were no significant correlations between the changes abnormalities of serotonergic transmission and autonomic
in FRC and oxygen saturation between positions either post dysfunction.18–23 The serotonergic (5 hydroxytryptamine, 5-
term or at a median of 36 weeks PMA. It is possible that our HT) neurons in the medulla oblongata project extensively to
inability to find significant correlations reflects the number of autonomic and respiratory nuclei in the brainstem and help to
infants studied, but Numa et al10 also reported a lack of regulate homoeostatic function.21 Altered 5-HT receptor binding
correlation between changes in FRC and oxygen saturation density in components of the medullary 5-HT system has been
related to position in ventilated infants and children receiving found in SIDS cases in two independent datasets.19 20
neuromuscular blocking drugs. An alternative explanation for Polysomnographic sleep recordings of 18 future SIDS infants
the improvement in oxygenation in the prone position is better showed differences in their sympathovagal balance compared
ventilation perfusion matching. Decreased intrapulmonary with controls.18 In addition, infants with apparent life-
shunting in the prone position has been shown in animal threatening events and obstructive sleep apnoeas have been
models,13 and a more uniform distribution of blood is expected shown to have a reduced heart rate response to 45u head-up
as gravitational forces oppose rather than augment differences tilts,22 and decreased heart rate variability has been demon-
in pulmonary vascular resistance.10 Redistribution of ventila- strated in the prone position.23
tion, however, is the principal mechanism of improved
ventilation perfusion matching in an oleic acid injury lung Funding: TS is, and HR was, supported by the Foundation for the Study of Infant
model.13 In addition, in infants of 32–36 weeks PMA, improve- Deaths.
ment in oxygen saturation in the prone compared with the Competing interests: None.
supine position was associated with less chest wall distortion, Ethics approval: Obtained.
and hence the authors concluded that the improvement in
oxygenation was due to enhanced ventilation/perfusion ratios.14
REFERENCES
Chest wall and abdominal movements were measured by 1. Oyen N, Makestad T, Skjaerven R, et al. Combined effects of sleeping position and
mercury strain gauges taped over the upper rib cage and prenatal risk factors in sudden death syndrome: the Nordic Epidemiological SIDS
abdomen.14 A limitation of the present study was that we did Study. Pediatrics 2007;100:613–19.
not use such a system and thus are unable to comment on 2. Bhat RY, Leipala JA, Singh NR, et al. Effect of posture on oxygenation, lung volume
and respiratory mechanics in premature infants studied before discharge. Pediatrics
whether the improvement in oxygenation demonstrated was 2003;112:29–32.
associated with a reduction in chest wall distortion in the prone 3. Malloy MH, Hoffman JH. Prematurity, sudden infant death syndrome and age of
position. death. Pediatrics 1995;95:464–71.
4. Hey EN. The relationship between environmental temperature and oxygen
None of the infants when studied post term remained
consumption in the newborn baby. J Physiol 1969;200:589–603.
oxygen-dependent. Yet, the effect of positioning post term 5. Dimitriou G, Greenough A, Laubscher B. Lung volume measurements immediately
differed according to BPD status—that is, significantly higher after extubation by prediction of ‘‘extubation failure’’ in premature infants. Pediatr
oxygen saturation levels occurred in the prone position in the Pulmonol 1996;21:250–4.
6. Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med
no-BPD group only. This appears to contrast with our earlier 2001;163:1723–9.
findings,15 16 in which we showed superior oxygenation in the 7. Aiton NR, Fox GF, Alexander J, et al. The influence of sleeping position on functional
prone position only in oxygen-dependent infants. In one of the residual capacity and effective pulmonary blood flow in healthy neonates. Pediatr
studies,15 the infants were only placed in each position for 1 h; Pulmonol 1996;22:342–7.
8. Poets CF, Rudolph A, Neuberk Buch U, et al. Arterial oxygen saturation in infants at
hence, we recommended that infants should be monitored for a risk of sudden infant death syndrome: influence of sleeping position. Acta Paediatr
longer time period to be certain that longer periods of supine 1995;84:379–82.
sleeping were not associated with loss of lung volume and 9. Levene S, McKenzie SA. Transcutaneous oxygen saturation in sleeping infants:
hypoxaemia.17 In the second study,12 as in the present, infants prone and supine. Arch Dis Child 1990;65:524–6.
10. Numa AH, Hammer J, Newth CJL. Effect of prone and supine positions on functional
were examined in each position for 3 h; although effects were residual capacity, oxygenation and respiratory mechanics in ventilated infants and
only statistically significant in one of the two groups, there children. Am J Respir Crit Care Med 1997;156:1185–9.
were trends in both. We therefore suggest that a diagnosis of 11. Lehtonen L, Martin RJ. Ontogeny of sleep and awake states in relation to breathing
in preterm infants. Semin Neonatol 2004;3:229–38.
BPD is not an accurate predictor of the response of lung volume
12. Bhat RJ, Hannam S, Pressler R, et al. Effect of prone and supine position on sleep,
and oxygen saturation to positioning. We also wished to apneas and arousal in preterm infants. Pediatrics 2006;118:101–7.
determine whether the influence of positioning on lung 13. Lamm WJE, Graham MM, Albert RK. Mechanisms by which the prone position
function post term was similar to that at 36 weeks PMA. We improves oxygenation in acute lung injury. Am J Respir Crit Care Med 1994;150:184–93.
14. Martin RJ, Herrell N, Rubin D, et al. Effect of supine and prone positions on
show that most very prematurely born infants have superior
antenatal oxygen tension in the preterm infant. Pediatrics 1979;63:528–31.
lung function post term as well as at 36 weeks PMA in the 15. Kassim Z, Donaldson N, Khetriwal B, et al. Sleeping position, oxygen saturation and lung
prone position. How long this influence of positioning on lung volume in convalescent, prematurely born infants. Arch Dis Child 2006;92:347–50.

Original article
16. Bhat RY, Leipala JA, Rafferty GF, et al. Survey of sleeping position recommendations 20. Panigrahy A, Filiano J, Sleeper LA, et al. Decreased serotonergic receptor binding in
for prematurely born infants on neonatal intensive care unit discharge. Eur J Pediatr rhombic lip-derived regions of the meduall oblongata in the sudden infant death
2003;162:426–7. syndrome. J Neuropathol Exp Neurol 2000;59:377–84.
17. Albert RK, Leasa D, Sanderson M, et al. The prone position improves arterial 21. Paterson DS, Trachtenberg FL, Thompson EG, et al. Multiple serotonergic
oxygenation and reduces shunt in oleic-acid-induced acute lung injury. Am Rev Respir brainstem abnormalities in sudden infant death syndrome. JAMA
Dis 1987;135:628–33. 2006;296:2124–32.
18. Franco P, Szliwowski H, Dramaix M, et al. Decreased autonomic responses to obstructive 22. Harrington C, Kirjavainen T, Teng A, et al. Altered autonomic function and reduced
sleep events in future victims of sudden infant death syndrome. Pediatr Res 1999;46:33–9. arousability in apparent life-threatening event infants with obstructive sleep apnea.
19. Kinney HC, Randall LL, Sleeper LA, et al. Serotonergic brainstem abnormalities in Am J Respir Crit Care Med 2002;165:1048–54.
Northern Plains Indians with the sudden infant death syndrome. J Neuropathol Exp 23. Galland BC, Reeves G, Taylor BJ, et al. Sleep position, autonomic function and
Neurol 2003;62:1178–91. arousal. Arch Dis Child Fetal Neonatal Ed 1998;78:189–94.
Depressed skull fracture in a newborn

baby
A term baby was delivered by emergency caesarean section for
failure to progress with an unstable lie. Forceps or vacuum
extractor was not used. She was born in good condition
requiring no resuscitation. Initial examination revealed a large
depression over the right parietal bone measuring ,6 cm in
diameter and 2 cm in depth (fig 1). A subsequent skull
radiograph revealed a depressed skull fracture involving the
right parietal bone (fig 2). Neurological examination was
normal. Neurosurgical elevation of the fracture was performed
2 weeks after birth with good postoperative recovery and no
residual neurological deficit.
The incidence of bone injuries during delivery has been
described as ,1 in 1000 live births. Of these, clavicular fractures
Figure 2 Anteroposterior skull radiograph showing the right parietal
represent half of all cases, with depressed skull fractures
depressed skull fracture.
accounting for ,11%.1 Intrauterine or obstetric depressed skull
fractures diagnosed at birth may be ‘‘spontaneous’’ or ‘‘instru-
ment associated’’. They have also been described as ‘‘neonatal milk extractor) has also been described.5 Persistent disabilities
skull depressions’’ or ‘‘ping pong’’ fractures and are due to are rare, and fractures following instrumental deliveries are
inward buckling of the calvarial bones without an associated more likely to be associated with intracranial lesions such as
cortical break.2 In babies with no abnormal neurological signs, subdural haematomas.2
expectant treatment has been associated with spontaneous S T Dharmaraj,1 N D Embleton,1 A Jenkins,2 G Jones3
resolution.3 Fracture reduction by neurosurgical elevation
1
should be considered when the depth is more than 2 cm.4 Royal Victoria Infirmary, Newcastle upon Tyne, UK; 2 Department of Neurosurgery,
Newcastle General Hospital, Newcastle upon Tyne, UK; 3 Department of Paediatrics,
Reduction by vacuum extraction (obstetric vacuum or breast
Cumberland Infirmary, Carlisle, Cumbria, UK
Correspondence to: Dr S T Dharmaraj, Royal Victoria Infirmary, Queen Victoria Road,

Newcastle upon Tyne NE1 4LP, UK; s.t.dharmaraj@doctors.org.uk
Competing interests: None.

Patient consent: Parental consent obtained.
Arch Dis Child Fetal Neonatal Ed 2009;94:F137. doi:10.1136/adc.2008.148718
REFERENCES
1. Bhat BV, Kumar A, Oumachigui A. Bone injuries during delivery. Indian J Pediatr
1994;61:401–5.
2. Dupuis O, Silveira R, Dupont C, et al. Comparison of ‘‘instrument-associated’’ and
‘‘spontaneous’’ obstetric depressed skull fractures in a cohort of 68 neonates.
3. Loeser JD, Kilburn HL, Jolley T. Management of depressed skull fracture in the
newborn. J Neurosurg 1976;44:62–4.
4. Hung KL, Liao HT, Huang JS. Rational management of simple depressed skull
fractures in infants. J Neurosurg 2005;103:69–72.
Figure 1 Clinical photograph showing a right parietal neonatal skull 5. Saunders BS, Lazoritz S, McArtor RD, et al. Depressed skull fracture in the neonate.
depression. Report of three cases. J Neurosurg 1979;50:512–14.
Arch Dis Child Fetal Neonatal Ed March 2009 Vol 94 No 2 F137

Depressed skull fracture in a newborn baby

S T Dharmaraj, N D Embleton, A Jenkins and G Jones

doi:10.1136/adc.2008.148718

These include:
Notes


Original article
16. Bhat RY, Leipala JA, Rafferty GF, et al. Survey of sleeping position recommendations 20. Panigrahy A, Filiano J, Sleeper LA, et al. Decreased serotonergic receptor binding in
for prematurely born infants on neonatal intensive care unit discharge. Eur J Pediatr rhombic lip-derived regions of the meduall oblongata in the sudden infant death
2003;162:426–7. syndrome. J Neuropathol Exp Neurol 2000;59:377–84.
17. Albert RK, Leasa D, Sanderson M, et al. The prone position improves arterial 21. Paterson DS, Trachtenberg FL, Thompson EG, et al. Multiple serotonergic
oxygenation and reduces shunt in oleic-acid-induced acute lung injury. Am Rev Respir brainstem abnormalities in sudden infant death syndrome. JAMA
Dis 1987;135:628–33. 2006;296:2124–32.
18. Franco P, Szliwowski H, Dramaix M, et al. Decreased autonomic responses to obstructive 22. Harrington C, Kirjavainen T, Teng A, et al. Altered autonomic function and reduced
sleep events in future victims of sudden infant death syndrome. Pediatr Res 1999;46:33–9. arousability in apparent life-threatening event infants with obstructive sleep apnea.
19. Kinney HC, Randall LL, Sleeper LA, et al. Serotonergic brainstem abnormalities in Am J Respir Crit Care Med 2002;165:1048–54.
Northern Plains Indians with the sudden infant death syndrome. J Neuropathol Exp 23. Galland BC, Reeves G, Taylor BJ, et al. Sleep position, autonomic function and
Neurol 2003;62:1178–91. arousal. Arch Dis Child Fetal Neonatal Ed 1998;78:189–94.
Depressed skull fracture in a newborn

baby
A term baby was delivered by emergency caesarean section for
failure to progress with an unstable lie. Forceps or vacuum
extractor was not used. She was born in good condition
requiring no resuscitation. Initial examination revealed a large
depression over the right parietal bone measuring ,6 cm in
diameter and 2 cm in depth (fig 1). A subsequent skull
radiograph revealed a depressed skull fracture involving the
right parietal bone (fig 2). Neurological examination was
normal. Neurosurgical elevation of the fracture was performed
2 weeks after birth with good postoperative recovery and no
residual neurological deficit.
The incidence of bone injuries during delivery has been
described as ,1 in 1000 live births. Of these, clavicular fractures
Figure 2 Anteroposterior skull radiograph showing the right parietal
represent half of all cases, with depressed skull fractures
depressed skull fracture.
accounting for ,11%.1 Intrauterine or obstetric depressed skull
fractures diagnosed at birth may be ‘‘spontaneous’’ or ‘‘instru-
ment associated’’. They have also been described as ‘‘neonatal milk extractor) has also been described.5 Persistent disabilities
skull depressions’’ or ‘‘ping pong’’ fractures and are due to are rare, and fractures following instrumental deliveries are
inward buckling of the calvarial bones without an associated more likely to be associated with intracranial lesions such as
cortical break.2 In babies with no abnormal neurological signs, subdural haematomas.2
expectant treatment has been associated with spontaneous S T Dharmaraj,1 N D Embleton,1 A Jenkins,2 G Jones3
resolution.3 Fracture reduction by neurosurgical elevation
1
should be considered when the depth is more than 2 cm.4 Royal Victoria Infirmary, Newcastle upon Tyne, UK; 2 Department of Neurosurgery,
Newcastle General Hospital, Newcastle upon Tyne, UK; 3 Department of Paediatrics,
Reduction by vacuum extraction (obstetric vacuum or breast
Cumberland Infirmary, Carlisle, Cumbria, UK
Correspondence to: Dr S T Dharmaraj, Royal Victoria Infirmary, Queen Victoria Road,

Newcastle upon Tyne NE1 4LP, UK; s.t.dharmaraj@doctors.org.uk

Patient consent: Parental consent obtained.
Arch Dis Child Fetal Neonatal Ed 2009;94:F137. doi:10.1136/adc.2008.148718
REFERENCES
1. Bhat BV, Kumar A, Oumachigui A. Bone injuries during delivery. Indian J Pediatr
1994;61:401–5.
2. Dupuis O, Silveira R, Dupont C, et al. Comparison of ‘‘instrument-associated’’ and
‘‘spontaneous’’ obstetric depressed skull fractures in a cohort of 68 neonates.
3. Loeser JD, Kilburn HL, Jolley T. Management of depressed skull fracture in the
newborn. J Neurosurg 1976;44:62–4.
4. Hung KL, Liao HT, Huang JS. Rational management of simple depressed skull
fractures in infants. J Neurosurg 2005;103:69–72.
Figure 1 Clinical photograph showing a right parietal neonatal skull 5. Saunders BS, Lazoritz S, McArtor RD, et al. Depressed skull fracture in the neonate.
depression. Report of three cases. J Neurosurg 1979;50:512–14.

How common are rib fractures in extremely low

birth weight preterm infants?
D Smurthwaite, N B Wright, S Russell, A J Emmerson and M Z Mughal

online 6 Aug 2008;
doi:10.1136/adc.2007.136853

These include:

Notes


Original article
How common are rib fractures in extremely low birth

weight preterm infants?
D Smurthwaite,1 N B Wright,2 S Russell,3 A J Emmerson,4 M Z Mughal1
1
Paediatric Medicine, St. Mary’s ABSTRACT
Hospital for Women and Background: This study was prompted by incidental What is already known on this topic
Children, Manchester, UK;
2
Paediatric Radiology, Royal
finding of healing rib fractures on chest radiographs of ex-
Manchester Children’s Hospital, preterm born infants, who were admitted to hospital with 1. Factures of ribs can occur in very low birth
Manchester, UK; 3 Perinatal & acute respiratory illnesses within a few weeks of weight infants with metabolic bone disease of
Paediatric Radiology, St. Mary’s discharge from the neonatal intensive care unit (NICU). prematurity.
Hospital for Women and Rib fractures in infants, particularly those situated
Children, Manchester, UK; 2. In young infants, posterior rib fractures are
4
Neonatal Medicine, St. Mary’s posteriorly, are considered to be specific for non- considered to be specific of non-accidental injury.
Hospital for Women and accidental injury (NAI).
Children, Manchester, UK Methods: Retrospective examination of radiographs of
extremely low birth weight (ELBW) infants ((1000 g)
Correspondence to: What this study adds
Zulf Mughal, Consultant with a gestation range of 22 of 33 weeks, cared for at a
Paediatrician & Honorary Senior tertiary NICU, between 1998 and 2002, and who had
Lecturer in Child Health, survived >4 weeks. 1. Radiologically apparent rib fractures were
Department of Paediatrics, Saint
Mary’s Hospital for Women &
Results: Five out of 72 (7%) infants studied had present in 7% of extremely low birth weight
Children, Hathersage Road, radiologically apparent rib fractures. None involved posterior preterm infants who had survived >4 weeks.
Manchester, M13 0JH UK; zulf. rib shafts. All infants with rib fractures died on the NICU. 2. No infant had posterior rib fractures and all
mughal@cmmc.nhs.uk Conclusions: The possibility of NAI should be considered infants with rib fractures died on the neonatal
in ex-ELBW infants found to have rib fractures. intensive care unit.
Accepted 11 July 2008
Accept for Online First
6 August 2008
Rib fractures in young infants, particularly those December 2002 were examined for number and
situated posteriorly, are considered to be specific sites of rib fractures. During this 5-year period,
for non-accidental skeletal injury.1 Fractures at this 1284 ELBW infants of 22 to 33 weeks’ gestation
site are thought to occur due to the posterior shafts were admitted to the NICU. Inclusion criteria for
of ribs levering over the transverse processes of the study were ELBW who:
vertebrae, when the chest of a young infant is (a) were born at our Regional Perinatal centre or
manually squeezed.2 This study was prompted by transferred to the NICU within 2 weeks of life
our experience of the incidental finding of healing (b) had survived for >4 weeks after birth
rib fractures on chest radiographs of ex-very low (c) were discharged home from NICU or died on
birth weight infants (VLBW; (1500 g) or extre- the NICU
mely low birth weight infants (ELBW; (1000 g),
(d) had a complete set of radiographs available
who were admitted to hospital with acute respira-
for review
tory illnesses within a few weeks of discharge from
One hundred and six ELBW infants fulfilled
the neonatal intensive care unit (NICU). This
criteria (a) to (c), however, some of these infants
raised the question as to whether rib fractures in
never had chest radiographs taken and in others
these infants arose as a result of non-accidental
radiographs taken could not be found. Chest
injury (NAI), other trauma (for example cardio-
radiographs of 72 ELBW infants who fulfilled the
pulmonary resuscitation (CPR)), or due to meta-
inclusion criteria were examined by a consultant
bolic bone disease of prematurity (MBDP). MBDP
paediatric radiologist (NBW). Any ELBW infant
is a condition of VLBW and ELBW infants,
found to have rib fractures, and randomly selected
characterised by diminished bone strength, due to
radiographs of infants without fractures (in total
the inadequate post-natal provision of bone miner-
20% of all infants studied), had their radiographs
als and/or increased bone resorption. Rib fractures
reviewed by a consultant perinatal & paediatric
have been described in VLBW infants with MBDP.
radiologist (SR). The following clinical information
The incidence of radiologically apparent rib frac-
relevant to causation of MBDP was extracted from
tures among VLBW infants receiving neonatal care
the neonatal medical and nursing case records by
in the mid- to late 1980s ranged from 2.1% to over
one investigator (DS): gestation, birth weight,
30%.3 4 The aim of this retrospective study was to
number of days of total parenteral nutrition
establish the incidence and sites of radiologically
(TPN), number of days of treatment with dexa-
apparent rib fractures in ELBW infants, receiving
methasone (Dx), whether or not the infants
contemporary neonatal care.
developed chronic lung disease of prematurity
(CLD; defined as the need for supplemental oxygen
SUBJECTS AND METHODS .30% at 28 days of age), number of doses of
Chest radiographs of all ELBW infants cared for on frusemide administered and the highest serum
our tertiary NICU between January 1998 and alkaline phosphatise activity. Information on

Original article
number of episodes of CPR and their timing in relation to vertebrae, when the chest of a young infant is squeezed. The
radiological appearance of rib fractures was collected. Similar amount of force required to break ribs is considerable; they
information about insertion of chest drains to drain pneu- occur only rarely in other circumstances such as chest
mothoraces and any episodes of cardiothoracic surgery was also compressions during CPR or in children who have suffered
collected. severe blunt chest trauma in a road traffic accident.5 Although
Statistical analysis was performed using the SPSS (Version much less common, pathological rib fractures can also occur
13) for Windows. The Mann–Whitney U test was used to secondary to conditions that result in bone fragility, such as
compare factors that are relevant to causation of MBDP in severe forms of osteogenesis imperfecta and MBDP.3 4 6
ELBW infants with and without radiologically apparent rib Results of this study suggest that radiologically apparent rib
fractures. fractures are rare in ELBW preterm infants receiving contempor-
ary neonatal care. Furthermore, all the infants with rib fractures
RESULTS in this study died before discharge from the NICU. None of the
One thousand, seven hundred and sixty-two chest radiographs fractures found involved the posterior shafts of ribs, which are
from 72 ELBW infants who met the inclusion criteria were considered to be specific of NAI. We found that infants with rib
examined. Their gestation ranged from 22 to 33 weeks and birth fractures had received significantly higher doses of frusemide. This
weight ranged from 450 to 990 g. Both the radiologists agreed might be due to frusemide contributing to MBDP through causing
that five ELBW infants (7%) had rib fractures. According to loss of minerals from the skeleton. On the other hand, the number
NBW, 2 infants had single and 3 had multiple rib fractures, of doses of frusemide might simply reflect the overall disease
while according to SR, 3 had single and 2 had multiple fractures. severity in ELBW infants with rib fractures.
The disagreement in one subject was caused by abnormal This study has a number of limitations. As this was a
anterior rib ends, which were judged to be healing anterior rib retrospective study, information about factors relevant to
fractures by NBW, while in SR’s opinion the abnormal causation of rib fractures, such as CPR, were obtained from
appearance of the anterior rib ends was thought to be due to perusal of clinical notes. It was not documented in the case
artefact (the tangential direction of the x ray beam in relation to notes whether the CPR was administered by compression of the
the rib ends, thereby producing the appearance of metaphyseal sternum with the infant lying on his/her back, or using the
fraying). All fractures were of anterior or lateral rib shafts; no hand-encircling technique.7 Acute rib fractures may not be
infant had a posterior rib fracture. Skeletons of all ELBW infants detectable on initial chest radiographs but become apparent
with rib fractures and 60 out of 67 (89%) without rib fractures later due to callus formation. Consequently, rib fractures may
were also judged to be osteopaenic by both the radiologists. have been missed in healthy infants in whom there were no
Three of the five ELBW infants with rib fractures had received clinical indications for performing delayed chest radiographs.
CPR, however, there was no temporal relationship between Many of the ELBW infants in the non-rib fracture group
CPR events and appearance of rib fractures on radiographs. survived and so it is likely that towards the end of their stay on
Likewise, there was no temporal relationship between insertion the NICU they were well and would therefore not require
of chest drains and appearance of rib fractures on radiographs. further chest radiographs to be taken. Thus it is entirely possible
All the ELBW infants with rib fractures died of complications that a surviving ELBW infant might suffer a rib fracture on
associated with prematurity prior to discharge from the NICU. discharge from the NICU. This question can only be answered
As shown in table 1, only the number of days of treatment by a prospective study in which all ELBW infants underwent
with frusemide was different between those ELBW infants with chest radiographs prior to, and at regular intervals after
and without rib fractures. The median peak serum alkaline discharge from NICU.
phosphatase activity, number of days of total parenteral In spite of these limitations, we recommend that when an
nutrition and number of days of treatment with Dx approached ELBW preterm infant, after discharge from the NICU is found
statistical significance. to have rib fractures, the possibility of NAI needs to be
considered, irrespective of the neonatal history. This is
particularly the case when the rib fractures are sited posteriorly.
DISCUSSION
Rib fractures in young infants are generally considered to be Acknowledgements: We thank Dr Alan Sprigg and Dr Stephen Chapman for their
specific of non-accidental skeletal injury.1 In an abused child very helpful comments on the manuscript.
they are frequently multiple (although they can be single), Competing interests: None declared.
bilateral and most often sited posteriorly, near the costotrans- Ethics approval: The study was approved by the Central Manchester Local Research
verse process articulation. Injuries at this site occur due to Ethics Committee [Ref. No. 03/CM/343].
posterior shafts of ribs levering over the transverse processes of
REFERENCES
Table 1 Variables relevant to causation of metabolic bone disease of 1. Shouldice M, Huyer D. Non-accidental rib fractures. In: David TJ, ed. Recent
advances in paediatrics 18. Edinburgh: Churchill Livingston, 2000: 63–76.
prematurity in infants with and without rib fractures 2. Kleinman PK, Marks SC Jr, Richmond JM, et al. Inflicted skeletal injury: a post-
Variable p Value mortem radiological-histopathologic study in 31 infants. AJR 1995;165:647–50.
3. Amir J, Katz K, Grunebaum M, et al. Fractures in premature infants. J Paediatr Orthop
Birth weight p = 0.52 1988;8:41–4.
Gestation p = 0.22 4. Koo WK, Sherman R, Succop P, et al. Fractures and rickets in very low birth weight
Number of days of treatment with dexamethasone p = 0.07 infants: conservative management and outcome. J Pediatr Orthop 1989;9:326–30.
5. Worlock P, Stower M, Barbor P. Patterns of fractures in accidental and non-
Number of doses of administered frusemide p = 0.03 accidental injury in children: a comparative study. Br Med J 1986;293:100–2.
Highest serum alkaline phosphatase activity p = 0.08 6. Bishop N, Sprigg A, Dalton A. Unexplained fractures in infancy: looking for fragile
Whether or not the infant developed chronic lung disease (need for p = 0.36 bones. Arch Dis Child 2007:92:189–282.
supplemental oxygen .30% at 28 days of age) 7. Mackway-Jones K, Molynenx E, Philips B, et al, eds. Advanced paediatric life
Number of days of total parenteral nutrition p = 0.08 support: the practical approach: In: Advanced life support group. Manchester, UK:
Blackwell BMJ Books, 2005.

Mortality of twin and singleton livebirths under 30

weeks’ gestation: a population-based study
B Ray and M P Ward Platt

online 6 Oct 2008;
doi:10.1136/adc.2008.143016

These include:

Notes


Original article
Mortality of twin and singleton livebirths under

30 weeks’ gestation: a population-based study
B Ray,1 M P Ward Platt2
1
Royal Victoria Infirmary, ABSTRACT
Newcastle upon Tyne, UK;
2
Objective: To determine the mortality rates of liveborn What is already known on this topic
Newcastle University,
Newcastle upon Tyne, UK
twins compared with singletons of less than 30 weeks’
gestation in relation to gestational age, mode of delivery c There is contradictory evidence about the
Correspondence to: and year of birth in a geographically defined population. differential mortality of very preterm twins and
M P Ward Platt, Newcastle Study design: Comparison of early neonatal, late singletons.
University, Newcastle upon
Tyne, UK; m.p.ward-platt@ncl. neonatal and infant death rates in 479 twin babies and c Monochorionic twins are thought to be at
ac.uk 1538 singletons, liveborn between 23 and 29 completed increased risk of neonatal death compared to
weeks of gestation in the north of England over two dichorionic twins.
Accepted 22 August 2008 epochs, 1998–2001 and 2002–5.
Published Online First Results: Twins and singletons had similar mortality rates
6 October 2008
except at the extreme of gestation (23–25 weeks) where
twins had higher infant mortality (OR 2.04, 95% CI 1.37 to
3.02). This higher rate was attributable to early and late What this study adds
neonatal deaths (OR 1.86, 95% CI 1.28 to 2.72, and 2.11,
95% CI 1.13–3.94, respectively). When analysed in two c The excess mortality among twins of less than
epochs, the excess mortality was confined to babies born 30 weeks’ gestation was confined to neonatal
in 1998–2001. There was no effect of gender or deaths among babies of 25 weeks’ or less
chorionicity. gestation.
Conclusions: The excess mortality among twins of less c The excess mortality among twins of less than
than 30 weeks’ gestation was confined to neonatal 25 weeks’ gestation was confined to the epoch
deaths in babies of 25 weeks or less, and to the earlier 1998–2001.
epoch (1998–2001). In the modern era, there appears to c Chorionicity has no effect on mortality among
be no excess mortality in neonates less than 30 weeks’ liveborn very preterm twins.
gestation when compared with singletons.
The UK Birth Statistics in 2004 showed that the in liveborn twins and singletons born at less than
likelihood of women having multiple births was 30 completed weeks in a defined geographical area.
higher at every age in 2004 than 10 years’ We wished to ascertain any associations related to
previously, and data from our own region are in mode of delivery, gender and chorionicity. We also
line with this continuing trend.1 Women aged 40 investigated any changes in gestation-specific
and over experienced the highest multiple mater- mortality over the time course of the study
nity rate (21.6/1000 all maternities).2 It is known (1998–2005).
that multiple pregnancies are associated with a
higher risk of perinatal death than singletons, METHODS
which may be the result of many factors,3–5 but To ascertain the twins we used the Multiple
many of the excess deaths are ultimately due to Pregnancy Register.1 To ascertain all deaths we
prematurity.6 7 used data from the Perinatal Morbidity and
Some of the existing studies of twinning in Mortality Survey.15 The registers have been
relation to gestational age-specific mortality show approved by ethics committees and have clearance
that extremely premature twins do not suffer more from the Patient Information Advisory Group
deaths than singletons,8–10 but others have shown (PIAG) to hold named patient information without
an increased mortality compared with single- individual consent under section 60 of the Health
tons.11 12 Furthermore, birth weight does not and Social Care Act (2001).
necessarily account for any twin disadvantage in Denominator data on numbers of singleton
very preterm babies. Garite et al13 demonstrated livebirths were obtained from the Regional data-
that statistically significant differences between base of neonatal admissions for intensive care held
mean weights of twins compared with singletons at the neonatal unit of Royal Victoria Infirmary,
occur only from 32 weeks although Alexander et Newcastle upon Tyne. This database collects audit
al14 suggested the difference may start at 28 weeks data for all admissions for neonatal intensive care
of gestation. in the four regional neonatal intensive care units.
Most of the deaths in preterm babies now occur We also enlisted the help of the special care units in
at gestational ages of 29 weeks or less, with very the other peripheral hospitals in the region to
low mortality from 30 weeks onwards.13 We there- ascertain the existence of any babies under
fore focused this study on the mortality outcome 30 weeks who were not transferred to one of the

Original article
Figure 1 Distribution of cases.
Table 1 Demographic comparison between two groups presented as absolute numbers and rates per 1000 livebirths or
Twins Singletons per 1000 survivors where appropriate. Differences in outcome
were calculated using the Fisher exact test and are presented as
28–29 weeks (n (%)) 208 (43) 665 (43)
odds ratio and confidence intervals. SPSS V.13.0 and Epi-info
26–27 weeks (n (%)) 119 (25) 475 (31)
statistical software were used to analyse the data.
23–25 weeks (n (%)) 152 (32) 398 (26)
Total (n (%)) 479 (100) 1538 (100)
Gestational age (median) 27 27 RESULTS
Birth weight (median (range)) 980 (300–2290) 990 (183–2025) In the 8-year period from January 1998 to December 2005, 2207
Male gender (%) 56 54 babies were born alive at between 23 weeks and 29 weeks of
Deliver by caesarean section 41 41 gestation in the region; 34 higher order multiple births were
(%)
excluded from the study. In the remaining 2173 liveborn babies,
1674 were singletons and 499 were twins. A total of 156 babies
(singletons = 136, twin I = 7 and twin II = 13) were excluded
four intensive care units and who would not have been captured
for congenital anomalies. This left a total of 479 twins (twin I
by the routine audit database; however, no patient identifiable
= 245 and twin II = 234) who were compared with 1538
data were used nor did we access any health records. singletons (fig 1).
The geographical area is that of North Cumbria, The characteristics of the study population are presented in
Northumberland, Tyne and Wear, Durham, Darlington and table 1. The median gestations, birth weights, gestation
Teesside. We used data from January 1998 to December 2005 distribution, gender ratio and rate of caesarean section delivery
and excluded all babies with significant congenital anomalies by are comparable between the groups. There was a male
cross-checking with the Northern Congenital Abnormality preponderance of 56% in twins and 54% in singletons.
Survey (NORCAS) and using the definitions of the European After subdividing into the gestation bands 23–25, 26–27, and
Surveillance of Congenital Anomalies (EUROCAT) central 28–29 weeks, the mortality outcome was computed for 0–
registry (see http://www.eurocat.ulster.ac.uk/pdf/EUROCAT- 7 days, 7–28 days, and 29–365 days (table 2). Twins had an
Guide-1.3.pdf). We also excluded babies born in this region but overall higher mortality rate compared with singletons only in
resident outside it. Twinning was determined according to the the 23–25 weeks band (OR 2.04), with the statistically
order of delivery. We categorised the timing of death using the significant excess mortality confined to the early and late
standard epidemiological definitions of early, late and post- neonatal periods.
neonatal death. Table 3 shows the effect on mortality rates by mode of
For analytical purposes, the babies were grouped by gesta- delivery for twins and singletons. Only among the 23–25 weeks
tional age bands 23–25, 26–27, and 28–29 weeks. Data are babies was there any relationship with mode of delivery, babies
Table 2 Early, late and post-neonatal deaths by gestational age

Early neonatal death Late neonatal death Post-neonatal death Total infant deaths in 1 year
Twin Singleton Twin Singleton Twin Singleton Twin Singleton
Gestation OR OR OR OR
(weeks) n (MR) n (MR) (95% CI) n (MR) n (MR) (95% CI) n (MR) n (MR) (95% CI) n (MR) n (MR) 95% CI)
28–29 7 (33) 33 (49) 0.67 7 (34) 14 (22) 1.59 1 (5) 13 (21) 0.24 15 (72) 60 (90) 0.78
(0.29 to 1.50) (0.65 to 3.80) (0.04 to 1.45) (0.43 to 1.40)
26–27 17 (143) 64 (134) 1.06 2 (20) 20 (49) 0.39 2 (20) 19 (49) 0.40 21 (176) 103 (216) 0.77
(0.60 to 1.89) (0.10 to 1.53) (0.10 to 1.57) (0.46 to 1.29)
23–25 79 (520) 146 (367) 1.86 19 (260) 36 (143) 2.11 6 (111) 23 (106) 1.05 104 (684) 205 (515) 2.04
(1.28 to 2.72) (1.13 to 3.94) (0.40 to 2.60) (1.37 to 3.02)
n (MR), number (mortality rate expressed per 1000 livebirths in cases of total infant death and early neonatal death, and per 1000 survivors in cases of late neonatal death and post-
neonatal death).

Original article
Table 3 Mortality by mode of delivery

Vaginal delivery Caesarean section
Gestation in Twins Singletons Twins Singletons
weeks n (MR) n (MR) OR (95% CI) n (MR) n (MR) OR (95% CI)
28–29 7 (78) 25 (90) 0.86 (0.37 to 2.01) 8 (78) 18 (68) 1.16 (0.49 to 2.69)
26–27 12 (187) 41 (189) 0.99 (0.49 to 2.00) 9 (191) 35 (203) 0.93 (0.42 to 2.06)
23–25 78 (736) 96 (407) 4.06 (2.46 to 6.70) 14 (467) 38 (500) 0.87 (0.38 to 2.02)
n (MR), number (mortality rate expressed per 1000 livebirths).
OR calculated using total infant deaths in each subgroup of gestation.
born vaginally having a mortality odds ratio of 4.06 (CI 2.46 to dichorionic twins, but our observations, confined to liveborn
6.70) relative to caesarean section. twins, have failed to confirm this.
We did not find any effect of gender either for all gestations or The strength of our study is that it presents population-based
among babies of 23–25 weeks. There was no effect of data in a society where high-quality medical facilities are readily
chorionicity. Figure 2 illustrates the gestational age-specific accessible to all. The neonatal service of the whole region is
mortality for twins and singletons from 1998 to 2001 and 2002 provided by a single neonatal network, which assures a degree
to 2005. There is noticeable improvement in both the twins and of uniformity of management. The use of data from reliable
singleton mortality outcome between the two epochs, with regional databases, together with cross-checking at the level of
convergence of outcomes at a gestation of 27 weeks. individual units, means that we can be confident about the
quality of the data.
Our study has a number of limitations. We may have missed
DISCUSSION
a very small number of babies born outside the region to
We have demonstrated that any disadvantage to being a twin,
mothers resident in the region who received all their neonatal
in this population of very premature babies, is confined to the
care outside the region, which may affect our denominator
extreme of viability and to the epoch of 1998 to 2001, rather
numbers. Although when considering the whole group we
than in more recent years. Indeed, inspection of fig 2 suggests
compared 479 twins with 1538 singletons over an 8-year period,
that in the earlier epoch, the disadvantage to being a twin
the numbers are quite small when they are subdivided into
persisted up to 26 weeks of gestation, and the impression given
groups. Hence some of our results need cautious interpretation.
by table 2, that the higher mortality is confined to 23–25 weeks,
Also some data were missing such as birth weight (singletons
is an artefact of the pre-study choice of gestational age bands for
4%), mode of delivery (singleton 16% and twins 7%) and
the analysis. The figure also suggests that any disadvantage
chorionicity (17%). Furthermore our data do not allow us to
from being a twin has now disappeared.
correct for other confounders, but in any case this does not
While it is true that in general twins have more congenital
matter for those more mature babies in whom being a twin or
anomaly than singletons, our data showed an opposite trend.
singleton had no impact on mortality. It is only of importance
This might have been because we have focused on a highly
when trying to explain the difference between twins and
selected set of births, those occurring very preterm. We are not
aware of any other data in this highly selected group to suggest singletons in the least mature babies, where there are too few
whether the ratio of congenital anomaly in twins/singletons is subjects to allow for a meaningful multivariable analysis even if
similar to or different from that when all babies are considered. we had access to the relevant obstetric data.
Alternatively, perhaps the twins with congenital anomalies had We would be particularly cautious about reading too much
more antenatal or early pregnancy loss, or that the reasons why into the findings for mode of delivery. In the first place the
singletons and twins go into preterm labour may themselves be numbers are small, and in the second, the decision-making for
different, which may also help to account for the unexpectedly choosing a particular mode of delivery is complex, so the factors
different rates of malformation in the two groups. giving rise to the apparent association probably have more to do
Monochorionic twins are known to be at higher risk during with the reasons for choosing the mode of delivery than the
the pregnancy, and it has been suggested (from unit-based data) delivery itself.
that they suffer more postnatal deaths9 compared with Our data are compatible with both the view that twins are at
a disadvantage and the view that they are not. It all depends
which period of time is chosen for study. However, we do not
feel that we can confirm the notion that twins actually do
better than singletons.13 16 17 An analysis of birth registry data in
Sweden11 in 1992 showed similar findings to ours, although this
must be interpreted with caution because we defined our cohort
in terms of gestational age, not birth weight. On the other hand,
studies by Shinwell et al18 and Donovan et al19 using babies
,1500 g showed no significant difference in the mortality
outcome between singletons and twins. However, analysing our
own data by birth weight, with a cut-off at 700 g, gave a similar
result for the infant mortality rate among twins to that of
gestation (OR 2.39, 95% CI 1.47 to 3.86).
Finally, it is encouraging to note that the number of deaths in
extreme preterm gestation appears to have fallen in twins as well
Figure 2 Mortality comparison of singletons and twins in two epochs as singletons, and that the apparent disadvantage of very preterm
(1998–2001 and 2002–5). twins was greatly reduced in the more recent epoch (2002–5).

Original article
Acknowledgements: The authors are grateful for the assistance provided by M 8. Wolf EJ, Vintzileos AM, Rosenkrantz TS, et al. A comparison of pre-discharge
Bythell and M Renwick at the Regional Maternity Survey Office. survival and morbidity in singletons and twin very-low-birth-weight infants. Obstet
Gynecol 1992;80:436–9.
Funding: Funding for the Northern Congenital Abnomality Survey is provided by the 9. Asztalos EV, Barret JF, Lacy M, et al. Evaluating 2-year outcome in twins
Department of Health. (30 weeks gestation at birth: a regional perinatal unit’s experience. Twin Res
Competing interests: None. 2001;4:431–8.
10. Buekens P, Wilcox A. Why do small twins have a lower mortality rate than small
Ethics approval: The registers used in this study have been approved by ethics singletons? Am J Obstetr Gynecol 1993;168:937–41.
committees. 11. Ericson A, Gunnarskog J, Kallen B, et al. A registry study of very low birth weight live
Patient consent: We used the Multiple Pregnancy Register and data from the born infants in Sweden, 1973–1988. Acta Obstet Gynaecol Scand 1992;71:104–11.
12. Synnes AR, Ling EW, Whitfield MF, et al. Perinatal outcomes of a large cohort of
Perinatal Morbidity and Mortality Survey. The registers have clearance from the Patient
extremely low gestational age infants (twenty-three to twenty-eight completed
Information Advisory Group (PIAG) to hold named patient information without individual
weeks of gestation). J Pediatr 1994;125(6 Pt 1):952–60.
consent under section 60 of the Health and Social Care Act (2001). 13. Garite TJ, Clark RH, Elliott JP, et al. The Pediatrix/Obstetrix Perinatal Research
Group. Twins and triplets: the effect of plurality and growth on neonatal outcome
compared with singleton infants. Am J Obstet Gynecol 2004;191:700–7.
REFERENCES 14. Alexander GR, Kogan M, Martin J, et al. What are the fetal growth patterns of
1. Ward Platt MP, Glinianaia S, Rankin J, et al. The Multiple Pregnancy Register in the singletons, twins and triplets in the United States? Clin Obstet Gynecol 1998;41:115–25.
North of England: an update and five-year results. Twin Res Hum Genet 2006;9:913–18. 15. North East Public Health Observatory. Perinatal mortality and morbidity survey.
2. Office for National Statistics. Birth Statistics 2004 Series FM1 no.33. http://www. http://www.nepho.org.uk (accessed 7 Nov 2008).
statistics.gov.uk/statbase/Product.asp?vlnk = 5768 (accessed 7 Nov 2008). 16. Northern Regional Health Authority Coordinating Group. Perinatal mortality: a
3. Powers WF, Kiley JL. The risks confronting twins: a national perspective. Am J continuing collaborative regional survey. BMJ 1984;288:1717–20.
Obstet Gynecol 170:456–61. 17. Northern Regional Maternity Survey Office. Annual report, 2000. Newcastle
4. Sontag J, Waltz S, Schollmeyer T, et al. Morbidity and mortality of discordant twins upon Tyne, 2002.
up to 34 weeks of gestational age. Eur J Pediatr 1996;155:224–9. 18. Shinwell ES, Blickstein I, Lusky A, et al. Excess risk of mortality in very low
5. Victoria A, Mora G, Arias F. Perinatal outcome, placental pathology and severity of birthweight triplets: a national, population based study. Arch Dis Child Fetal Neonatal
discordance in monochorionic and dichorionic twins. Obstet Gyenecol 2001;97:310–15. Ed 2003;88:F36–40.
6. Sassoon DA, Castro LC, Davis JL, et al. Perinatal outcome in triplet versus twin 19. Donovan EF, Ehrenkrantz RA, Shankaran S, et al. Outcomes of very low birth weight
gestations. Obstet Gynecol 1990;75:817–20. twins cared for in the National Institute of Child Health and Human Development
7. Weissman A, Yoffe N, Jakobi P, et al. Management of triplet pregnancies in Neonatal Research Network’s intensive care units. Am J Obstet Gynecol
1980s—are we doing better? Am J Perinatal 1991;8:333–7. 1998;179:742–9.
Take advantage of BMJ Journals’ remarkable catalogue of titles with Related Collections
No busy professional has time to browse through all pertinent journals to find relevant articles, but with
Related Collections you no longer have to. Follow the ‘‘Related Collections’’ link from any article and use
the ‘‘Show Collections from other Journals’’ to expand your search across all BMJ Journals. Or simply
follow the ‘‘Browse by topic’’ link on the home page. By setting up your own collections and receiving
email alerts every time an article is added to your chosen area, you can build up your own significant
body of knowledge.

Neonatal infections in Asia

R Tiskumara, S H Fakharee, C Q Liu, P Nuntnarumit, K M Lui, M Hammoud, J K
F Lee, C B Chow, A Shenoi, R Halliday, D Isaacs and on behalf of APNIS
(Asia-Pacific Neonatal Infections Study)

online 19 Sep 2008;
doi:10.1136/adc.2008.139865

These include:

Notes


Original article
Neonatal infections in Asia

R Tiskumara,1 S H Fakharee,2 C Q Liu,3 P Nuntnarumit,4 K M Lui,5 M Hammoud,6
J K F Lee,7 C B Chow,8 A Shenoi,9 R Halliday,1 D Isaacs,1,10 on behalf of APNIS (Asia-
Pacific Neonatal Infections Study)
1
Children’s Hospital at ABSTRACT
Westmead, Westmead, NSW, Objective: To study the epidemiology (including inci- What is already known on this topic
Australia; 2 Mofid Children’s
Hospital, Teheran, Iran;
dence, antibiotic sensitivity and mortality) of neonatal unit
3
Children’s Hospital of Hebei infections in countries in Asia. c Hospital-acquired infections with multi-resistant
Province, China; 4 Ramathibodi Methods: One year prospective study of neonatal organisms are common in resource-poor Asian
Hospital, Bangkok, Thailand; infections in eight neonatal units in Asia. countries.
5
Centro Hospitalar Conde Sao
Januario, Macau SAR, China;
Results: There were 453 episodes of sepsis affecting c Early-onset infections in Asia are caused by
6
Ab Sarah Maternity Hospital, 394 babies. Mortality from neonatal sepsis was 10.4%, multiple different pathogens.
Kuwait; 7 Kuala Terengganu with an incidence of 0.69 deaths/1000 live births. Group B
Hospital, Terengganu, Malaysia; streptococcus was the most common early-onset
8
Princess Margaret Hospital
HKSAR, Hong Kong; 9 Manipal
organism causing 38% of episodes of early-onset (,48 h
old) sepsis, with a rate of 0.51 episodes per 1000 live What this study adds
Hospital, Bangalore, India;
10
University of Sydney, Sydney, births and a mortality of 22%. Gram-negative bacillary
NSW, Australia early-onset sepsis occurred at a rate of 0.15 episodes per c Although the rate of hospital-acquired infection
1000 live births with a mortality of 12%. There were 406 with multi-resistant Gram-negative organisms in
Correspondence to:
Dr D Isaacs, Department of episodes of late-onset sepsis. The incidence was high at Asia was high, the organisms isolated and
Infectious Diseases, Children’s 11.6 per 1000 live births, and mortality was 8.9%. mortality were similar to those in developed
Hospital at Westmead, Locked Coagulase-negative staphylococcus caused 34.1% of countries.
Bag 4001, Westmead, NSW c Group B streptococcus was the most common
2145, Australia; davidi@chw. episodes, whereas Staphlococcus aureus caused only
edu.au 5.4%. Gram-negative bacilli caused 189 episodes (46.6%). cause of early-onset sepsis.
Only 44% of Gram-negative bacilli were sensitive to both
Accepted 11 August 2008 gentamicin and a third-generation cephalosporin, whereas
Published Online First 30% were resistant to both antibiotics. Meningitis mainly resource-poor countries in Asia, and at
19 September 2008 antibiotic-sensitivity patterns.
occurred in 17.2% of episodes of late sepsis, with a
mortality of 20%.
Conclusions: The incidence of late-onset sepsis was MATERIALS AND METHODS
higher in Asia than in resource-rich countries, but the Neonatologists in Asia identified from a database
organisms isolated and mortality were similar. Over half of of neonatologists working in level 3 neonatal units
all Gram-negative bacilli were antibiotic resistant. (defined as those that can manage babies with
artificial ventilation) were invited to participate in
the study. Questionnaires were sent to those
Neonatal infections are an important cause of neonatologists who agreed.
mortality and morbidity world wide. In their Neonatal sepsis was defined as the pure growth
2000–2003 report, the World Health Organization of a single potentially pathogenic organism (bac-
estimated that neonatal sepsis and pneumonia are terium or fungus) from the blood of a baby who
responsible for about 1.6 million deaths each year, was clinically septic according to defined criteria6 7
mainly in resource-poor countries.1 Antibiotic and had supportive laboratory evidence of sepsis
resistance is an important problem in resource- (eg, one or more of low or high white cell count or
poor countries,2–5 and a survey of neonatologists in abnormal immature: total (I:T) ratio, low platelets,
Asian countries suggested that there is a significant or raised serum C-reactive protein, as defined
problem with sepsis caused by multi-resistant previously).6 Laboratory variables were age-depen-
Gram-negative organisms and meticillin-resistant dent. We did not further define clinical sepsis. We
Staphylococcus aureus (MRSA).6 Previous studies did not request two positive blood cultures because
have reported rates of hospital-acquired neonatal antibiotics are usually started empirically in Asia
infections that are 3–20 times higher in resource- after only one set of blood cultures had been taken.
poor than resource-rich countries.3 The most Likely contaminants were excluded. The decision
common reported organisms are Gram-negative as to whether a baby had true sepsis or if the
bacilli and S aureus.3 Antibiotic resistance rates are cultured organism was a contaminant was made
high: in South-East Asia up to 86% of Klebsiella by the local clinician, according to clinical judge-
species are resistant to cefotaxime and gentamicin, ment but using the above criteria. Early-onset
56% of Escherichia coli are resistant to gentamicin, sepsis was defined as the onset of sepsis within
and 28% of S aureus are resistant to meticillin.3 48 h of delivery, and late-onset sepsis as the onset
In this study, we look at the microorganisms of sepsis more than 48 h after delivery. Outcome
(bacteria and fungi), the incidence and the mortal- was defined as: died from sepsis, died possibly from
ity of early-onset and late-onset neonatal sepsis in sepsis, died from unrelated cause, or survived.

Original article
Reported cases were excluded if they did not meet the defined Forty-one babies had two episodes of infection, six had three
criteria for sepsis or if the data were inconsistent and could not episodes of infection, and one had five episodes of infection.
be verified. We relied on clinicians to report all cases of babies There were 41 deaths reported as a direct result of sepsis, giving
with positive blood cultures and clinical sepsis, and we did not a mortality of 10.4%. If mortality in only the maternity
search microbiology records for unreported episodes of neonatal hospitals was analysed, the overall mortality as a direct result of
septicaemia. sepsis can be expressed as 0.69 deaths/1000 live births. If we
All babies with sepsis meeting our criteria on neonatal units include deaths probably occurring as a result of sepsis in the
in the study were included, regardless of their postnatal age. analysis, the overall mortality increases to 1 death/1000 live
The babies were de-identified and recorded by their initials only. births.
The following data on babies with neonatal sepsis were The most common associated conditions or infections were
collected: gestational age, birth weight, gender, and details of meningitis (76 episodes), pneumonia (49 episodes), necrotising
whether the baby was born in the hospital or outside the enterocolitis (38 episodes) and skin sepsis (12 episodes). Thirty-
hospital and then transferred to the nursery. Denominator data four of the 38 bacteraemic episodes of necrotising enterocolitis
were collected on the number of babies managed at each site, (89%) were associated with episodes of late infection, and 44 of
categorised by birth weight. The age in days when the positive the 49 episodes of pneumonia (90%) occurred in association
blood culture was obtained was also recorded. with episodes of late infection. In 248 episodes of infection, no
The organisms isolated on blood culture and their antibiotic co-existing condition was recorded.
sensitivities were recorded. Local laboratories used recognised
methods of antibiotic susceptibility testing, but these were not
standardised. For Gram-negative bacilli, data were only sought Early-onset sepsis
on sensitivity to third-generation cephalosporins (cefotaxime or Early-onset infections, defined as infection with onset within
ceftazidime for Pseudomonas) and to gentamicin. The organisms 2 days of birth (47 episodes in total), caused 10.4% of all
were recorded as sensitive to both cephalosporin and gentami- infections reported. The rate of early-onset sepsis from any
cin, sensitive to cephalosporin and resistant to gentamicin, organism was 0.72 infections/1000 live births. Six out of 47
resistant to cephalosporin and sensitive to gentamicin, or babies with early-onset sepsis died, a mortality of 13%. Two
resistant to both cephalosporin and gentamicin. For S aureus, babies with early-onset sepsis went on to have late-onset sepsis
data were requested on sensitivity to meticillin. (one baby had one episode of late-onset sepsis, and one baby
For each episode of sepsis, it was recorded whether or not a had two episodes of late-onset sepsis).
lumbar puncture had been performed. The results of the Table 1 shows the organisms that caused early-onset sepsis.
cerebrospinal fluid (CSF) culture were recorded as well as the Group B streptococcus (GBS) was the most common organism,
presence of CSF leucocytosis (defined6 as a CSF white blood cell causing 18 out of 47 episodes of early-onset sepsis (38%). There
count .100/ml). Meningitis was defined as the presence of an were eight episodes of early GBS sepsis reported from Kuwait,
organism cultured from CSF or the presence of CSF leucocytosis five from Malaysia, four from Hong Kong, and one from China.
in the presence of a positive blood culture. Co-existing Seventeen episodes of early-onset GBS infection occurred in
conditions associated with infection, such as pneumonia or inborn babies in maternity hospitals, giving an incidence of 0.51
necrotising enterocolitis, were also recorded. per 1000 live births for early-onset GBS. Four of the 18 babies
The data were recorded on standard proformas by local with early-onset GBS died, a mortality of 22%. This was higher
clinicians, and these were either posted or e-mailed back. The than the mortality from Gram-negative infection in early-onset
data were then collated into a dedicated database created for the sepsis (2 of 16, 12.5%) and the total overall mortality in early
purpose. Mortality per 1000 live births was calculated using sepsis (12.8%), but the differences did not reach statistical
only data from neonatal units attached to maternity hospitals. significance.
Statistical calculations were performed using SPSS V15. There were 17 early-onset episodes of Gram-negative bacillary
This study was approved by the ethics committee of the infection. The most common Gram-negative bacillus in this
Royal Alexandra Hospital for Children. Local sites were asked if group was E coli (five episodes). The incidence of Gram-negative
they required local ethics approval and all felt that the Royal bacillary early-onset sepsis was 0.15 episodes per 1000 live
Alexandra Hospital for Children ethics approval was sufficient. births, and mortality was 12% (two deaths from 17 infections).
Sensitivities were reported for 16 episodes of early-onset
RESULTS Gram-negative bacillary sepsis. Seven organisms (44%) were
Seventeen level 3 neonatal units were approached. Data were resistant to either a third-generation cephalosporin or gentami-
received from eight units (47%), two from China and one each cin, six organisms (37%) were resistant to both, and three
from Hong Kong, India, Iran, Kuwait, Malaysia and Thailand. organisms (19%) were sensitive to both antibiotics.
We present the prospective data from 1 January through 31 There were 30 episodes of early-onset sepsis in which Gram-
December 2005. Five of the neonatal units were attached to positive organisms were isolated. Eighteen of these episodes
maternity units, and we were able to obtain the number of live
born infants in the maternity unit over the year. The other Table 1 Organisms causing early-onset sepsis
three neonatal units only accepted babies born elsewhere. The Organism n (%)
neonatal units had similar case mixes; all were able to ventilate
preterm babies and all cared for babies who underwent Group B streptococcus 18 (38)
Coagulase-negative staphylococcus 8 (17)
abdominal surgery.
Escherichia coli 6 (13)
There were 453 episodes of sepsis documented in the study,
Staphylococcus aureus 2 (4)
affecting 394 babies. The overall rate of a baby developing either
Other Gram-negative bacilli 11 (23)
early or late sepsis in our study, calculated from neonatal units
Other Gram-positive cocci 2 (4)
attached to maternity hospitals, varied from 3.0 per 1000 live Total 47 (100)
births in Hong Kong to 15.0 per 1000 live births in Kuwait.

Original article
Table 2 Organisms causing late-onset sepsis Two of the 12 S aureus infections (17%) for which sensitivities
were reported were meticillin resistant.
Organism n (%)
Regarding the 406 episodes of late-onset sepsis, 36 (8.9%)
Coagulase-negative staphylococcus 136 (33.5) babies died as a direct result of sepsis. If the 18 babies who
Klebsiella species 69 (17.0) probably died from sepsis are included in the analysis, mortality
Enterobacter species 29 (7.1) from late-onset sepsis increases to 13.3%. If the 136 episodes of
Escherichia coli 26 (6.4)
CONS infection are excluded from the total (because they are
Staphylococcus aureus 22 (5.4)
potential contaminants), mortality as a direct result of sepsis
Acinetobacter species 21 (5.2)
increases to 13.3% and as a direct result of definite plus probable
Pseudomonas species 18 (4.4)
sepsis to 20%. Gram-negative bacillary septicaemia had a higher
Serratia species 15 (3.7)
mortality as a direct result of infection (14.8%) than CONS
Other Gram-negative bacilli 10 (2.5)
Group B streptococcus 3 (0.7)
(1.5%) and S aureus (4.5%), but this did not reach statistical
Other Gram-positive cocci 26 (6.4) significance.
Candida species 25 (6.2) The incidence of late-onset infection was inversely propor-
Miscellaneous 6 (1.5) tional to birth weight (table 4). The proportion of babies
Total 406 (100.0) developing late-onset sepsis varied from 2.0 per 1000 live births
in Hong Kong to 22.0 per 1000 in Thailand. The overall figure
was 11.6 per 1000 live births (table 4).
(60%) were due to GBS. Eight episodes were reported to be due
to coagulase-negative staphylococcus (CONS), but we were
Meningitis
unable to exclude that they may have been contaminants. One
Meningitis was reported in 76 episodes of sepsis (16.8% of all
of the two S aureus isolated in early sepsis was meticillin
episodes) affecting 75 babies (table 5). Six of 47 babies with
resistant.
early-onset sepsis (13%) were reported with meningitis com-
Six episodes of early sepsis (12.8%) were associated with
pared with 70 of 406 episodes of late-onset sepsis (17.2%)
meningitis (see below and table 5).
(p.0.05). It was not possible to calculate the incidence of
meningitis in early sepsis per 1000 live births, as none of the
Late-onset sepsis episodes of meningitis occurred in a maternity hospital. One of
There were 406 episodes of late-onset sepsis, affecting 347 six babies with early-onset meningitis died.
babies. Gram-negative bacilli caused 189 (46.6%) of the episodes. Seventy episodes of late sepsis (17.2%) were associated with
Table 2 shows the organisms causing late-onset sepsis. The meningitis (table 5). Gram-negative bacilli caused 53 episodes of
most common Gram-negative bacillus was Klebsiella pneumoniae late-onset meningitis (75.7%). Meningitis occurred in 28.0% of
(69 episodes), followed by Enterobacter species (29) and E coli the 189 episodes of late-onset Gram-negative sepsis. The
(26). predominant species causing late-onset Gram-negative menin-
Sensitivities were recorded for 180 of the 189 Gram-negative gitis were Klebsiella (25, of which 23 were reported as Klebsiella
bacilli (95.2%): 44% were sensitive to both gentamicin and to a pneumoniae) and E coli (nine episodes). Meningitis was reported
third-generation cephalosporin, 27% were resistant to either in 18 (9.6%) of 187 episodes of late-onset sepsis due to Gram-
gentamicin (6%) or a third-generation cephalosporin (21%), and positive cocci: 10 were coagulase-negative staphylococci, two S
30% were resistant to both (table 3). Resistance was most aureus, and only one GBS (table 5). The reports of meningitis
common in Bangalore, India, where 21 (75%) of 28 Gram- due to CONS may represent contaminants, but we were unable
negative bacilli were resistant to both third-generation cepha- to verify this. One episode of meningitis was caused by an
losporins and gentamicin, and only three (11%) were sensitive anaerobe. We received no reports of fungal meningitis.
to both. Meningitis was significantly more common with late-onset
There were 187 episodes of late-onset Gram-positive sepsis. Gram-negative bacillary sepsis than with sepsis due to Gram-
CONS was the single most common isolate, causing 136 positive cocci (x2 = 11.3, p = 0.001).
episodes of Gram-positive sepsis (33.5% of all late sepsis The episodes of meningitis associated with late sepsis were
episodes). The next most common Gram-positive organism spread across all gestations. One baby had two episodes of late-
was S aureus, accounting for 22 late-onset infections (5.4%). onset meningitis. Fourteen of 70 babies with late-onset
meningitis died (20%).
Table 3 Sensitivities of Gram-negative organisms causing late-onset DISCUSSION

sepsis The rates of sepsis in our study varied from 2 per 1000 live births
Organism CSGS CSGR CRGS CRGR (%) Total in Hong Kong to 22 per 1000 live births in Thailand, with an
Acinetobacter species 6 4 7 3 (15) 20
overall figure of 11.6 per 1000 live births, a number dominated
Escherichia coli 14 1 5 5 (20) 25 by the data from Kuwait. Our study dealt with neonatal units
Enterobacter species 11 2 12 3 (11) 28 and is not representative of neonatal sepsis in the community.
Klebsiella species 35 2 2 31 (44) 70 The rates of sepsis we calculated were from units with attached
Proteus species 0 0 2 1 (33) 3 maternity wards, and the figures we obtained were comparable
Pseudomonas species 3 1 6 6 (37) 16 to the reported overall incidence of neonatal sepsis in Asia, of
Serratia species 9 0 1 3 (23) 13 7.1–38.0 per 1000 live births,2 8–12 and higher than the rates
Other Gram-negative bacilli 1 0 2 2 (40) 5 usually reported from resource-rich countries in North
Total 79 (44%) 10 (6%) 37 (21%) 54 (30) 180 America,7 Europe7 and Australia6 of 1.0–8.1 per 1000 live births.
C, third-generation cephalosporin (cefotaxime or ceftazidime for Pseudomonas); G, The mortality figures in this study, of 13% for early sepsis and
gentamicin; S, sensitive; R, resistant. 8.9% for late sepsis, are similar to recent mortality figures from

Original article
Table 4 Proportion of babies admitted to neonatal intensive care who developed late-onset sepsis, by birth
weight
Maternity hospital ,1000 g 1000–1499 g 1500–1999 g 2000–2499 g >2500 g Total
Centro Hospitalar Conde S Januario, China 4/6 0/14 1/32 0/23 0/333 5/408
Kuala Terengganu, Malaysia 4/35 5/87 7/294 5/564 12/3609 33/4589
Ramathibodi Hospital, Thailand 3/18 5/35 0/88 0/133 3/225 11/499
Ab Sarah Maternity Hospital, Kuwait 61/131 45/157 27/285 8/698 22/10211 163/11482
Princess Margaret Hospital, Hong Kong 3/22 1/19 1/100 3/1857 8/1998
Total 75/211 56/312 49/1234 40/16235 220/18976
% of total 35.5 17.9 4.0 0.25 1.2
the USA,7 and are very similar to the Australian mortality The proportion of babies with early-onset sepsis who had
figures for 1991–2 of 15% for early-onset sepsis and 9% for late- meningitis in our study (12.8%) was comparable to figures from
onset sepsis.6 The mortality from late-onset Gram-negative North America and Europe.12 The rate of meningitis in late
bacillary sepsis in the present study was also comparable to sepsis (17.2%), however, was slightly higher than the rate of
recent North American13 and Australian14 data. ,10% generally reported from North America, Europe and
The most common pathogen reported causing early-onset Australia.6 7 This was despite our use of stringent criteria (.100
sepsis in our study was GBS, responsible for 38.3% of early white cells/ml) to define meningitis, although lowering these
sepsis. CONS was next most common, either due to rapid early criteria to .20 cells did not significantly alter our figures, as
postnatal acquisition of the organism or as blood culture most babies with meningitis grew organisms from CSF.
contaminants. E coli and other Gram-negative bacilli caused Mortality from late-onset meningitis was 20% in the present
most of the other early-onset infections. The pattern of study, a rate that is perhaps surprisingly low and comparable to
organisms identified in early sepsis in Asia in this study is mortality in resource-rich countries.7 This may be due to
similar to that described in resource-rich countries. selection bias and/or to a high quality of care in the nurseries
The incidence of early-onset sepsis due to GBS in this study that participated in the study.
was 0.51 per 1000 live births, which is lower than the incidence Of the Gram-negative bacillary infections for which sensitiv-
in the USA15 and Australia6 before the widespread use of ities were reported, only 30% of Gram-negative organisms
intrapartum antibiotics. GBS is usually reported to be an responsible for late-onset sepsis and 19% of Gram-negative
uncommon cause of early-onset sepsis in resource-poor coun- organisms responsible for early-onset sepsis were reported as
tries.2 8 In previous Asian studies, the incidence of early-onset being sensitive to both a third-generation cephalosporin and
sepsis due to GBS was 0.1–0.27 cases per 1000 live births in gentamicin. The rest were resistant to either antibiotic or both.
Thailand,9 0.11–1.39 in Taiwan,10 0.27 in Singapore11 and 0.4 in This is in line with data from Asia suggesting very high rates of
Malaysia.12 In contrast, the incidence of early-onset GBS disease antibiotic resistance among Gram-negative bacilli.3 8 16 In con-
in the USA before the use of intrapartum antibiotics was 0.7–3.7 trast, only 17% of S aureus species responsible for late infection
per 1000 live births.15 that were tested were meticillin resistant, whereas many Asian
units report higher rates of MRSA.16
The organisms causing late-onset infection in Asian neonatal
The data we obtained were surprisingly similar to those from
units in the present study were similar to those reported from
resource-rich countries, in terms of the organisms isolated from
resource-rich countries since the 1980s, with CONS being the
babies with early and late sepsis, the incidence of meningitis,
most common cause of infection, followed by Gram-negative
and mortality from meningitis and overall sepsis. The major
bacilli.1 2 6
differences are that the rate of late-onset sepsis in Asia is higher
than in resource-rich countries and that a very high percentage
of Gram-negative organisms are resistant to either gentamicin
Table 5 Organisms causing meningitis or a third-generation cephalosporin or both. However, this does
Late-onset Early-onset not apparently result in increased mortality, at least in the units
Organism meningitis meningitis studied here, although, because the incidence of sepsis is higher
Acinetobacter species 3 – than in resource-rich countries, more Asian babies are dying
Bacillus species 1 – from sepsis. Furthermore, if rates of antibiotic resistance
Coagulase-negative staphylococcus 10 1 continue to rise, we expect that mortality will also rise in
Escherichia coli 9 1 Asia as clinicians run out of antibiotic options.
Edwardsiella tarda 1 – The present study is not a systematic survey of Asian
Enterobacter species 3 – neonatal units. Selection bias is one of the major criticisms of
Haemophilus influenzae 1 – the study, and it is likely that the neonatal units reporting data
Klebsiella species 25 2 are better resourced than many other neonatal units in the same
Listeria species 1 – country and many others in the region. However, the data are
Proteus species 1 – recent, prospective and from various centres and add valuable
Pseudomonas species 1 1 information on neonatal infections in Asia, particularly with
Salmonella species 1 – regard to antibiotic resistance. One of the strengths of this
Serratia marcescens 6 –
study is that it is an ongoing project, and hence there is the
Staphylococcus aureus 2 1
capacity to monitor changes in pathogens and their changing
Group B streptococcus 1 –
antibiotic-resistance patterns with time. We intend to continue
Streptococcus species 1 –
collecting data for several years and to recruit more neonatal
Total 70 6
units.

Original article
Acknowledgements: This study was performed by the members of the Asia-Pacific 6. Isaacs D, Barfield C, Grimwood K, et al. Systemic bacterial and fungal infections in
Neonatal Infections Study: R Tiskumara, R Halliday, D Isaacs (Children’s Hospital at infants in Australian neonatal units. Med J Aust 1995;162:198–200.
Westmead, Sydney, Australia), SH Fakharee (Mofid Children’s Hospital, Teheran, Iran), 7. Palazzi DL, Klein JO, Baker CJ. Bacterial sepsis and meningitis. In: Remington JS,
C-Q Lui (Children’s Hospital of Hebei Province, China), P Nuntnarumit (Ramathibodi Klein JO, Wilson CB, et al, eds. Infectious diseases of the fetus and newborn infant.
Hospital, Bangkok, Thailand), K-M Lui (Centro Hospitalar Conde Sao Januario, Macau 6th edition. Philadelphia: Elsevier-Saunders, 2006:247–96.
SAR, China), M Hammoud, SK Seema, A Mazen (Ab Sarah Maternity Hospital, 8. Orsin D, Vergnano S, Anthony C. Serious bacterial infections in newborn infants in
Kuwait), JFK Lee, SCO Zuraidan (Kuala Terengganu Hospital, Terengganu, Malaysia), developing countries. Curr Opin Infect Dis 2004;17:217–24.
CB Chow, CC Shek (Princess Margaret Hospital HKSAR, Hong Kong), A Shenoi, NN 9. Yossuck P, Kanchana P. Neonatal group B streptococcal infection: incidence and
Nagesh (Manipal Hospital, Bangalore, India). The study was unfunded, there are no clinical manifestation in Siriraj Hospital. J Med Assoc Thai 2002;85(Suppl
known conflicts of interest, and the data are provided through the generosity of the 2):S479–87.
10. Ho MY, Wu CT, Ku YT, et al. Group B Streptococcal infection in neonates: an 11-year
participants.
review. Acta Paediatr Taiwan 1999;40:83–6.
Competing interests: None. 11. Tan K, Tay L, Lin R, et al. Group B Streptococcal septicaemia/meningitis in neonates
Ethics approval: Obtained. in a Singapore teaching hospital. Aust N Z J Obstet Gynaecol 1998;38:418–23.
12. Lim CT, Thong MK, Parasakthi N, et al. Group B streptococcus: maternal
carriage rate and early neonatal septicaemia. Ann Acad Med Singapore
REFERENCES 1997;26:421–5.
1. Bryce J, Boschi-Pinto C, Shibuya K, et al. WHO estimates of the causes of death in 13. Karlowicz MG, Buescher ES, Surka AE. Fulminant late onset bacterial infection in a
children. Lancet 2005;365:1147–53. neonatal intensive care unit, 1988–1997, and the impact of avoiding empiric
2. Vergnano S, Sharland M, Kazembe P, et al. Neonatal sepsis: an international vancomycin therapy. Pediatrics 2000;107:1387–90.
perspective. Arch Dis Child Fetal Neonatal Ed 2005;90:F220–4. 14. Gordon A, Isaacs D. Late onset Gram negative bacillary infection in Australia and
3. Zaidi A, Huskins C, Thaver D, et al. Hospital-acquired neonatal infections in New Zealand 1992–2002. Pediatr Infect Dis J 2006;25:25–9.
developing countries. Lancet 2005;365:1175–89. 15. Edwards MS, Nizet V, Baker CJ. Group B streptococcal infections. In: Remington
4. Kapoor L, Randhawa V, Deb M. Microbiological profile of neonatal septicemia in a JS, Klein JO, Wilson CB, et al, eds. Infectious diseases of the fetus and newborn
pediatric care hospital in Delhi. J Commun Dis 2005;37:227–32. infant. 6th edn. Philadelphia: Elsevier-Saunders, 2006:403–64.
5. Waheed M, Laeeq A, Maqbool S. The etiology of neonatal sepsis and patterns of 16. Isaacs D. Unnatural selection: reducing antibiotic resistance in neonatal units. Arch
antibiotic resistance. J Coll Physicians Surg Pak 2003;13:449–52. Dis Child Fetal Neonatal Ed 2006;91:F72–4.

Neonatal blood pressure waves are associated

with surges of systemic noradrenaline
B Wefers, S Cunningham, R Stephen and N McIntosh

online 1 May 2008;
doi:10.1136/adc.2007.123273

These include:

Notes


Short report
Neonatal blood pressure waves are associated with

surges of systemic noradrenaline
B Wefers,1 S Cunningham,2 R Stephen,3 N McIntosh3
1
Department of Neonatology, ABSTRACT adrenaline and ,10 pg per injection for dopamine,
Section of Child Life and Health, Neonatal blood pressure (BP) waves have been linked to and the inter-assay coefficients of variation for
University of Edinburgh, noradrenaline, adrenaline and dopamine were
Edinburgh, UK; 2 Department of
neonatal illness. We investigated plasma levels of
Respiratory and Sleep Medicine, vasoactive hormones when BP waves were observed. 1.6%, 2.9% and 3.4%, respectively.
Royal Hospital for Sick Children, Peak and trough noradrenaline levels correlated with Endothelin levels were measured by radio
Edinburgh, UK; 3 Department of mean BP (p = 0.028). There was no relationship to immunoassay in a subgroup of samples obtained
Reproductive and adrenaline, dopamine or endothelin levels. from a single patient. The mean BP value recorded
Developmental Sciences,
University of Edinburgh,
by the CPTM at the time of sampling was
Edinburgh, UK correlated to the vasoactive hormone levels using
Maintenance of blood pressure (BP) is important the Wilcoxon signed rank test.
Correspondence to: for perfusion of vital organs and avoidance of
Birgit Wefers, NICU, Ninewells
Hospital and Medical School,
perfusion-related brain injury in the newborn. RESULTS
Dundee DD1 9SY, UK; Normal reference ranges relate to gestation and Seven paired sets of consecutive peak/trough blood
b.wefers@nhs.net postnatal age.1 Hypertensive neonatal BP waves samples were obtained from five infants with
have previously been described2 and have been typical hypertensive BP waves. In two infants we
Accepted 2 April 2008 linked to more severe neonatal illness. An increased
Published Online First 2 April 2008 were able to obtain two sets of peak/trough levels.
risk of cerebral pathology in these infants has been In one infant we were able to obtain a total of
postulated. However the mechanism behind these seven consecutive samples.
waves is unclear. As BP control normally involves a The results of the paired consecutive samples and
number of vasoactive chemicals, we aimed to associated mean BP values are shown in table 1. The
measure these during the peaks and troughs of five infants had a median birth weight of 2620 g
the waves. (interquartile range; IQR 956–3001) and median
gestation of 33 weeks (IQR 27–38). The median
METHODS wavelength of BP waves was 14.3 minutes (IQR
The infants were receiving intensive care because 7.3–16.2), with the median amplitude 13.2 mmHg
of prematurity or sickness and thus had contin- (IQR 8.3–19.4). The median noradrenaline peak level
uous invasive measurement of BP. Measurements was 10.5 nmol/l (IQR 5.1–38.1) and was signifi-
are sampled each second and stored on our cantly greater than the trough level 5.7 nmol/l (IQR
computerised trend monitoring system (CPTM) 3.1–25.8) (Wilcoxon signed rank test, p = 0.021). In
with mean BP routinely displayed as a trend graph all cases the level at the wave peak was greater than
at the cot side. This revealed cyclical BP waves that at the corresponding trough. There was no signifi-
were of a degree which led to clinical concern, cant difference in the peak and trough levels of other
especially with regards to possible effects on hormones; adrenaline was generally undetectable.
cerebral perfusion. It was therefore thought clini- Infant 2 had a total of seven blood samples taken
cally important to try and explore the underlying over 3 h. These samples showed a significant
mechanisms. An example of the pattern is shown correlation between noradrenaline and mean BP
in fig 1. Paired blood samples (0.5–1 ml from the (r2 = 0.64, p = 0.0280). No significant correlation
indwelling arterial line used for monitoring BP) was found between dopamine, adrenaline or
were taken at the peak and trough of a BP wave endothelin levels and mean BP in this infant
with parental permission in order to determine the (r2 = 0.17, r2 = 0.03 and r2 = 0.04, respectively).
cause of the waves. Samples were collected into
cooled ethylenediaminetetraacetic acid tubes con- DISCUSSION
taining sodium metabisulphite, separated immedi- BP is usually maintained within a tight physiolo-
ately and stored at 270uC prior to analysis. gical range by a combination of physiological
Noradrenaline, adrenaline and dopamine were control mechanisms. These include the autonomic
measured by reverse-phase, ion-pair high perfor- nervous system, the cardiovascular system and the
mance liquid chromatography (HPLC) with dual- endocrine system. The stability of BP achieved in
electrode coulometric detection (which gave an this manner is important for the perfusion of
additional order of sensitivity compared to the more essential organs. We have previously described
usual amperometric electrochemical detection). A hypertensive BP waves in both term and preterm
simple solvent extraction technique was used to newborn infants, suggesting that under certain
isolate the catecholamines from plasma prior to their circumstances this balanced regulation is dis-
separation by HPLC. Dihydroxybenzylamine was turbed.2 We have also previously described that
used as the internal standard and the average such hypertensive BP waves were associated with
recovery was 88.3% (n = 100). The limits of detec- more severe neonatal illness.2 Many of the neural
tion were ,5 pg per injection for noradrenaline and and endocrine controls of the cardiovascular

Short report
Figure 1 31-week gestation infant with

heart rate and mean blood pressure (BP)
trend graphs over 6 h showing waves
(the infant was not receiving inotropes).
system have intrinsic rhythms which may be exaggerated by Although a rise in noradrenaline levels was seen in two infants
hypoxia.3 Repeated measurements of neonatal endocrine func- when a dopamine infusion was started (infants 1 + 5), the
tion are scarce due to technical and ethical difficulties; however relationship between BP and noradrenaline was maintained
arginine vasopressin has previously been demonstrated to independently. Both infants with two sets of peak and trough
induce BP waves.4 samples showed a reduction in noradrenaline levels over time.
The advent of micro assays and the widespread use of The samples were taken at 8- and 4-minute intervals,
invasive BP monitoring in Neonatal Intensive Care allowed us respectively. This is in excess of the plasma noradrenaline
to investigate the relationship between some vasoactive half-life of 2–3 minutes described by Esler et al5 which would
hormones and BP waves in a small group of infants. These suggest that the hypertensive BP waves were related to repeated
assays demonstrated a strong relationship between mean BP noradrenaline surges. In our opinion it is less likely that BP
during episodes of waves and serum noradrenaline levels. waves induce surges in noradrenaline. Noradrenaline release can
Table 1 Infant characteristics and vasoactive hormone levels

Birth
Gestation weight Wavelength Amplitude Age mBP
Infant (weeks) (g) (minutes) (mmHg) (h) Peak/trough (mm Hg) NA (nmol/l) DA (nmol/l) A (nmol/l) Outcome
1 36 2620 16.2 31.4 31 p 69 118 280 ,0.1 Meconium aspiration,

died 6 months
t 42 71.2 146 ,0.1
1 16.2 31.4 31 p 70 43.12 83 ,0.1
t 38 21.52 51 ,0.1
2 29 886 8.9 8.3 93 p 49 10.5 ,0.16 ,0.1 Problems of prematurity,
normal at F/U
t 41 3.87 0.48 ,0.1
2 8.9 8.3 93 p 49 2.71 ,0.16 ,0.1
t 40 1.82 0.48 ,0.1
3 33 2960 14.3 8.3 60 p 42 3.97 257 ,0.1 Hydrops unknown
aetiology, normal at F/U
t 36 2.42 269 ,0.1
4 28 560 16.2 13.2 192 p 43 6.3 1.4 ,0.1 Problems of prematurity
and growth retardation,
neurodevelopmental
delay
t 34 5.7 0.7 ,0.1
5 40 2933 2.9 15.4 40 p 62 33.5 19 1 Birth asphyxia, outcome
unknown
t 48 30 21.1 0.8
A, adrenaline; DA, dopamine; F/U, follow-up; mBP, mean blood pressure; NA, noradrenaline; p, peak; t, trough.

Short report
be caused by raised intracranial pressure. This relationship was other cardiovascular regulatory mechanisms would be neces-
first explored by Harris et al who mimicked the effect of the sary.
contractions of labour on the fetal head in the near-term fetal
sheep. They showed that phasic increase in intracranial pressure Acknowledgements: We would like to thank Dr Alberto Smith, Reader in Vascular
was followed within 30 seconds by a rise in mean BP, but that Science at Kings College London.
after 10 cycles or so the rise in BP became tonic.6 This is different Competing interests: None declared.
to what we describe. In our infants we have no measures of
intracranial pressure, and would not a priori believe it to be REFERENCES
cycling. Lagercrantz’s group showed that there is a postnatal 1. Cunningham S, Symon AG, Elton RA, et al. Intra-arterial blood pressure reference
surge of noradrenaline, which can be exaggerated by perinatal ranges, death and morbidity in very low birthweight infants during the first seven days
hypoxia (certainly present in some of our infants) but this surge of life. Early Hum Dev 1999;56(2–3):151–65.
2. Cunningham S, Deere S, McIntosh N. Cyclical variation of blood pressure and heart
lasts for hours or days, and we are not aware that cycles have rate in neonates. Arch Dis Child 1993;69:64–7.
been described within it.7 Noradrenaline is released as a stress 3. Kocsis B, Fedina L, Pasztor E. Two phase change of sympathetic rhythms in brain
hormone, and in a separate small number of infants we have ischaemia, Cushing reaction and asphyxia. Am J Physiol 1989;256:R120–32.
found that BP waves can be induced by a stressful event such as 4. Murata Y, Miyake Y, Yakamoto T, et al. Experimentally produced sinusoidal fetal heart
rate pattern in the chronically instrumented fetal lamb. Am J Obstet Gynecol
bradycardia, a blocked endotracheal tube or transient hypoxia. 1985;153:693–702.
However infants demonstrating stress-induced BP waves had 5. Esler M, Jennings G, Korner P, et al. Measurement of total and organ specific
waves that were somewhat smaller in amplitude, median norepinephrine kinetics in humans. Am J Physiol 1984;247:E21–8.
5 mmHg (IQR 5–10), and wavelength, median 1 minute (IQR 6. Harris AP, Koehler RC, Nishijima MK, et al. Circulatory dynamics during periodic
intracranial hypertension in fetal sheep. Am J Physiol 1992;263:R95–102.
0.6–7.8). Before firm conclusions could be drawn regarding the 7. Soulier V, Dalmaz Y, Cottet-Emard JM, et al. Long-term influence of neonatal hypoxia
role of noradrenaline in the development of hypertensive BP on catecholamine activity in carotid bodies and brainstem cell groups of the rat.
waves further investigation of vasoactive hormone levels and J Physiol 1997;498:523–30.

Early individualised parenteral nutrition for

preterm infants
S Eleni-dit-Trolli, E Kermorvant-Duchemin, C Huon, M Mokthari, K Husseini, M-L
Brunet, C Dupont and A Lapillonne

online 6 Oct 2008;
doi:10.1136/adc.2007.136333

These include:

Notes


Short report
Early individualised parenteral nutrition for preterm

infants
S Eleni-dit-Trolli,1 E Kermorvant-Duchemin,1,2 C Huon,1 M Mokthari,1 K Husseini,1
M-L Brunet,3 C Dupont,1,2 A Lapillonne1,2
1
APHP, Department of ABSTRACT unit from the first day of life. Infants who had
Neonatology and Nutrition, Considerable effort should be made to optimise parenteral congenital malformations or metabolic disease,
Saint-Vincent de Paul Hospital, nutrition of preterm infants in order to limit the were transferred to another hospital, or died before
Paris, France; 2 Paris Descartes
University, Paris, France; 3 APHP, development of postnatal growth restriction. A mono- the 28th day of life were excluded from the study.
Parenteral Nutrition Unit, Cochin centric before-and-after study design was used to The last 20 preterm infants who matched the
Hospital, Paris, France determine the effects of computerising parenteral inclusion criteria and who were consecutively
nutrition ordering on the composition of parenteral admitted before the date of implementation of
Correspondence to:
Professor A Lapillonne, nutrition (PN) solutions and early clinical outcomes of the CPN ordering system were included in the
Department of Neonatology and preterm infants born (28 weeks of gestation. Parenteral control group, and the first 20 preterm infants
Nutrition, Saint-Vincent de Paul protein intake during the first week of life and parenteral admitted after this date were included in the CPN
Hospital, 74 Avenue Denfert lipid, glucose and energy intakes during the first and group. The nutritional protocol remained
Rochereau, 75014 Paris, France;
alexandre.lapillonne@svp.aphp.fr second week of life were significantly higher in infants unchanged during the two phases of the study.
assessed after the introduction of computerised parent- Perinatal and nutritional data were collected
Accepted 4 September 2008 eral nutrition ordering. This led to a significant reduction in from medical records. Human milk was assumed to
Published Online First the cumulative energy deficit over the first 28 days of life contain 70 kcal/100 ml and 1.05 g protein/100 ml.
6 October 2008 and to an improvement in both early growth and Formula intakes were based on published manu-
pulmonary outcome. Computerising the PN ordering facturer’s figures. Weight was converted into
process improves the nutrient content of the PN solutions standard deviation scores (z scores). Cumulative
and early postnatal outcome. intakes over the first 28 days of life were obtained
by adding the daily enteral and parenteral intakes
of the first 28 days of life. Actual daily intakes were
Postnatal growth deficit is the most commonly subtracted from recommended intake (120 kcal/
observed morbidity in very low birthweight kg/day for energy and 3.5 g/kg/day for protein) to
(VLBW) infants and is due, at least in part, to calculate daily deficits. These were subsequently
inadequate early nutritional intake during hospita- added together to calculate cumulative deficits.
lisation.1 In order to appropriately manage nutri- The statistical analysis was carried out using
tional intake of VLBW infants, daily changes in the Minitab13.3 (Minitab Inc, State College,
composition of parenteral nutrition (PN) are Pennsylvania, USA). Comparative analyses were
usually required, but the ordering of individualised performed using a x2 test, or an unpaired t test
PN solutions is associated with a high incidence of when appropriate. The level of significance was
medical errors and protocol deviations. Our fixed at p(0.05.
hypothesis was that optimisation of early PN by
using a computerised PN (CPN) ordering process
would improve early nutritional intake, reduce RESULTS
early nutritional deficit, and, in turn, influence Mean (SD) birth weight and gestational age of the
early growth and neonatal outcomes. 40 infants were 973 (208) g and 27.2 (0.9) weeks,
respectively. Maternal and infant clinical charac-
teristics, PN duration and parenteral fluid intake
MATERIALS AND METHODS were similar in the two groups (data not shown).
On 6 June 2005, we implemented an automated
computerised calculation for determining, for each
infant, the optimal daily PN solution according to
Improvement in early nutritional intakes
Parenteral protein intake during the first week of
our nutritional protocol. Before this date, all
life and parenteral lipid, glucose and energy intakes
individualised PN orders were calculated at the
during the first and second week of life were
bedside, whereas, after this date, the desired PN
significantly higher in the CPN group than in the
intake was calculated using standardised spread-
control group (table 1).
sheet software (Microsoft Excel), which takes into
account the desired total volume and nutrient
intake, the nutrition provided by the enteral Reduction in early nutritional deficit
nutrition, and the volume of drugs prescribed. Owing to a higher parenteral intake over the first
A before-and-after monocentric study was car- 28 days, the cumulative total intake was signifi-
ried out to determine the effects of the CPN cantly higher in the CPN group for lipid (+16%),
ordering process on nutritional intake and clinical carbohydrate (+8%) and energy (+11.5%), but the
outcome of VLBW infants during the first 28 days difference did not reach significance for protein
of life. The inclusion criteria were birth at a (+4%) (table 2). The 28-day cumulative energy
gestational age (28 weeks and admission to the deficit was significantly smaller in the CPN than in

Short report
Table 1 Weekly parenteral intakes received by preterm infants before Table 2 Cumulative parenteral, enteral and total intake during the first
(ie, control) and after (ie, CPN) the introduction of a computerised 4 weeks of life received by 20 preterm infants before (ie, control) and
parenteral nutrition (CPN) ordering process after (ie, CPN) the introduction of a computerised parenteral nutrition
Control CPN (CPN) ordering process
Week (n = 20) (n = 20) p Value Control CPN p
(n = 20) (n = 20) Value
Protein (g/kg/week) 1 19.6 (3.7) 23.2 (3.0) 0.002
2 20.1 (3.0) 21.9 (4.8) 0.16 Protein
3 15.0 (8.5) 15.1 (9.4) 0.96 Cumulative parenteral intake (g/kg) 62.9 (18.1) 66.2 (21.5) 0.60
4 8.3 (8.5) 6.1 (8.6) 0.42 Mean intake per day (g/kg) 2.2 (0.7) 2.4 (0.8)
Glucose (g/kg/week) 1 65.4 (6.0) 79.1 (11.1) ,0.001 Cumulative enteral intake (g/kg) 36.0 (18.8) 36.9 (16.9) 0.87
3 47.2 (29.5) 52.9 (34.1) 0.57 Cumulative total intake (g/kg) 98.9 (7.1) 103.1 (8.3) 0.09
Lipid (g/kg/week) 1 5.2 (2.8) 15.9 (4.5) ,0.001 Lipid
2 19.8 (9.5) 26.7 (3.6) 0.006 Cumulative parenteral intake (g/kg) 44.9 (22.1) 65.2 (22.0) 0.004
4 7.0 (8.2) 5.7 (9.0) 0.65 Cumulative enteral intake (g/kg) 75.4 (36.3) 73.8 (31.8) 0.88
Energy (kcal/kg/week) 1 387 (50) 551 (78) ,0.001 Mean intake per day (g/kg) 2.7 (1.3) 2.6 (1.1)
2 565 (105) 692 (133) 0.002 Cumulative total intake (g/kg) 120.4 (30.9) 139.0 (18.2) 0.02
3 368 (225) 425 (252) 0.45 Mean intake per day (g/kg) 4.3 (1.1) 5.0 (0.7)
4 214 (211) 159 (232) 0.44 Glucose
Values are mean (SD). Cumulative parenteral intake (g/kg) 218.8 (69.1) 243.9 (79.3) 0.29
Mean intake per day (g/kg) 7.8 (2.5) 8.7 (2.8)
the control group (252 (165) vs 2389 (344) kcal/kg, respec- Cumulative enteral intake (g/kg) 154.2 (77.1) 160.5 (64.8) 0.78
tively; p = 0.001), whereas there was no protein deficit at Mean intake per day (g/kg) 5.5 (2.8) 5.7 (2.3)
28 days of life for either group (5 (8) vs 1 (7) g/kg; p = 0.096). Cumulative total intake (g/kg) 373.0 (25.9) 404.5 (26.5) 0.001
Mean intake per day (g/kg) 13.3 (0.9) 14.4 (1.0)
Energy
Improvement in early growth and neonatal outcomes Cumulative parenteral intake (kcal/kg) 1534 (480) 1827 (571) 0.087
The importance of early weight loss was similar in both groups, Mean intake per day (kcal/kg) 55 (17) 65 (20)
but the time to regain birth weight was significantly shorter in Cumulative enteral intake (kcal/kg) 1437 (702) 1481 (605) 0.84
the CPN group (10.4 (2.5) vs 12.8 (3.3) days, p = 0.01). There Mean intake per day (kcal/kg) 51 (25) 53 (22)
was a trend for a higher weight gain (+6.5%; 11 (4.1) vs 13.2 Cumulative total intake (kcal/kg) 2971 (344) 3308 (165) 0.001
(3.8) days, p = 0.079) and a smaller loss of weight z score Mean intake per day (kcal/kg) 106 (12) 118 (6)
between birth and the 28th day of life in the CPN group (20.76 Results are mean (SD) expressed as cumulative intake over the 4-week period and as
(0.24) vs 2 0.95 (0.39); p = 0.073). Other growth parameters did mean intake per day during this period.
not differ. The number of infants with necrotising enterocolitis,
patent ductus arteriosus, nosocomial infection, intraventricular
overall rate of infants with BPD remained unchanged. We cannot
haemorrhage, white matter disease, or a requirement for insulin
rule out the possibility that factors other than nutrition changed
therapy was similar in the two groups. The number of infants
between the two study periods. However, our observations are in
with bronchopulmonary dysplasia (BPD) at 28 days of life was
line with previous experimental and epidemiological studies,
similar in the two groups (n = 14 vs 15), but the number of
which have shown that early reduction in energy intake is a risk
those with severe BPD (ie, need for ventilator support) was
factor for the development of BDP and that early inadequate
significantly lower in the CPN group (n = 0 vs 5; p = 0.049).
nutrition, especially protein deficiency, affects lung cell division
and diminishes alveolar formation and lung volume.5
DISCUSSION Our data show the nutritional advantages of computerising
This study shows that providing doctors with an adequate tool the PN ordering process of individualised parenteral nutrition
for prescribing individualised PN is effective in improving the and also suggest advantages in terms of better clinical outcome
nutritional management of VLBW infants, which may lead, in for VLBW infants.
turn, to better clinical outcome. Together with previously
shown advantages of time saving, efficiency and accuracy of Acknowledgements: We thank F Lemare, A Lescoat, J Rambaud, L Fellous, M
calculations, these data emphasise the need to computerise the Moreno and A Guiseppi for their contribution to the preparation of PN solutions, data
process of ordering individualised PN.2 acquisition and/or infant management.
The difference between the two groups was more apparent for Competing interests: None.
energy intake than protein intake. This is not surprising, as the
nutritional protocol used in the unit is very close to the most REFERENCES
recent recommendations for protein and leads to no apparent 1. Thureen P, Heird WC. Protein and energy requirements of the preterm/low
protein deficit over the first month of life. However, the birthweight (LBW) infant. Pediatr Res 2005;57:95R–98R.
introduction of CPN ordering further improved early protein 2. Puangco MA, Schanler RJ. Computerized PN ordering optimizes timely nutrition
therapy in a neonatal intensive care unit. J Am Diet Assoc 1997;97:258–61
intake. We postulate that the early loss of body protein was 3. Embleton NE, Pang N, Cooke RJ. Postnatal malnutrition and growth retardation: an
minimised in our study, and this, in turn, may explain why the inevitable consequence of current recommendations in preterm infants? Pediatrics
time to regain birth weight was shorter and the loss of weight z 2001;107:270–3.
4. Ehrenkranz R, Younes N, Lemons JA, et al. Longitudinal growth of hospitalized very
score is smaller than those in previously published studies.3 4
low birthweight infants. Pediatrics 1999;104:280–9.
We also showed that the change in mode of PN ordering was 5. Biniwale MA, Ehrenkranz RA. The role of nutrition in the prevention and management
associated with a reduction in the severity of BPD, although the of bronchopulmonary dysplasia. Semin Perinatol 2006;30:200–8.

Impact of delayed screening for prolonged

jaundice in the newborn
M Tyrell, S Hingley, C Giles and J O Menakaya

doi:10.1136/adc.2008.145268

These include:

Notes


PostScript
LETTERS 3. Hannam S, McDonnell M, Rennie JM. Investigation of

prolonged neonatal jaundice. Acta Paediatr
2000;89:694–7.
Impact of delayed screening for 4. Mowat AP, Davidson LL, Dick MC. Earlier
prolonged jaundice in the identification of biliary atresia and hepatobiliary
disease: selective screening in the third week of life.
newborn Arch Dis Child 1995;72:90–2.
Jaundice persisting for longer than 14 days in

the newborn is a trigger to screen for serious Hydrocortisone treatment for
underlying disorders such as biliary atresia severe evolving
and other hepatobiliary disorders.1 However,
screening too early could result in consider- bronchopulmonary dysplasia and
able anxiety for the parents and unnecessary cerebral haemodynamics
tests for the baby.2 We evaluated our
practice of delaying screening tests for 1 We read with interest the paper by
week after referral to assess its effect on Rademaker et al, in which the authors raised
babies with prolonged jaundice. the question of replacing dexamethasone
with hydrocortisone to treat severe evolving
bronchopulmonary dysplasia (BPD).1 In
METHODS Figure 1 Jaundice status at review of 183 2006, a national survey revealed that 91%
The outcome for all babies referred aged 14 babies referred with prolonged jaundice. DNA, of French neonatologists used betametha-
days to a dedicated prolonged jaundice clinic did not attend (and therefore jaundice status not sone for delayed treatment of BPD. From
at Hillingdon Hospital, Middlesex, UK known). October 2004 to June 2006, we treated 12
between 1 April 2006 and 31 July 2007 was infants with betamethasone for severe evol-
reviewed. The babies had a detailed clinical evaluated clinically even if no further tests ving BPD and observed a pronounced
assessment carried out the age of at least were carried out. This ensured that no co- decrease in cerebral blood flow velocities
21 days by a paediatrician. All jaundiced morbidities were missed. The low pick-up (CBFVs).2 Moreover, a quarter of these
babies had additional screening investiga- rate for hepatobiliary abnormalities was patients also presented an enlargement of
tions (full blood count and film, liver similar to the findings observed by the subarachnoid space and the lateral
function test, total and conjugated bilirubin, Hannam et al.3 In addition, this study ventricles, suggesting a transitory impair-
thyroid function test, blood group and direct enabled us to determine the effect of a 1- ment in brain growth. These morphological
Coombs test, urine for reducing substances week delay for assessment on referrals to the and functional cerebral effects prompted us
and culture), and glucose-6-phosphate dehy- prolonged jaundice clinic. During this inter- to replace betamethasone with hydrocorti-
drogenase assay was carried out in at risk val, jaundice had resolved in many babies, sone in .4-week-old infants who still
babies. Babies with abnormal results were reducing anxiety for parents, unnecessary required the combination of high-frequency
referred immediately to a named consultant tests to babies and the cost to the NHS. oscillatory ventilation, inhaled nitric oxide
paediatrician. Biliary atresia is an important but rare and FiO2.50%. This change in corticosteroid
disorder, which must be treated within was the only modification in their ventila-
RESULTS 8 weeks of life.1 This approach will not tory care.
A total of 183 (62% male) babies with compromise early detection of babies with The hydrocortisone was administered in
prolonged jaundice lasting 2 weeks were hepatobiliary disorders, indeed it is consis- four intravenous injections daily for 6 days,
referred. The mean (SD) gestational age tent with recommendations for selective at successive doses of 5, 5, 3, 3, 1 and 1 mg/
was 38.1 (2.0) weeks, and the mean (SD) screening at 3 weeks.4 kg/day (adapted from Lodygensky et al3).
birth weight was 3170 (630) g. More than CBFVs were measured by Doppler ultraso-
half (59%) of the babies were exclusively nography, using the computer-controlled
breast fed, and the mode age at review was CONCLUSION pulsed High Q Auto Doppler Trace (Philips
3 weeks (range 2–9). Jaundice had resolved Delayed assessment of babies with pro- Medical Systems, Eindhoven, The
in 79 babies (43%) between day 14 (referral longed jaundice reduced significantly the Netherlands) in the anterior cerebral artery
day) and day 21 (review day). Ninety-four number of babies that required screening and the lenticulostriate artery. The anterior
persistently jaundiced babies were investi- investigations without missing any impor- horn width and the pericerebral space were
gated, and pathology was identified in 16: tant pathologies. measured at the level of the interventricular
direct Coombs test positive (5), cardiac foramen just before and after completion of
murmur (5), raised thyroid-stimulating hor- M Tyrell, S Hingley, C Giles, J O Menakaya the treatment and then every week until
mone .6.5 mU/l (2), glucose-6-phosphate Department of Neonatal Paediatrics, Hillingdon Hospital, term, on ultrasound coronal scans. Signed
dehydrogenase deficiency (1), urinary tract Uxbridge, UK permission was obtained from the parents,
infection (1), failure to thrive (1) and bloody Correspondence to: Dr J O Menakaya, Department of and the study was approved by the ethics
stools (1). Overall, 80% of the babies were Neonatal Paediatrics, Hillingdon Hospital, Pield Heath Road, committee of our institution.
discharged after one visit. However, 18% of Uxbridge, Middlesex UB8 3NN, UK; jide.menakaya@thh. From July 2006 to September 2007, 12
babies seen at 3 weeks and above were
nhs.uk infants (median (range)) birth weight: 725
recalled to clinic compared with 40% of Competing interests: None. (620–865) g, median (range) gestational age:
those seen at 2 weeks or less. Biliary atresia 25.9 (24.9–26.7) weeks) were enrolled at a
was not identified in any baby during the median (range) post-natal age of 31.1 (29.3–
Arch Dis Child Fetal Neonatal Ed 2009;94:F154. 33.2) days. No significant variation in
study. doi:10.1136/adc.2008.145268 CBFVs or pulsatility index was observed in
either artery during hydrocortisone treat-
DISCUSSION ment (table 1). Before hydrocortisone, the
REFERENCES
The 1-week deferment for assessment of 1. Kelly DA Davenport M. Current management of
mean (SD) values for anterior horn width
referred babies resulted in resolution of biliary atresia. Arch Dis Child 2007;92:1132–5. and the pericerebral space were, respectively,
jaundice in nearly half of the babies (fig 1). 2. Hall DMD, Michel J-M. Screening in infancy. Arch Dis 1.30 (0.56) and 2.8 (0.9) mm and did not
All babies who attended the clinic were Child 1995;72:93–6. show significant change thereafter.

Hydrocortisone treatment for severe evolving

bronchopulmonary dysplasia and cerebral
haemodynamics
G Cambonie, R Mesnage, C Milési, A Rideau, C Veyrac and J-C Picaud

http://fn.bmj.com/cgi/content/full/94/2/F154-a
These include:
http://fn.bmj.com/cgi/content/full/94/2/F154-a#BIBL

http://fn.bmj.com/cgi/eletter-submit/94/2/F154-a
Notes


PostScript
LETTERS 3. Hannam S, McDonnell M, Rennie JM. Investigation of

prolonged neonatal jaundice. Acta Paediatr
2000;89:694–7.
Impact of delayed screening for 4. Mowat AP, Davidson LL, Dick MC. Earlier
prolonged jaundice in the identification of biliary atresia and hepatobiliary
disease: selective screening in the third week of life.
newborn Arch Dis Child 1995;72:90–2.
Jaundice persisting for longer than 14 days in

the newborn is a trigger to screen for serious Hydrocortisone treatment for
underlying disorders such as biliary atresia severe evolving
and other hepatobiliary disorders.1 However,
screening too early could result in consider- bronchopulmonary dysplasia and
able anxiety for the parents and unnecessary cerebral haemodynamics
tests for the baby.2 We evaluated our
practice of delaying screening tests for 1 We read with interest the paper by
week after referral to assess its effect on Rademaker et al, in which the authors raised
babies with prolonged jaundice. the question of replacing dexamethasone
with hydrocortisone to treat severe evolving
bronchopulmonary dysplasia (BPD).1 In
METHODS Figure 1 Jaundice status at review of 183 2006, a national survey revealed that 91%
The outcome for all babies referred aged 14 babies referred with prolonged jaundice. DNA, of French neonatologists used betametha-
days to a dedicated prolonged jaundice clinic did not attend (and therefore jaundice status not sone for delayed treatment of BPD. From
at Hillingdon Hospital, Middlesex, UK known). October 2004 to June 2006, we treated 12
between 1 April 2006 and 31 July 2007 was infants with betamethasone for severe evol-
reviewed. The babies had a detailed clinical evaluated clinically even if no further tests ving BPD and observed a pronounced
assessment carried out the age of at least were carried out. This ensured that no co- decrease in cerebral blood flow velocities
21 days by a paediatrician. All jaundiced morbidities were missed. The low pick-up (CBFVs).2 Moreover, a quarter of these
babies had additional screening investiga- rate for hepatobiliary abnormalities was patients also presented an enlargement of
tions (full blood count and film, liver similar to the findings observed by the subarachnoid space and the lateral
function test, total and conjugated bilirubin, Hannam et al.3 In addition, this study ventricles, suggesting a transitory impair-
thyroid function test, blood group and direct enabled us to determine the effect of a 1- ment in brain growth. These morphological
Coombs test, urine for reducing substances week delay for assessment on referrals to the and functional cerebral effects prompted us
and culture), and glucose-6-phosphate dehy- prolonged jaundice clinic. During this inter- to replace betamethasone with hydrocorti-
drogenase assay was carried out in at risk val, jaundice had resolved in many babies, sone in .4-week-old infants who still
babies. Babies with abnormal results were reducing anxiety for parents, unnecessary required the combination of high-frequency
referred immediately to a named consultant tests to babies and the cost to the NHS. oscillatory ventilation, inhaled nitric oxide
paediatrician. Biliary atresia is an important but rare and FiO2.50%. This change in corticosteroid
disorder, which must be treated within was the only modification in their ventila-
RESULTS 8 weeks of life.1 This approach will not tory care.
A total of 183 (62% male) babies with compromise early detection of babies with The hydrocortisone was administered in
prolonged jaundice lasting 2 weeks were hepatobiliary disorders, indeed it is consis- four intravenous injections daily for 6 days,
referred. The mean (SD) gestational age tent with recommendations for selective at successive doses of 5, 5, 3, 3, 1 and 1 mg/
was 38.1 (2.0) weeks, and the mean (SD) screening at 3 weeks.4 kg/day (adapted from Lodygensky et al3).
birth weight was 3170 (630) g. More than CBFVs were measured by Doppler ultraso-
half (59%) of the babies were exclusively nography, using the computer-controlled
breast fed, and the mode age at review was CONCLUSION pulsed High Q Auto Doppler Trace (Philips
3 weeks (range 2–9). Jaundice had resolved Delayed assessment of babies with pro- Medical Systems, Eindhoven, The
in 79 babies (43%) between day 14 (referral longed jaundice reduced significantly the Netherlands) in the anterior cerebral artery
day) and day 21 (review day). Ninety-four number of babies that required screening and the lenticulostriate artery. The anterior
persistently jaundiced babies were investi- investigations without missing any impor- horn width and the pericerebral space were
gated, and pathology was identified in 16: tant pathologies. measured at the level of the interventricular
direct Coombs test positive (5), cardiac foramen just before and after completion of
murmur (5), raised thyroid-stimulating hor- M Tyrell, S Hingley, C Giles, J O Menakaya the treatment and then every week until
mone .6.5 mU/l (2), glucose-6-phosphate Department of Neonatal Paediatrics, Hillingdon Hospital, term, on ultrasound coronal scans. Signed
dehydrogenase deficiency (1), urinary tract Uxbridge, UK permission was obtained from the parents,
infection (1), failure to thrive (1) and bloody Correspondence to: Dr J O Menakaya, Department of and the study was approved by the ethics
stools (1). Overall, 80% of the babies were Neonatal Paediatrics, Hillingdon Hospital, Pield Heath Road, committee of our institution.
discharged after one visit. However, 18% of Uxbridge, Middlesex UB8 3NN, UK; jide.menakaya@thh. From July 2006 to September 2007, 12
babies seen at 3 weeks and above were
nhs.uk infants (median (range)) birth weight: 725
recalled to clinic compared with 40% of Competing interests: None. (620–865) g, median (range) gestational age:
those seen at 2 weeks or less. Biliary atresia 25.9 (24.9–26.7) weeks) were enrolled at a
was not identified in any baby during the median (range) post-natal age of 31.1 (29.3–
Arch Dis Child Fetal Neonatal Ed 2009;94:F154. 33.2) days. No significant variation in
study. doi:10.1136/adc.2008.145268 CBFVs or pulsatility index was observed in
either artery during hydrocortisone treat-
DISCUSSION ment (table 1). Before hydrocortisone, the
REFERENCES
The 1-week deferment for assessment of 1. Kelly DA Davenport M. Current management of
mean (SD) values for anterior horn width
referred babies resulted in resolution of biliary atresia. Arch Dis Child 2007;92:1132–5. and the pericerebral space were, respectively,
jaundice in nearly half of the babies (fig 1). 2. Hall DMD, Michel J-M. Screening in infancy. Arch Dis 1.30 (0.56) and 2.8 (0.9) mm and did not
All babies who attended the clinic were Child 1995;72:93–6. show significant change thereafter.

PostScript
Table 1 Mean flow velocities and pulsatility index in the anterior cerebral artery and the automatic amino acid analyser (L-8800;
lenticulostriate artery during hydrocortisone treatment Hitachi, Tokyo, Japan).1 Both UV and UA
samples from each neonate were analysed in
Baseline Day 1 Day 3 Day 5 ANOVA
the same HPLC run, with a variance of 1.4%.
ACA Figure 1 shows the difference between the
MFV 18.4 (7.8) 19.0 (8.5) 17.9 (7.3) 18.6 (8.4) 0.17 UV and UA amino acid concentrations (the
PI 1.52 (0.37) 1.50 (0.32) 1.53 (0.35) 1.51 (0.43) 0.41 umbilical flux) of the neonates. Cetin et al2
LSA compared UV or UA plasma amino acid
MFV 8.2 (3.6) 8.4 (2.5) 8.3 (3.0) 8.1 (3.3) 0.30 concentrations of normal and gestational
PI 0.94 (0.25) 0.92 (0.18) 0.94 (0.27) 0.93 (0.22) 0.29 diabetes mellitus (GDM) pregnancies. In
plasma samples of GDM pregnancies, valine,
Data are mean (SD). Each data point corresponds to an average of 792 (10) measurements in the ACA and 736 (9)
measurements in the LSA. ACA, anterior cerebral artery; ANOVA, repeated-measure analysis of variance; LSA, lenticulostriate methionine, phenylalanine, isoleucine, leu-
artery; MFV, mean flow velocity; PI, pulsatility index. cine, ornithine, glutamate, proline, and
alanine were increased, whereas glutamine
Thus, even at pharmacological doses, which amino acids are required by the fetus was markedly decreased. They also showed
hydrocortisone did not produce the consis- and which ones are sufficient. We analysed that umbilical venoarterial plasma amino
tent decrease in CBFVs observed with amino acid concentrations in the UV and acid differences were not remarkably differ-
betamethasone in comparable conditions. UA blood obtained at birth by using high- ent from zero. However, many of the amino
Decreased blood flow could be one of the performance liquid chromatography acid flux patterns seen in fig 1 were
mechanisms involved in the deleterious (HPLC). We examined 31 singleton, full- significantly different (p,0.05) and seemed
effects of synthetic glucocorticoids on brain term neonates appropriate for gestational reasonable. All essential amino acids—
development. From a haemodynamic point age (according to Japanese standards). The lysine, valine, leucine, threonine, trypto-
of view, these results suggest the use of Fukuda Hospital Board approved the proto- phan, histidine, isoleucine, phenylalanine
hydrocortisone to rescue infants with severe col of this investigation. Informed consent and methionine—showed a positive differ-
BPD. was obtained from all the mothers who ence (positive flux). Alanine, which is the
participated. The mothers underwent sched- most gluconeogenic in fasting, showed the
G Cambonie,1 R Mesnage,1 C Milési,1 A Rideau,1 uled caesarean sections (for breech position, highest flux. The excessive flow of lysine
C Veyrac,2 J-C Picaud1 into the fetus may be explained by the
history of caesarean section, or maternal
1
Neonatal Intensive Care Unit, Arnaud de Villeneuve preference) and refrained from drinking or hypothesis that there is a greater proportion
Hospital, University Hospital of Montpellier, Montpellier, of lysine transporters (cationic amino acid
eating for at least 6 h before the operation.
France; 2 Department of Pediatric Radiology, Arnaud de transporter) on the fetal side of placenta
Villeneuve Hospital, University Hospital of Montpellier, During the caesarean, the umbilical cord was
Montpellier, France double-clamped by the obstetrician. UV and than on the maternal side.3 A negative
UA blood samples were obtained from this difference (negative flux) was seen with
Correspondence to: Dr Gilles Cambonie, Unité de
Réanimation Néonatale, Pédiatrie II, Hôpital Arnaud segment of cord and were stored separately glutamic acid, taurine, glycine, etc.
de Villeneuve, 371 Avenue du Doyen G. Giraud, in test tubes containing ethylenediaminete- Negative flux of glutamic acid and serine
34295 Montpellier Cedex 5, France; traacetic acid. Blood samples were centri- may reflect the fact that these amino acids
g-cambonie@chu-montpellier.fr fuged (4uC, 20 min) immediately to remove are produced in the fetal liver and not
Competing interests: None. precipitated protein. The plasma was sepa- supplied by the placenta.4
Ethics approval: The study was approved by the ethics rated after centrifugation and mixed with 2 Our results confirm the differences in
committee of our institution. volumes of 5% (w/w) trichloroacetic acid. amino acid concentrations between UV
Patient consent: Parental consent obtained. The samples were kept in a freezer at 280uC and UA blood. Pathological conditions in
until they were analysed by HPLC, using an fetuses that may affect the differences in the
REFERENCES
1. Rademaker KJ, de Vries LS, Uiterwaal CS, et al.
Postnatal hydrocortisone treatment for chronic lung
disease in the preterm newborn and long-term
neurodevelopmental follow-up. Arch Dis Child Fetal
2. Cambonie G, Mesnage R, Milési C, et al.
Betamethasone impairs cerebral blood flow velocities
in very premature infants with severe chronic lung
disease. J Pediatr 2008;152:270–5.
3. Lodygensky GA, Rademaker K, Zimine S, et al.
Structural and functional brain development after
hydrocortisone treatment for neonatal chronic lung
disease. Pediatrics 2005;116:1–7.
Differences between the amino

acid concentrations of umbilical
venous and arterial blood
Amino acids are supplied to the fetus via the Figure 1 The umbilical flux of amino acids in fetuses. The umbilical flux denotes the difference
umbilical vein (UV). The difference between between the concentration of each amino acid in the umbilical venous blood and the umbilical
the amino acid concentrations of the UVs arterial blood, in other words, the net uptake of each amino acid into the fetus. The median
and umbilical arteries (UAs), in other words venoarterial difference with 95% confidence interval on either side is shown for each amino acid
the umbilical amino acid flux, determines flux. Hyp, hydroxyproline; aAbu, a-aminobutyric acid; MEA, monoethanolamine.

Differences between the amino acid

concentrations of umbilical venous and arterial
blood
H Tsuchiya, K Matsui, T Muramatsu, T Ando and F Endo

doi:10.1136/adc.2008.147256

These include:

Notes


PostScript
Table 1 Mean flow velocities and pulsatility index in the anterior cerebral artery and the automatic amino acid analyser (L-8800;
lenticulostriate artery during hydrocortisone treatment Hitachi, Tokyo, Japan).1 Both UV and UA
samples from each neonate were analysed in
Baseline Day 1 Day 3 Day 5 ANOVA
the same HPLC run, with a variance of 1.4%.
ACA Figure 1 shows the difference between the
MFV 18.4 (7.8) 19.0 (8.5) 17.9 (7.3) 18.6 (8.4) 0.17 UV and UA amino acid concentrations (the
PI 1.52 (0.37) 1.50 (0.32) 1.53 (0.35) 1.51 (0.43) 0.41 umbilical flux) of the neonates. Cetin et al2
LSA compared UV or UA plasma amino acid
MFV 8.2 (3.6) 8.4 (2.5) 8.3 (3.0) 8.1 (3.3) 0.30 concentrations of normal and gestational
PI 0.94 (0.25) 0.92 (0.18) 0.94 (0.27) 0.93 (0.22) 0.29 diabetes mellitus (GDM) pregnancies. In
plasma samples of GDM pregnancies, valine,
Data are mean (SD). Each data point corresponds to an average of 792 (10) measurements in the ACA and 736 (9)
measurements in the LSA. ACA, anterior cerebral artery; ANOVA, repeated-measure analysis of variance; LSA, lenticulostriate methionine, phenylalanine, isoleucine, leu-
artery; MFV, mean flow velocity; PI, pulsatility index. cine, ornithine, glutamate, proline, and
alanine were increased, whereas glutamine
Thus, even at pharmacological doses, which amino acids are required by the fetus was markedly decreased. They also showed
hydrocortisone did not produce the consis- and which ones are sufficient. We analysed that umbilical venoarterial plasma amino
tent decrease in CBFVs observed with amino acid concentrations in the UV and acid differences were not remarkably differ-
betamethasone in comparable conditions. UA blood obtained at birth by using high- ent from zero. However, many of the amino
Decreased blood flow could be one of the performance liquid chromatography acid flux patterns seen in fig 1 were
mechanisms involved in the deleterious (HPLC). We examined 31 singleton, full- significantly different (p,0.05) and seemed
effects of synthetic glucocorticoids on brain term neonates appropriate for gestational reasonable. All essential amino acids—
development. From a haemodynamic point age (according to Japanese standards). The lysine, valine, leucine, threonine, trypto-
of view, these results suggest the use of Fukuda Hospital Board approved the proto- phan, histidine, isoleucine, phenylalanine
hydrocortisone to rescue infants with severe col of this investigation. Informed consent and methionine—showed a positive differ-
BPD. was obtained from all the mothers who ence (positive flux). Alanine, which is the
participated. The mothers underwent sched- most gluconeogenic in fasting, showed the
G Cambonie,1 R Mesnage,1 C Milési,1 A Rideau,1 uled caesarean sections (for breech position, highest flux. The excessive flow of lysine
C Veyrac,2 J-C Picaud1 into the fetus may be explained by the
history of caesarean section, or maternal
1
Neonatal Intensive Care Unit, Arnaud de Villeneuve preference) and refrained from drinking or hypothesis that there is a greater proportion
Hospital, University Hospital of Montpellier, Montpellier, of lysine transporters (cationic amino acid
eating for at least 6 h before the operation.
France; 2 Department of Pediatric Radiology, Arnaud de transporter) on the fetal side of placenta
Villeneuve Hospital, University Hospital of Montpellier, During the caesarean, the umbilical cord was
Montpellier, France double-clamped by the obstetrician. UV and than on the maternal side.3 A negative
UA blood samples were obtained from this difference (negative flux) was seen with
Correspondence to: Dr Gilles Cambonie, Unité de
Réanimation Néonatale, Pédiatrie II, Hôpital Arnaud segment of cord and were stored separately glutamic acid, taurine, glycine, etc.
de Villeneuve, 371 Avenue du Doyen G. Giraud, in test tubes containing ethylenediaminete- Negative flux of glutamic acid and serine
34295 Montpellier Cedex 5, France; traacetic acid. Blood samples were centri- may reflect the fact that these amino acids
g-cambonie@chu-montpellier.fr fuged (4uC, 20 min) immediately to remove are produced in the fetal liver and not
Competing interests: None. precipitated protein. The plasma was sepa- supplied by the placenta.4
Ethics approval: The study was approved by the ethics rated after centrifugation and mixed with 2 Our results confirm the differences in
committee of our institution. volumes of 5% (w/w) trichloroacetic acid. amino acid concentrations between UV
Patient consent: Parental consent obtained. The samples were kept in a freezer at 280uC and UA blood. Pathological conditions in
until they were analysed by HPLC, using an fetuses that may affect the differences in the
REFERENCES
1. Rademaker KJ, de Vries LS, Uiterwaal CS, et al.
Postnatal hydrocortisone treatment for chronic lung
disease in the preterm newborn and long-term
neurodevelopmental follow-up. Arch Dis Child Fetal
2. Cambonie G, Mesnage R, Milési C, et al.
Betamethasone impairs cerebral blood flow velocities
in very premature infants with severe chronic lung
disease. J Pediatr 2008;152:270–5.
3. Lodygensky GA, Rademaker K, Zimine S, et al.
Structural and functional brain development after
hydrocortisone treatment for neonatal chronic lung
disease. Pediatrics 2005;116:1–7.
Differences between the amino

acid concentrations of umbilical
venous and arterial blood
Amino acids are supplied to the fetus via the Figure 1 The umbilical flux of amino acids in fetuses. The umbilical flux denotes the difference
umbilical vein (UV). The difference between between the concentration of each amino acid in the umbilical venous blood and the umbilical
the amino acid concentrations of the UVs arterial blood, in other words, the net uptake of each amino acid into the fetus. The median
and umbilical arteries (UAs), in other words venoarterial difference with 95% confidence interval on either side is shown for each amino acid
the umbilical amino acid flux, determines flux. Hyp, hydroxyproline; aAbu, a-aminobutyric acid; MEA, monoethanolamine.

PostScript
amino acid concentrations is an interesting Table 1 Demographic and clinical

issue for further study on umbilical flux characteristics of the 20 infants at time of
analysis. We are developing a system that measurement*
allows analysing amino acids with much
smaller amounts of samples. Neonates 20
Male to female ratio 13/7
H Tsuchiya,1 K Matsui,2 T Muramatsu,3 T Ando,3 Gestational age (weeks) 35 (3.8)
F Endo4
Age (days) 16 (18)
1
Department of Pediatrics, Fukuda Hospital, Kumamoto, Actual weight (g) 2418 (616)
Japan; 2 Department of Obstetrics, Fukuda Hospital,
Mean arterial pressure (mm Hg) 46.1 (8.5)
Kumamoto, Japan; 3 Institute of Life Sciences, Ajinomoto
Co. Inc., Kawasaki-shi, Japan; 4 Department of Pediatrics, Heart rate/min 133 (13)
Kumamoto University Graduate School of Medicine, Arterial oxygen saturation (%) 96.0 (2.5)
Kumamoto, Japan Haemoglobin concentration (g/l) 129 (20)
Correspondence to: Dr H Tsuchiya, Department of Figure 1 Comparison of cerebral and periph- Temperature rectal (uC) 36.8 (0.5)
Pediatrics, Fukuda Hospital, Shin-machi 2-2-6, Kumamoto, eral ‘‘tissue oxygenation index’’ (c-TOI and p- Temperature peripheral (uC) 34.5 (0.9)
860-0004 Japan; tsuchiya@fukuda-hp.or.jp TOI) in 20 healthy newborn infants. Calf circumference (cm) 9.4 (1.1)
Competing interests: None. Thigh circumference (cm) 12.5 (1.9)
Accepted 4 September 2008 the mid-sternum. To increase the precision1 Calf diameter (cm) 3.1 (0.3)
Arch Dis Child Fetal Neonatal Ed 2009;94:F155–F156. the optodes were reapplied five times. After Calf subcutaneous adipose tissue (cm) 0.3 (0.1)
doi:10.1136/adc.2008.147256 each application there was a rest period of at *Values are mean (SD) except number of neonates and the
least 3 min and repeated measurements male to female ratio.
REFERENCES lasting 20 s each were performed five times.
1. Noguchi Y, Zhang QW, Sugimoto T, et al. Network Heart rate and arterial oxygen saturation
analysis of plasma and tissue amino acids and the were measured by pulse oximetry. Central
generation of an amino index for potential diagnostic comparison of c-TOI and p-TOI in compro-
and peripheral temperatures and mean blood
use. Am J Clin Nutr 2006;83:S513–19. mised infants.
pressure were measured before and after
2. Cetin I, de Santis MS, Taricco E, et al. Maternal and
fetal amino acid concentrations in normal pregnancies NIRS measurements. Diameter of calf and K Grossauer, G Pichler, G Schmölzer, H Zotter,
and in pregnancies with gestational diabetes mellitus. subcutaneous adipose tissue were measured W Mueller, B Urlesberger
Am J Obstet Gynecol 2005;192:610–17. with ultrasound. c-TOI and p-TOI were Division of Neonatology, Department of Pediatrics, Medical
3. Battaglia FC, Regnault TR Placental transport and determined as mean values of the repeated University of Graz, Austria
metabolism of amino acids. Placenta 2001;22:145–61. measurements in each newborn and com-
4. Regnault TR, Friedman JE, Wilkening RB, et al. Correspondence to: Dr G Pichler, Division of Neonatology,
Fetoplacental transport and utilization of amino acids in
pared using paired t test. Data are presented Department of Pediatrics, University of Graz,
IUGR—a review. Placenta 2005;26(Suppl A):S52–62. as mean (SD). Auenbruggerplatz 30, A-8036 Graz, Austria; pichler.ger-
Demographic and clinical characteristics hard@klinikum-graz.at
of the infants at time of measurements are Acknowledgements: The authors would like to thank
Comparison of peripheral and presented in table 1. At time of measure- E Ziehenberger for her assistance.
ment all infants had a weight .2000 g. Of
cerebral tissue oxygenation index the 500 measurements (25 in each neonate)
Accepted 4 September 2008
in neonates 135 measurements were excluded because of
body movements causing artefacts. c-TOI Arch Dis Child Fetal Neonatal Ed 2009;94:F156.
Near-infrared spectroscopy (NIRS) is a non- doi:10.1136/adc.2008.146654
invasive method to measure haemoglobin was significantly higher than p-TOI (70.4%
and tissue oxygenation continuously. (6.7) vs 62.1% (5.7), respectively; p,0.001)
Measurement of ‘‘cerebral tissue oxygena- (fig 1). The c-TOI/p-TOI ratio was 1.14¡ REFERENCES
tion index’’ (c-TOI)1 and ‘‘peripheral tissue 0.14. 1. Sorensen LC, Greisen G. Precision of measurement of
oxygenation index’’ (p-TOI)2 is based on This is the first report of comparison of cerebral tissue oxygenation index using near-infrared
simultaneously measured c-TOI and p-TOI spectroscopy in preterm neonates. J Biomed Opt
spatially resolved spectroscopy (SRS). SRS is
in healthy term and preterm infants, and 2006;11:054005.
realised with one light detector having 2. Pichler G, Grossauer K, Klaritsch P, et al. Peripheral
sensors at different distances. The aim of therefore, the present study is the first to
oxygenation in term neonates. Arch Dis Child Fetal
the present study was to measure c-TOI and introduce an index (c-TOI/p-TOI ratio). Neonatal Ed 2007;92:F51–52.
p-TOI simultaneously to compare the two We found that our values for c-TOI and p- 3. Boushel R, Langberg H, Olesen J, et al. Monitoring
values. TOI were similar to recent studies.1 2 There tissue oxygen availability with near infrared
NIRS measurements were carried out in may be several reasons for differences spectroscopy (NIRS) in health and disease.
between c-TOI and p-TOI. Differences in Scand J Med Sci Sports 2001;11:213–222.
20 term and preterm infants (gestational age 4. Watzman H, Kurth C, Montenegro L, et al. Arterial and
.29 weeks and birth weight .1200 g) the ratio of the three vascular compartments venous contributions to near-infrared cerebral oximetry.
within the first 8 weeks after birth. At time (arterial:capillary:venous) in muscle and Anesthesiology 2000;93:947–53.
of measurements the infants had to be brain can influence the results. Within the
clinically stable, without any cardiorespira- muscle, the estimated ratio is 10%:20%:70%,
respectively.3 In the brain, the mean arter-
CORRECTION
tory support. The measurements were car-
ried out with the NIRO-300 (Hamamatsu, ial:venous ratio is thought to be 16:84.4 doi:10.1136/adc.2006.105577corr1
Japan). Near-infrared light was transmitted Furthermore, cerebral autoregulation of oxy-
through the left frontoparietal side of the gen delivery may also have an important S H Dijkstra, A van Beek, J W Janssen, et al.
head (interoptode distance of 4 cm) and the role. High prevalence of vitamin D deficiency in
left lateral calf (interoptode distance of Simultaneous measurements and compar- newborns of high-risk mothers. This article
3 cm). Measurements were performed dur- isons of c-TOI and p-TOI might help in was published in print in Archives of Disease
ing undisturbed daytime sleep after a feed. future to detect early disturbances in circu- in Childhood (Arch Dis Child 2007;92:750–3)
The infants were lying in a horizontal lation and oxygenation, especially in states but was in fact an article for the Fetal &
position with the calf positioned just above of shock. Further studies should address Neonatal edition.

Comparison of peripheral and cerebral tissue

oxygenation index in neonates
K Grossauer, G Pichler, G Schmölzer, H Zotter, W Mueller and B Urlesberger

doi:10.1136/adc.2008.146654

These include:

Notes


PostScript

F Endo4
Age (days) 16 (18)
1
CORRECTION

CORRECTION

doi:10.1136/adc.2006.105577corr1

http://fn.bmj.com/cgi/content/full/94/2/F156-a
These include:
http://fn.bmj.com/cgi/eletter-submit/94/2/F156-a
Notes


PostScript

F Endo4
Age (days) 16 (18)
1
CORRECTION

Fantoms
Martin Ward Platt

These include:

Notes


Fantoms
Martin Ward Platt, Deputy Editor
Safety culture and the NICU uncommon the trial would need to be have a ‘better’, or more pleasant, NICU
Over recent years, we have carried a large, but the question is important. See experience—it just means that this
number of papers examining rates of page F120 dimension could not be measured using
adverse events in babies receiving inten- the rather crude tools at our disposal for
sive care. These have contained salutary measuring outcomes. The last word on
reminders of the possible harms that can
ECMO in the UK this has to go to Lewis Carroll again:
Extracorporeal membrane oxygenation ‘‘Contrariwise,’’ continued Tweedledee,
happen, and their frequency, but they
has now been an established mode of ‘‘if it was so, it might be; and if it were
have been less helpful in terms of gen-
therapy in the UK since the findings of so, it would be; but as it isn’t, it ain’t.
erating and testing practical measures by
the UK trial were published in 19961, yet That’s logic.’’ 3 See page F92
which errors might be reduced. We would
last year we carried a paper showing that
all sign up to the laudable aims of better
there was still unexplained differential
education, tight and simple systems, and
referral of babies for ECMO in the UK, Prone versus supine
close monitoring of errors and learning How times have changed. The paper by
begging the question as to whether all
from them when they occur, but even Saiki et al demonstrates that the prone
neonatologists are equally well informed
these do not reduce rates of error as far as position has clear benefits in terms of
about the potential benefits of ECMO
we would all wish. So it is particularly respiratory function for preterm babies
therapy2. This month, Karimova et al
valuable to have the paper by Lee et al studied at around 36 weeks postmenstrual
report the results of ECMO from the UK
bearing a very positive message by report- age. Thirty years ago such data would
registry for its first thirteen years, with
ing the application of structured random have been taken as good evidence that ex-
outcomes very similar to those achieved in
safety audits (a system widely used in prem babies would be ‘safer’ nursed prone
reports from around the world. This is
industry) to the NICU setting. In short: it than supine, and doubtless even more
good news, and should stimulate clin-
demonstrably works, and other units would have gone home with strict
icians who may be reluctant to consider
should give serious consideration to emu- instructions to their parents to put them
referral for ECMO to reconsider. See page
lating this system. See page F116 down to sleep on their tummies. How
F129
wrong we would have been (indeed, how
wrong we were). Now, in the light of
Feeding and NEC more sophisticated knowledge, we turn
You might have thought that the many Developmental care through a
these results on their head, and conclude
published case control studies investigat- looking glass that relative impairment of lung function
ing risks for necrotising enterocolitis ‘‘When I use a word,’’ Humpty Dumpty is not the reason that ex-prem babies are at
would have said everything that there is said, in rather a scornful tone, ‘‘it means higher risk of sudden unexplained death
to say on the subject. Not so. Henderson what I choose it to mean—neither more when placed prone, rather than supine,
et al, while confirming both that breast nor less.’’ 3 So with ‘developmental care’. for sleep. See page F133
feeding was protective and starting early Maguire et al have shown that when you
trophic feeding was not associated with choose it to mean incubator covers and
References
any increased risk, found that higher rates nesting, there is no demonstrable bene- 1. UK Collaborative ECMO Trial Group. UK
of increase of enteral feeding (with both ficial effect on babies’ outcomes. This is collaborative randomised trial of neonatal
shorter duration of trophic feeds and interesting as far as it goes, but we should extracorporeal membrane oxygenation. Lancet
1996;348:75–82.
earlier attainment of full feeds), was be careful. First, as the authors acknowl- 2. Tiruvoipati R, Pandya H, Manktelow B, et al.
associated with increased risk. The chal- edge, this does not invalidate or contradict Referral pattern of neonates with severe
lenge is now to design a simple pragmatic the studies demonstrating beneficial respiratory failure for extracorporeal membrane
trial comparing two different rates of effects of more integrated and extensive oxygenation. Arch Dis Child Fetal Neonatal Ed
2008;93:F104–F107.
increase in enteral feeds, with NEC free developmental care (eg NIDCAP). Second, 3. Carroll L. Through the Looking-Glass, and What Alice
survival as the endpoint. Since NEC is it does not mean that the babies did not Found There. London: Macmillan; 1871.

The effect of two levels of pressure support

ventilation on tidal volume delivery and minute
ventilation in preterm infants
S Gupta, S K Sinha and S M Donn
Arch. Dis. Child. Fetal Neonatal Ed. 2009;94;F80-F83; originally published online
1 Aug 2008;
doi:10.1136/adc.2007.123679

These include:

Notes


Original article
The effect of two levels of pressure support

ventilation on tidal volume delivery and minute
ventilation in preterm infants
S Gupta,1 S K Sinha,2 S M Donn3
1
University Hospital of North ABSTRACT
Tees, Stockton-on-Tees, UK; Objective: To study the effect of different levels of What is already known on this topic
2
University of Durham, The
James Cook University Hospital,
pressure support ventilation (PSV) on respiratory para-
Middlesbrough, UK; 3 Division of meters in preterm infants during the weaning phase of c Pressure support ventilation (PSV) supports
Neonatal-Perinatal Medicine, C. mechanical ventilation. spontaneous breaths by providing pressure
S. Mott Children’s Hospital, Design/methods: In this quasi-experimental crossover boost and thus should decrease the work of
University of Michigan Health
System, Ann Arbor, Michigan,
study, a total of 19 154 breaths were analysed from 10 breathing.
USA; Smdonnmd@med.umich. ventilated infants of ,32 weeks’ gestation. Breath-to- c Patients have control on how much to breath
edu breath data on minute ventilation, tidal volume, respiratory and for how long to breathe.
rate, peak inspiratory pressure and mean airway pressure c PSV has been described in adult and paediatric
Correspondence to: were collected during three study epochs: synchronised
Sunil K Sinha, Professor of intensive care but the data regarding its use in
Paediatrics, University of intermittent mandatory ventilation (SIMV) alone, SIMV the neonatal population are limited.
Durham, Consultant in with partial PSV (PSmin), and SIMV with full PSV (PSmax).
Paediatrics and Neonatal PSmin was set to provide an exhaled tidal volume (VTe)
Medicine, The James Cook
University Hospital, between 2.5–4 ml/kg and PSmax 5–8 ml/kg VTe.
Middlesbrough, UK; Sunil. Statistical analyses were performed using analysis of What this study adds
sinha@stees.nhs.uk variance (ANOVA) for repeated measures.
Results: The addition of full PSV (PSmax) was associated with c Pressure support ventilation increases total
Accepted 3 July 2008 a significant increase in total minute ventilation as compared minute ventilation and stabilises the breathing
Published Online First pattern in preterm infants.
1 August 2008 with SIMV alone (392 ml/kg/min vs 270 ml/kg/min,
respectively; p,0.05). This difference in minute ventilation c The increase in respiratory efficiency is
was still present when PSmin was used (332 ml/kg/min as proportional to the level of pressure support-
compared with 270 ml/kg/min in SIMV; p,0.05). There was provided ventilation.
also a concomitant decrease in the respiratory rate with both
PSmax (59 breaths per minute) and PSmin (65 breaths per
minute) compared with SIMV alone (72 breaths per min) flow (fig 1).1 There are important differences
(p,0.05). between PSV and traditional time-cycled, pres-
Conclusions: Pressure support ventilation increases total sure-limited (TCPL) ventilation. In TCPL, circuit
minute ventilation and stabilises breathing in proportion to gas flow is fixed and the inspiratory time is set by
the level of pressure support used. This may be the clinician, whereas in PSV, inspiratory gas flow
advantageous and provide a useful ventilation strategy for is variable and is proportional to patient demand.
use during weaning stages of mechanical ventilation in The inspiratory phase in PSV is flow-cycled, and
preterm infants. inspiration ends when flow has decelerated to a
small percentage of peak. Because of the ability to
control inspiratory time and rate, synchronisation
Pressure support ventilation (PSV) is a ventilatory occurs during both the inspiratory and expiratory
modality in which patients’ spontaneous breaths phases. Thus, patients have control of how much to
are supported by an inspiratory pressure ‘‘boost’’ breathe (inspiratory flow and tidal volume) and for
above the baseline pressure. This is designed to how long to breathe (inspiratory time), thus mimick-
decrease the imposed work of breathing created by ing physiological breathing.2
the narrow lumen endotracheal tube and ventilator PSV has been shown to reduce the work of
circuit, and to facilitate weaning. It is a form of breathing and oxygen requirement in intubated
patient-triggered ventilation, which can be used adult and paediatric patients.3 4 There is no
either alone in patients with reliable respiratory consensus as to what the most appropriate level
drive or in conjunction with other modes of of PSV is in neonatal patients. A recent study by
ventilation in patients who have poor or unreliable Osorio et al5 reported the effect of two levels of
respiratory drive. When used in conjunction with pressure support as an adjunct to SIMV in
synchronised intermittent mandatory ventilation ventilated premature infants. They chose arbitrary
(SIMV) either in pressure- or volume-targeted levels of pressure support (3 and 6 cm H2O) and
modalities, PSV augments the tidal volume gener- weaned babies by decreasing the SIMV rate. As
ated by a patient’s own spontaneous breathing. tidal volume delivery was not targeted in this
The three distinct characteristics of PSV are: (1) study, the interpretation about the effectiveness of
triggering; (2) pressurisation; and (3) cycling, pressure support becomes difficult, especially at the
which are accomplished using changes in airway lower level, which might not have been sufficient

Original article
Figure 1 Schematic diagram showing

features of pressure support ventilation
(PSV): (1) triggering; (2) pressurisation;
and (3) cycling.
to generate effective tidal volume delivery. In another published (Microsoft Excel, Microsoft Corp., Redmond, Washington) for
study, Migliori et al6 targeted the exhaled tidal volume (VTe) further analyses.
delivery only at 6 ml/kg and did not study the effect of different Babies were randomly assigned to one of the three study
levels of PSV. modes for 30-minute epochs — SIMV, SIMV with partial PSV
In the present study, we assessed the effects of two different (PSmin) or SIMV with full PSV (PSmax), and then switched to
levels of pressure support: full pressure support or partial another study mode in a random order. PSmax was targeted to
pressure support (PSmax and PSmin, respectively) on respiratory deliver a VTe of 5–8 ml/kg to match the SIMV breaths, and
parameters, and compared their effectiveness in relation to PSmin was adjusted to deliver 2.5–4 ml/kg VTe. The mandatory
SIMV during weaning from mechanical ventilation in preterm machine breaths in all three study modes were kept at 20
infants. breaths per minute. The spontaneous breaths in babies receiving
SIMV alone were supported only with PEEP. When pressure
support was added the same spontaneous breaths were either
PATIENTS AND METHODS half or fully supported to become partial PSV or full PSV,
This was a quasi-experimental study design, crossover trial, respectively. This allowed us to compare the effect of different
carried on the Neonatal Unit at James Cook University levels of pressure support by augmenting the spontaneous
Hospital, Middlesbrough, UK. The study was part of an breaths and comparing across the study groups.
ongoing, randomised, controlled trial, which was approved by Real-time breath-to-breath pulmonary mechanics data were
the Institutional Review Board, and written informed consent collected during each study epoch. This included inspired tidal
was obtained from each parent prior to study entry. Babies volume (VTi) and VTe, spontaneous and total minute ventila-
,32 weeks’ gestation and receiving assisted ventilation for tion (Ve), total respiratory rate (spontaneous plus mechanical
respiratory distress syndrome (RDS) were eligible for enrolment breaths, RR), FiO2, and peak inspiratory pressure (PIP). A re-
during the recovery phase. To meet entry criteria, the mean equilibration period of 15 minutes was used between the
airway pressure (mean Paw) had to be ,10 cm H2O, and epochs. In order to avoid spill-over from the previous mode,
fraction of inspired oxygen (FiO2) ,0.4. All infants received 10 minutes of artefact-free data were extracted from the latter
caffeine citrate (20 mg/kg loading, 5 mg/kg/day maintenance half of each epoch, and subsequently integrated for analysis.
dose) at entry into the study, and demonstrated reliable The demographic data for gender, gestational age, postnatal age,
respiratory drive, defined as a spontaneous respiratory rate at and weight at birth and at study entry were recorded for each
least 25% higher than the ventilator rate.6 Newborns with subject. During the recording of data, handling of babies was
systemic or thoracic malformations, neuromuscular diseases, or not permitted; however, FiO2 was adjusted to maintain the
thoracic air leaks were not eligible for inclusion. pulse oximeter reading in a target range of 88%–92%. The rapid
According to a standardised unit protocol, all babies were shallow breathing index (RSBI) was used as a measure of the
ventilated with the AveaH ventilator (Viasys Healthcare, Yorba efficiency of breathing. This was calculated by taking the ratio
Linda, California, USA) using either TCPL or volume-controlled of respiratory rate to tidal volume (RR/VTe) and expressed as
(VC) ventilation. At entry into the trial, they were on low-rate breaths/min/ml/kg and compared across the epochs. Using the
SIMV (20 breaths/min), combined with PSV. The peak RSBI, the lower the ratio, the better is the efficiency of
inspiratory pressure for SIMV breaths was set between 12 and spontaneous breathing.
16 cm H2O to provide a VTe of 5–8 ml/kg. The mean value for each outcome measure was calculated on
Both the mechanical and spontaneous respiratory parameters each ventilation mode for every baby and then compared
were displayed on the digital interface graphic monitor. The between the study epochs. This was then utilised for the
Medical Information Bus (MIB) interface from the monitor was statistical analyses and compared between the groups.
used to download continuous breath-to-breath data to a secure Statistical analyses were performed using analysis of variance
computer using a proprietary ‘‘GSP Interface Kit’’ and research (ANOVA) for repeated measures and the post hoc Bonferroni
tool software (Viasys Healthcare, Yorba Linda, California, test, to evaluate differences in respiratory parameters between
USA). These data were then exported to a spreadsheet the three study groups. Outcome comparisons between the

Original article
groups were also carried out using the two-sided t test for Table 1 Effect of different levels of pressure support on study
parametric data and the Mann–Whitney U test for non- parameters
parametric data. All statistical analyses were performed using SIMV SIMV + partial SIMV + full PSV
SPSS Inc., version 12 for WindowsTM (Chicago, Illinois, USA). Study parameter (SIMV) PSV (PSmin) (PSmax)
Mean (SD) exhaled tidal 3.9 (0.72) 5.2 (1.29)* 6.7 (0.69)*{
RESULTS volume(ml/kg)
Ten ventilated babies were enrolled in the study. Their mean Mean (SD) total Ve (ml/kg/min) 270 (47) 332 (53)* 392 (63)*{
gestational age and birth weight were 28 weeks and 1190 g, Mean (SD) total RR (breaths/ 72 (5.6) 65 (8.6)* 59 (9.8)*{
min)
respectively. The mean age at entry into the study was 16 days.
All babies successfully completed the study and provided a total *p,0.05 (PSV vs SIMV).
{p,0.05 (partial vs full PSV).
of 19 154 breaths for analyses. There were equal numbers of Ve, minute ventilation; PSV, pressure support ventilation; RR, respiratory rate; SIMV,
babies receiving TCPL and VC ventilation. synchronised intermittent mandatory ventilation.
The ventilator parameters were adjusted to achieve the
desired VTe and the data for partial pressure of CO2 (pCO2) significantly and proportionally with the addition of PSV
were collected before and after the cross-over study period on all (table 2).
infants. There were no differences in the pCO2 levels and
clinical status before or after the study period. There were no
major episodes of desaturation during the study period. DISCUSSION
There was a significant increase in minute ventilation during Pressure support ventilation is a ventilatory mode in which
PSmin compared with SIMV alone. This was associated with a spontaneous breaths are partially or fully supported by an
concomitant decrease in the total respiratory rate (table 1). inspiratory pressure assist above the baseline pressure. During
These differences in minute ventilation and total respiratory weaning of mechanical ventilation, spontaneous breathing
rate were even more marked when full pressure support was must overcome the work imposed by the presence of a high-
applied (fig 2). There was also a statistically significant resistance endotracheal tube, ventilatory circuit and the disease-
difference in the observed parameters when PSmax was induced respiratory load. Pressure support can be used as an
compared with PSmin (total minute ventilation 392 ml/kg/min adjunct to SIMV to partially or fully unload the spontaneous
in PSmax vs 332 ml/kg/min in PSmin; p,0.05; and total breaths.5 PSV increases the tidal volume proportionally to the
respiratory rate 59 breaths per min in PSmax vs 65 breaths per chosen pressure support level, achieves synchrony during both
min in PSmin; p,0.05). There was also a significant decrease in inspiration and expiration, and enhances the efficiency of
the mean (SD) RSBI during PSmin (17.4 (0.31) breaths/min/ml/ breathing.7 PSV has been shown to reduce the work of breathing
kg) and PSmax (10.2 (0.24) breaths/min/ml/kg) compared with and oxygen requirement in intubated adult and paediatric
SIMV alone (42.9 (1.00) breaths/min/ml/kg). Both of these patients3 4 and appears to accelerate weaning from mechanical
differences were statistically significant (p,0.05). ventilation and to reduce ventilatory dependency.8 9 PSV with
There was also an incremental effect on mean Paw, PIP and SIMV has been compared with SIMV alone in a randomised
FiO2 according to the level of pressure support. The mean Paw controlled trial and was reported to have advantages as
increased significantly with the addition of PSV compared with compared with SIMV alone for weaning. This study, however,
SIMV alone. The increase in mean Paw and PIP was greater only used pressure support of 30%–50%.10 PSV has been
with full PSV compared with partial PSV. The FiO2 decreased reported to be beneficial but there are not yet sufficient data
to compare its effectiveness at different levels. This is important
to know, as PSV can be used for both full and partial support
and may offer advantages over other modalities of ventilation,
especially during weaning, as it resembles physiological breaths.
In the present study, each baby served as his/her own control.
The mandatory SIMV rate was kept constant at 20 breaths per
minute throughout the study period to enable assessment of the
effects of two levels of pressure support, which were aimed to
maintain a desired range of VTe. This enabled us to standardise
the level of PS across the study epochs.
Data from this study suggest that the use of PSV increases
total minute ventilation and stabilises the breathing pattern,
Table 2 Effect of the level of pressure support on ventilation

parameters
SIMV SIMV + partial SIMV + full PSV
Study parameter (SIMV) PSV (PSmin) (PSmax)
Mean (SD) airway pressure (cm 5.60 (0.84) 6.83 (0.89)* 7.85 (1.3)*
H2O)
Figure 2 Diagram showing the effect of two different levels of pressure Mean (SD) peak inspiratory 14.4 (3.2) 11.3 (1.2)* 15.7 (2.1){
support on minute ventilation and respiratory rate. The mandatory rate is pressure (cm H2O)
fixed across the three groups: synchronised intermittent mandatory Mean (SD) FiO2 (%) 24.2 (2.2) 23.6 (1.5)* 21.5 (1.2)*{
ventilation (SIMV), SIMV with partial pressure support (PSmin) and SIMV
*p,0.05 (PSV vs SIMV).
with full pressure support (PSmax). The effect of partial or full pressure {p,0.05 (partial vs full PSV).
support on total minute ventilation (Ve/kg) and total respiratory rate is FiO2, fraction of inspired oxygen; PSmin, partial PSV; PSmax, full PSV; PSV, pressure
shown. RR, respiratory rate. support ventilation; SIMV, synchronised intermittent mandatory ventilation.

Original article
while enabling the baby to decrease the total respiratory rate. findings, different levels of pressure support can be used to
The significantly lower RSBI observed in this study when PSmax facilitate a gradual shift of work of breathing from the ventilator to
was used (compared with PSmin), suggests that there is an the baby, and conditioning the respiratory muscles for successful
incremental increase in respiratory efficiency by providing more extubation. Further clinical trials are needed to fully explore the
pressure support, perhaps by augmenting the unloading of safety and efficacy of different levels of pressure support in
respiratory muscles. This has not been reported before in newborn infants requiring mechanical ventilation.
newborns but mirrors the observations made in paediatric and
adult patients.11 There are limited data comparing various levels Competing interests: None.
of PSV,12–14 but these studies suggest that the addition of PSV Patient consent: Written informed consent was obtained from each parent prior to study
during weaning augments spontaneous breathing with better entry.
thoraco-abdominal synchrony, increase in minute ventilation,
and reduction in the total respiratory rate compared with SIMV REFERENCES
alone.15 1. Sinha SK, Donn SM. Pressure support ventilation. In: Donn SM, ed. Neonatal and
Addition of PSV in the present study increased mean Paw and pediatric pulmonary graphics: principles and clinical application. Armonk NY: Futura
Publishing Co., 1998: 301–12.
PIP. This has been observed before.6 While this may be a source 2. Sarkar S, Donn SM. In support of pressure support. Clin Perinatol 2007;34:117–28.
of anxiety, it should be realised that PSV is flow-cycled, and 3. El-Khatib MF, Chatburn RL, Potts DL, et al. Mechanical ventilators optimised for
inspiratory gas flow delivery is variable according to the patient paediatric use decrease work of breathing and oxygen consumption during pressure
effort. This, in fact, makes PSV more physiological than the support ventilation. Crit Care Med 1994;22:1942–8.
4. Tokioka H, Kinjo M, Hirakawa M. The effectiveness of pressure support ventilation
mechanically delivered mandatory breaths, which are set by the for mechanical ventilatory support in children. Anaesthesiology 1993;78:880–4.
operator and are thus more prone to result in patient–ventilator 5. Osorio W, Claure N, D’Ugard C, et al. Effects of pressure support during an acute
asynchrony. The effect of PSV in improving minute ventilation reduction of synchronized intermittent mandatory ventilation in preterm infants.
with a better RSBI through reduction in the spontaneous J Perinatol 2005;25:412–16.
6. Migliori C, Cavazza A, Motta M, et al. Effect on respiratory function of pressure
respiratory rate may be explained by the optimisation of tidal support ventilation versus synchronised intermittent mandatory ventilation in preterm
volume delivery and patient–ventilator synchrony. infants. Pediatr Pulmonol 2003;35:364–7.
The decrease in FiO2 observed with the addition of PSV may 7. Tokioka H, Saito S, Kosaka F. Effect of pressure support ventilation on breathing
have resulted from better alveolar recruitment. Another patterns and respiratory work. Intensive Care Med 1989;15:491–4.
8. Gulberg N, Winberg P, Selldén H. Pressure support ventilation increases cardiac
possibility is the opportunity for the infant to sigh during output in neonates and infants. Paediatr Anaesth 1996;6:311–15.
PSV through voluntary extension of the inspiratory time, which 9. Brochard L, Rauss A, Benito S, et al. Comparison of three methods of gradual
could also contribute to improved oxygenation. This observa- withdrawal from ventilatory support during weaning from mechanical ventilation.
tion of a decrease in the oxygen requirement with increasing Am J Respir Crit Care Med 1994;150:896–903.
10. Reys ZC, Claure N, Tauscher MK, et al. Randomized controlled trial comparing
pressure support is important while weaning preterm babies synchronised intermittent mandatory ventilation and synchronised intermittent
who require prolonged ventilatory support. mandatory ventilation plus pressure support in preterm infants. Pediatrics
Because of lack of availability of continuous blood gases 2006;118:1409–17.
monitoring devices, we were limited in our ability to collect 11. Leung P, Jubran A, Tobin MJ. Comparison of assisted ventilator modes on triggering,
patient effort and dyspnoea. Am J Respir Crit Care Med 1997;155:1940–8.
continuous data on pCO2. Nonetheless, our data suggest that 12. Olsen SL, Thibeault DW, Truog WE. Crossover trial comparing pressure support with
the infants probably adjust spontaneous breathing, and thus synchronized intermittent mandatory ventilation. J Perinatol 2002;22:461–6.
minute ventilation, as a way to maintain normocapnia, and 13. Nafday SM, Green RS, Lin J, et al. Is there an advantage of using pressure support
that higher PSV levels provide a higher tidal volume, requiring ventilation with volume guarantee in the initial management of premature infants with
respiratory distress syndrome? A pilot study. J Perinatol 2005;25:193–7.
fewer spontaneous breaths. The previous study by Osorio also 14. Abd El-Moneim ES, Fuerste HO, Krueger M, et al. Pressure support ventilation
did not show any changes in pCO2.5 combined with volume guarantee versus synchronized intermittent mandatory
In summary, the findings of this study confirm that PSV ventilation: a pilot crossover trial in premature infants in their weaning phase. Pediatr
Crit Care Med 2005;6:286–92.
enhances the efficiency of spontaneous breathing during weaning
15. Tokioka H, Nagano O, Ohta Y, et al. Pressure support ventilation augments
of preterm infants from mechanical ventilation. The increase is spontaneous breathing with improved thoracoabdominal synchrony in neonates with
proportional to the level of pressure support used. Based on our congenital heart disease. Anesth Analg 1997;85:789–93.

Volume-guarantee ventilation: pressure may

decrease during obstructed flow
K I Wheeler, C J Morley, C O F Kamlin and P G Davis
13 Aug 2008;
doi:10.1136/adc.2008.139980

These include:

Notes


Original article
Volume-guarantee ventilation: pressure may

decrease during obstructed flow
K I Wheeler, C J Morley, C O F Kamlin, P G Davis
Neonatal Services, Royal ABSTRACT

Women’s Hospital, Melbourne, Background: Two unexpected observations were made What is already known on this topic
Australia
during ventilation with the Dräger Babylog 8000+ in
Correspondence to: volume-guarantee mode: (a) during complete obstruction c During volume-guarantee ventilation with the
Dr K I Wheeler, Neonatal to gas flow down the endotracheal tube (ETT), positive Dräger Babylog 8000+, peak inspiratory
Services, Royal Women’s inspiratory pressure (PIP) was reduced to half way
Hospital, Locked Bag 300, Cnr pressure (PIP) changes to target a set expired
Grattan Street & Flemington between the maximum inflating pressure and the positive tidal volume.
Road, Parkville, VIC 3052, end expiratory pressure (PEEP) even though the set c When the endotracheal tube (ETT) flow is
Australia; kevin.wheeler@ expired tidal volume had not been achieved; (b) an partially obstructed during volume-guarantee
thewomens.org.au
external Dräger waveform monitor may stop displaying ventilation, the PIP increases to the set
real-time waveforms when a tube-obstructed alarm is maximum. When the manual breath button is
Published Online First activated. pushed, the maximum PIP is delivered
13 August 2008 Objective: To investigate these phenomena using a test regardless of the ventilation mode.
lung.
Method: A 50 ml Dräger test lung was attached to the
ventilation circuit of a Dräger Babylog 8000+. Partial
obstruction to ETT flow was induced by compressing the What this study adds
tubing leading to the test lung, and complete obstruction
was achieved by clamping. Recordings were made from c When the ETT flow is completely obstructed
the digital output of the ventilator at 125 Hz. during VG ventilation, the Dräger Babylog 8000+
Results: When the ETT flow was completely obstructed uses a constant PIP set midway between the set
during VG ventilation, a constant PIP was set midway maximum and positive end expiratory pressure.
between the set maximum and PEEP. This did not happen This does not happen during partial obstruction.
during partial obstruction. The external waveform monitor c An external waveform monitor display may
display froze when ETT flow was completely obstructed. freeze when ETT flow is completely obstructed.
Conclusions: During complete ETT obstruction, the PIP is
set to a pressure midway between maximum PIP and
PEEP even if this is less than the PIP used before the clinicians increased Pmax from 25 to 30 cm H2O
obstruction. Further research is needed to evaluate (fig 1, epoch D).
whether this reduction in PIP is associated with
prolongation of precipitating events. VG VENTILATION: TARGETING TIDAL VOLUME
VG is a mode of volume-targeted ventilation
provided by the Dräger Babylog 8000+ ventilator
An extremely-low-birthweight infant was deliv- (Drägerwerk AG, Lübeck, Germany).1 2 A flow
ered prematurely at less than 26 weeks’ gestation, sensor measures inspired and expired gas flow
requiring intubation in the delivery room. Assist- through the ETT and integrates these to inspired
control, volume-guarantee (AC/VG) ventilation and expired tidal volumes (VTi and VTe). The
with the Dräger Babylog 8000+ was started on ventilator adjusts PIP to achieve a set VTe.3 The PIP
the neonatal intensive care unit with a target can vary between the set PEEP and a set maximum
expired tidal volume (VTset) of 4.3 ml (4.5 ml/ pressure limit (Pmax), referred to as Pinsp in the
kg), a maximum inflating pressure limit (Pmax) of Dräger manual.2 Factors that affect the VTe and
25 cm H2O and a positive end expiratory pressure therefore the PIP include complete or partial
(PEEP) of 5 cm H2O (fig 1). Curosurf (100 ml/kg) airway occlusion, altered airway resistance,
was rapidly administered through a closed-circuit changes in lung or chest wall compliance, ETT
device (Trach Care Multi-Access Catheter; leak, and spontaneous breathing.4 5
Kimberley-Clark, Ballard Medical Products, Before observing this phenomenon, we assumed
Roswell, Georgia, USA) while AC/VG ventilation that in VG mode the ventilator would always
continued. Complete endotracheal tube (ETT) increase the PIP up to Pmax if the VTe was less than
obstruction occurred immediately and lasted for VTset. We therefore studied the response of the
45 s (fig 1, epoch B). Recorded ventilator pressure Dräger Babylog 8000+ to episodes of partially and
and flow waves showed that the baby received a completely obstructed ETT flow using a test lung.
peak inspiratory pressure (PIP) of only
15 cm H2O during this time. As soon as ETT METHODS
flow returned (fig 1, epoch C), the PIP increased A 50 ml Dräger test lung was attached to the wye
to Pmax over several inflations. Because of hypoxia piece of the ventilation circuit of a Dräger Babylog
and because the VTset was not achieved, the 8000+. The settings were: AC/VG mode, back-up

Original article
rate 50/min, PEEP 5 cm H2O, Pmax 25 cm H2O, VTset 5 ml and External waveform monitoring
circuit flow 8 litres/min. The PIP delivered was 17–18 cm H2O. The waveform displayed on the ventilator’s inbuilt LCD display
Partial obstruction to ETT flow was induced by compressing shows information in real-time. However, with an external
the tubing leading to the test lung. Clamping this produced Dräger waveform monitor (eg, VentView or BabyView), the
complete obstruction. reduction in PIP during complete ETT obstruction was not seen
Recordings were made from the digital output of the because this display froze during the tube-obstructed alarm. The
ventilator at 125 Hz using Spectra software (Grove Medical, waveform display recommenced when ETT flow was restored.
London, UK). Dräger acknowledged that this previously unrecognised phe-
nomenon was unintended (Dragerwerk AG Lubeck Germany,
Personal Communication, 4 April 2008).
RESULTS
Partial ETT obstruction (fig 2, epoch B)
When the ETT flow was partially obstructed, the VTe was DISCUSSION
reduced. The ventilator increased the PIP to Pmax to try to During partially obstructed ETT flow, the Babylog 8000+ in VG
achieve VTset. mode increased the PIP to Pmax. If VTset was not reached, the
ventilator activated the alarm. During complete ETT obstruc-
Complete ETT obstruction (fig 2, epoch C) tion, the ventilator immediately reduced the PIP to midway
During complete ETT obstruction, flow stopped and VTe was between Pmax and PEEP, despite not having delivered VTset. The
zero. After a single cycle with no flow, the ventilator reduced reduction occurred after one cycle with zero ETT flow.
the PIP from 25 to 15 cm H2O (the midpoint of Pmax and PEEP). Dräger have confirmed these observations, attached as
When ETT flow resumed (fig 2, epoch E), PIP increased stepwise appendix 1 (Dragerwerk, Personal Communication 26
to Pmax. This reproduced our clinical observations when the November 2007 and 28 March 2008). They informed us that
ETT was completely blocked after surfactant administration. this is designed to avoid excessive pressures after resolution of
When the obstruction resolved (fig 2, epoch F), the VTe the tube-obstructed alarm. For example, in VG mode with a
transiently exceeded the VTset and the PIP decreased until VTe Pmax of 30 cm H2O and PEEP 6 cm H2O (maximum inflating
equalled VTset. pressure = 24 cm H2O), the ventilator will reduce the inflating
pressure to 12 cm H2O above PEEP. The PIP is limited to
18 cm H2O until the tube-obstructed alarm resolves.
Manual breath inflations (fig 2, epoch D)
VG is used to automatically control PIP to target a tidal
The effect of manual inflations during complete obstruction to
volume. The doctors in our unit expected PIP to be increased to
ETT flow was investigated with inflations delivered at a
Pmax during partial or complete ETT obstruction. We were
constant PIP at Pmax. The inflation was sustained for as long
unaware that the Babylog ventilator reduces PIP during
as the manual breath button was held (up to 5 s).
complete ETT obstruction. We speculate that this is because
information about ventilator functioning during partial and
Ventilation without volume guarantee complete ETT obstruction is not included in the product
When the VG mode is not used, the PIP is constant for each manual. Therefore this attenuation of PIP may not be apparent
inflation. During both partial and complete obstruction to ETT in neonatal units that use an external Dräger VentView display
flow, inflations are delivered at the set PIP regardless of the tidal to show waveforms, and will be particularly marked in units
volume delivered. that set Pmax close to delivered PIP.
Figure 1 A recording from a Dräger

Babylog 8000+ ventilating an extremely
low birthweight infant, showing
inspiratory and expiratory endotracheal
flow, inflating pressures and tidal volumes
around the time of surfactant
administration. The initial ventilator
settings were: assist-control, volume-
guarantee mode, back-up rate 50/min,
VTset 4.3 ml, Pmax 25 cm H2O, positive
end expiratory pressure 5 cm H2O,
inflation time 0.3 s, circuit flow 8 litres/
min. Four epochs are shown. (A) Set tidal
volumes generally not obtained despite
peak inspiratory pressure (PIP) having
increased to Pmax. (B) Surfactant given
followed by complete endotracheal tube
(ETT) obstruction. PIP decreases from 25/
5 to 15/5. (C) Obstruction partially
resolves, PIP increases back up to Pmax.
Despite this, VTe remains below VTset. (D)
Clinical team increases Pmax to
30 cm H2O. VTe reaches VTset with
varying PIP.

Original article
Figure 2 Recording of study on a 50 ml

test lung with complete and partially
obstructed endotracheal tube (ETT) flow.
Ventilator settings: assist-control,
volume-guarantee mode, back-up rate 50/
min, VTset 5 ml, Pmax 25 cm H2O, positive
end expiratory pressure 5 cm H2O,
inflation time 0.3 s, circuit flow 8 litres/
min. Six epochs are shown. (A) Baseline
period with unobstructed ETT flow.
Ventilator delivers a VTe equal to the set
VTe using an inflating pressure of about
17–18 cm H2O. (B) Partially obstructed
ETT flow. VTe is initially reduced. The PIP
increases over three inflations until the
VTe reaches the VTset. The peak
inspiratory pressure (PIP) reaches Pmax.
(C) Completely obstructed ETT flow. After
the first inflation without flow, the PIP
falls to almost half the previous PIP, to
,15 cm H2O. (D) Five manual inflations
are delivered by pushing the manual
breath button. These are delivered at
Pmax. After these, the ventilator returns to
a PIP of 15 cm H2O. (E) Partial flow
restored. The ventilator increases the PIP
from 15 cm H2O to Pmax as VTe is less
than VTset. (F) Resolution: flow is no
longer obstructed. PIP decreases back to
17–18 cm H2O as initial VTe exceeds
VTset.
We speculate that delivering an increased PIP during REFERENCES

obstruction may help in overcoming the resistance to flow in 1. Grover A, Field D. Volume-targeted ventilation in the neonate: time to change? Arch
Dis Child Fetal Neonatal Ed 2008;93:F7–13.
some situations, and, if PIP is reduced during complete ETT
2. Dragerwerk AGLG. Babylog 8000 Plus: intensive care ventilator for neonates.
obstruction, the precipitating situation may take longer to Instructions for use. Software 5.n. 2008.
resolve. Further work is needed to clarify this, including 3. Ahluwalia JS, Morley CJ, Wahle H. Volume guarantee: new approaches in volume
determining the optimal settings for Pmax during VG ventila- controlled ventilation for neonates. Lubeck: Drager Medizintechnik GmbH.
4. Jaecklin T, Morel DR, Rimensberger PC. Volume-targeted modes of modern neonatal
tion, and minimisation of ETT obstruction events. ventilators: how stable is the delivered tidal volume? Intensive Care Med
2007;33:326–35.
SUMMARY 5. Esquer C, Claure N, D’Ugard C, et al. Role of abdominal muscles activity on duration
and severity of hypoxemia episodes in mechanically ventilated preterm infants.
During complete ETT obstruction, the Dräger Babylog 8000+ Neonatology 2007;92:182–6.
ventilator in VG mode sets PIP to a pressure midway between
maximum PIP and PEEP even if this is less than the PIP used
before the obstruction. Manual breath inflations at Pmax are APPENDIX 1
delivered in both VG and non-VG modes. In non-VG ventila- Communication from Dräger regarding volume guarantee mode during ETT
obstruction. ‘‘While working in Volume Guarantee, the Babylog needs the calculated
tion, the set PIP is delivered with each inflation whether there is tidal volumes in order to be able to regulate the pressure for the next breath. As far
ETT obstruction or not. as the Babylog detects flow going to or coming from the patient the regulator stays
Further research is needed to evaluate whether reducing the in action and tries to achieve the set target tidal volume. Even in case of partial
PIP during complete ETT obstruction is associated with a obstruction the regulator will increase the PIP in order to compensate for the
prolongation of precipitating events. additional resistance. But as soon as the ventilator cannot detect any flow going to
the patient or coming from the patient, it will alarm with a tube obstruction alarm,
as in this case the ventilator seems not to be able to solve the problem and needs
Acknowledgements: We thank Research Nurse Connie Wong for her assistance help and observation by the user. If the alarm is activated the regulator goes into a
with this project. kind of waiting status, stopping temporarily its regulation, setting the PIP in
Funding: Funded by Australian National Health and Medical Research Council Program between Pinsp and PEEP, providing enough pressure to detect a reopening of the
Grant No 384100. obstruction and at the same time preventing too much pressure with the first breath
after reopening the obstruction. As soon as the first flow is detected again the
Competing interests: None. regulator will restart its work again and provide the needed pressure after very short
Ethics approval: Obtained. time.’’

Oxygen saturation and heart rate during delivery

room resuscitation of infants <30 weeks’
gestation with air or 100% oxygen
J A Dawson, C O F Kamlin, C Wong, A B te Pas, C P F O’Donnell, S M Donath,
P G Davis and C J Morley
14 Aug 2008;
doi:10.1136/adc.2008.141341

These include:

Notes


Original article
Oxygen saturation and heart rate during delivery

room resuscitation of infants ,30 weeks’ gestation
with air or 100% oxygen
J A Dawson,1,2 C O F Kamlin,1 C Wong,1 A B te Pas,1 C P F O’Donnell,3 S M Donath,4
P G Davis,1,2 C J Morley1,2,5
1
Neonatal Services, The Royal ABSTRACT
Women’s Hospital, Melbourne, Background: Because of concerns about harmful effects What is already known on this topic
Australia; 2 Department of
Obstetrics and Gynaecology,
of 100% oxygen on newborn infants, air has started to be
University of Melbourne, used for resuscitation in the delivery room. c Brief exposure to supplemental oxygen in the
Australia; 3 National Maternity Objective: To describe changes in preductal oxygen delivery room can produce an SpO2 >95%.
Hospital, Dublin, Ireland; saturation (SpO2) and heart rate (HR) in the first 10 min
4 c Preductal oximetry measures SpO2 and heart
Clinical Epidemiology and
after birth in very preterm infants initially resuscitated rate within 90 s of birth.
Biostatistics Unit, Murdoch
Children’s Research Institute, with 100% oxygen (OX100) or air (OX21). c Using oximetry in the delivery room, clinicians
Melbourne, Australia; 5 Murdoch Patients and methods: In July 2006, policy changed can adjust the FiO2 to the SpO2.
Children’s Research Institute, from using 100% oxygen to air. Observations of SpO2 and
Melbourne, Australia HR before and after the change were recorded whenever
Correspondence to: a member of the research team was available to attend
Jennifer Dawson, Neonatal the birth. What this study adds
Services, The Royal Women’s Results: There were 20 infants in the OX100 group and
Hospital, 20 Flemington Road,
Parkville, VIC 3052, Australia;
106 in the OX21 group. In the OX100 group, SpO2 had risen c When very preterm infants are initially
jennifer.dawson@thewomens. to a median of 84% after 2 min and 94% by 5 min. In the resuscitated with air, some will require
org.au OX21 group, median SpO2 was 31% at 2 min and 54% at supplemental oxygen.
5 min. In the OX21 group, 92% received supplemental c If very preterm infants are initially resuscitated
Accepted 27 July 2008 oxygen at a median of 5 min; the SpO2 rose to a median
Published Online First with 100% oxygen, many will rapidly become
of 81% by 6 min. In the first 10 min after birth, 80% and hyperoxic.
14 August 2008
55% of infants in the OX100 and OX21 groups, respectively, c Titrating oxygen administration may reduce the
had an SpO2 >95%. Increases in HR over the first 10 min number of very preterm infants with SpO2
were very similar in the two groups. measurements >95%.
Conclusions: Most very preterm infants received
supplemental oxygen if air was used for the initial
resuscitation. In these infants, the use of backup 100% Australian Resuscitation Council.10 Before the
oxygen and titration against SpO2 resulted in a similar change, 100% oxygen was used throughout resusci-
course to ‘‘normal’’ term and preterm infants. Of the tation. After the change, air was used with 100%
infants resuscitated with 100% oxygen, 80% had SpO2 oxygen as a backup guided by ranges of oxygen
>95% during the first 10 min. The HR changes in the two saturations previously seen in term and preterm
groups were very similar. infants not requiring resuscitation.13 14 We aimed to
describe the changes in oxygen saturation (SpO2) and
heart rate (HR) of infants ,30 weeks’ gestation
For many years, 100% oxygen was recommended
resuscitated in the two eras.
for delivery room (DR) resuscitation of newborn
infants of all gestational ages.1 In recent years,
experts have suggested that even a brief exposure PATIENTS AND METHODS
to high oxygen concentrations at birth in very-low- In this prospective observational study, the cohorts
birthweight infants is harmful.2 There is also included infants ,30 weeks’ gestation, born at The
accumulating evidence of oxygen toxicity from Royal Women’s Hospital, Melbourne between 12
animal and in vitro studies.3–5 Several studies have January 2006 and 31 December 2007, when a
found evidence of oxidative damage in infants after member of the research team was available to
short exposure to 100% oxygen in the DR.6–9 This attend the birth. This included 6 months before
evidence has led to a change in national guidelines (OX100) and 18 months after (OX21) 17 July 2006,
for DR resuscitation, which now advise that 21% when the change in policy and implementation of
oxygen should be considered rather than 100% the new guideline took place.
oxygen during initial resuscitation of all infants.10 11 Immediately after birth, an oximetry sensor
Very preterm infants appear to be at greatest risk (LNOP Neo sensor; Masimo, Irvine, California,
of oxidative damage.7 9 12 Data from infants born at USA) was placed on the infant’s right hand and
,30 weeks’ gestation are limited. then connected to the oximeter (Radical7 V5;
The protocol for resuscitation of newborn infants Masimo), as previously described.15 SpO2, HR and
at The Royal Women’s Hospital, Melbourne was signal quality were stored by the oximeter every
changed in line with recommendations of the 2 s for at least the first 10 min after birth. We used

Original article
2 s averaging and maximum sensitivity. A member of the distribution was skewed. SpO2 and HR during the first 10 min
research team documented any interventions during resuscita- are illustrated by group using box plots showing the median,
tion including adjustments made to the fractional inspired interquartile range (IQR) and range with outliers. We did not
oxygen (FiO2). In the OX100 group, the FiO2 was not able to be define primary or secondary outcomes a priori, therefore
changed because there was no gas blender available until infants inferential statistics have not been used to compare these
were moved into a transport cot for transfer to the neonatal historical cohorts.
intensive care unit. Infants in the OX21 group were managed
according to the 2006 Royal Women’s Hospital DR protocol
with oxygen titrated according to SpO2 measurements (fig 1). If
infants reached an SpO2 .90%, the FiO2 was reduced in stages of
Table 1 Characteristics of infants in the two groups
,10% to target the SpO2 to 80–90%. Active, spontaneously
breathing infants, of any gestation, were started on continuous OX100 OX21
positive airway pressure (CPAP); infants requiring additional (n = 20) (n = 105)
support were intubated and ventilated. This did not change over
Gestational age (weeks)* 27 (1.6) 27 (1.6)
the time of this study. If free flow oxygen, CPAP or intermittent
Birth weight (g)* 915 (300) 930 (293)
positive pressure ventilation were required, this was given with
Male 13 (65) 67 (64)
a Neopuff (Fisher & Paykel, Auckland, New Zealand) T-piece
Labour started 3 (15) 51 (48)
resuscitation device. In the second time period, a small number Full course of antenatal 18 (90) 87 (82)
of infants were managed using a self-inflating resuscitation steroids
device (Laerdal, Stavanger, Norway). Apgar score at 1 min{ 5 (5–7) 5 (3–7)
After resuscitation, data from the oximeter (HR, SpO2 and Apgar score at 5 min{ 8 (7–9) 8 (6–9)
signal quality) were downloaded to a computer using the Cord pH{{ 7.28 (7.25–7.34) 7.3 (7.25–7.34)
NeO2m program16 (Dr Girvan Malcolm, Royal Prince Alfred Days in oxygen{ 3.5 (2–35.5) 20 (3–44)
Hospital, Sydney, Australia). The data were analysed with Stata Days treated with CPAP{ 9 (4–42) 9 (4–34)
(Intercooled 10). We only analysed measurements where the Days treated with 9 (3–18) 6 (3–12)
ventilation{
signal was considered normal, ie, no alarm messages (low IQ
Died before discharge/ 3 (15) 12 (11)
signal, low perfusion, sensor off, ambient light). transfer
These observational data are presented to illustrate the effects Values are number (%) unless indicated otherwise.
on SpO2 and HR of resuscitation with either 100% oxygen or air *Mean (SD).
with backup 100% oxygen if the SpO2 was ,70% at 5 min. The {Median (interquartile range).
{pH available for eight infants in the OX100 group and 77 infants in
data are presented as numbers and proportions (%) for
the OX21 group.
categorical variables, or means (SD) for normally distributed CPAP, continuous positive airway pressure; OX100, received 100%
continuous variables and median (interquartile range) when the oxygen; OX21, received 21% oxygen.
Figure 1 The Royal Women’s Hospital

guideline for managing administration and
titration of supplemental oxygen in the
delivery room. CPAP, continuous positive
airway pressure; FiO2, fractional inspired
oxygen; NICU, neonatal intensive care
unit; SpO2, oxygen saturation.

Original article
In the OX100 group, verbal consent was obtained from 2 min, over 140 beats/min by 5 min and over 150 beats/min at
parents to monitor their infants in the DR. In the later group, 10 min.
parental consent was not obtained to monitor infants, as
applying an oximeter sensor for monitoring in the DR was the DISCUSSION
standard of care for management of infants at ‘‘high risk’’ for Pulse oximetry is increasingly used during neonatal resuscita-
receiving active resuscitation in our institution. tion17 18 with some centres using it to target a specific SpO2
range.19 Several studies report SpO2 changes in term or near-term
RESULTS infants not requiring resuscitation in the first minutes after
A total of 126 infants were studied. Pulse oximeter data were birth.13 14 20–22 Recent studies used pulse oximeters with algo-
available for 125 infants (sensor failure in one infant in the OX21 rithms that deal with low perfusion and motion artefact, both
group). All 20 infants in the OX100 group received 100% oxygen of which are common in the DR.13 14 20–22 These studies reported
before 1 min of age. In the OX21 group, 97/105 (92%) were an SpO2 of ,60% at 1 min, with many infants taking at least
subsequently treated with supplemental oxygen at median 10 min to achieve >90%. Data from our very preterm infants
(IQR) of 5.05 (4–5.5) min. Eight infants (8%) in the OX21 group initially resuscitated with air and backup 100% oxygen had a
did not receive supplemental oxygen in the DR. Tables 1 and 2 similar course. In our very preterm infants resuscitated with
100% oxygen, the SpO2 rose more quickly.
present the clinical characteristics and DR interventions,
When the infants in our study were resuscitated initially with
respectively. No infants received external cardiac massage.
air, the SpO2 levels were at the lower end of the normal range for
healthy term infants13 and rose into the ‘‘normal’’ range when
Changes in oxygen saturation they were treated with supplemental oxygen at about 5 min. By
Figure 2 shows the changes in SpO2 values for the two groups 6 min, their median SpO2 was 81%. By 7 min, there was little
over the first 10 min. The median SpO2 at 1 min was 60% for difference in the SpO2 between the OX21 group and the OX100
the OX100 group and 55% for the OX21 group. By 2 min, the group. Fewer infants in the OX21 group had an SpO2 >95% in
OX100 group had a median SpO2 of 84%, which continued to rise the first 10 min than in the OX100 group.
steadily to a median of 94% at 5 min and 96% by 10 min. For A systematic review found that air was more effective than
the OX21 group, the median SpO2 fell to 31% at 2 min, then rose 100% oxygen for resuscitation of asphyxiated term infants.23
to a median of 54% at 5 min, followed by a sharp rise to 81% at SpO2 data available for some infants in these trials24–26 show no
6 min, after supplementary 100% oxygen was started, reaching significant difference in SpO2 measurements for infants rando-
a median SpO2 of 91% at 10 min. After 5 min, the median SpO2 mised to receive air or 100% oxygen. However, few very preterm
was very similar in the two groups. In the first 10 min after infants were enrolled in these studies.
birth, 80% and 55% of infants in the OX100 and OX21 groups, Our hospital changed policy to starting resuscitation in air on
respectively, had an SpO2 >95%. the basis of evidence from randomised trials comparing initial
The eight infants not receiving supplemental oxygen were DR resuscitation with 100% oxygen or air plus 100% oxygen as
similar in gestation to infants in both the OX100 and OX21 needed.8 9 26 When we developed our protocol, there were few
groups, with a mean (SD) gestational age of 27.5 (1) weeks; data from randomised trials comparing 100% oxygen with an
however, they were slightly larger with a mean (SD) birth oxygen concentration other than 21% in preterm infants. Two
weight of 1044 (167) g. Six received CPAP and three received studies had randomised infants to either ,100% or .21%
intermittent positive pressure ventilation via a face mask. The oxygen. Lundstrom et al27 randomised 70 infants ,33 weeks’
median SpO2 in these eight infants at 1, 2, 5 and 10 min was gestation to receive air or 80% oxygen in the DR. They
60%, 71%, 87% and 93%, respectively. hypothesised that cerebral blood flow at 2 h of age might be
reduced after a brief period of hyperoxia from 80% oxygen at
birth. They found that the cerebral blood flow was significantly
Changes in HR (p,0.0001) higher in the group treated with air. They also
Figure 3 shows the changes in HR over the first 10 min. The showed, in a subgroup of infants monitored with oximetry, that
median HR in both groups at 1 min was ,100 beats/min: 76 the mean SpO2 was significantly higher at 3, 5 and 7 min in the
beats/min (OX100 group) and 93 beats/min (OX21 group). In 80% oxygen group than the air group. In the air group, 74% did
both groups, the median HR increased to over 100 beats/min by not receive supplemental oxygen. In those who received oxygen,
the maximum was 50%. The reasons for the different oxygen
Table 2 Delivery room interventions requirements from those in our study are that the infants in the
OX100 OX21 study of Lundstrom et al were more mature and clinical
methods were used rather than SpO2 to titrate oxygen in the
(n = 20) (n = 105) air group. We have shown that clinicians’ ability to measure
Nasopharyngeal suction 9 (45) 51 (48) colour28 or HR29 in the DR is weak. Harling et al30 randomised 63
CPAP 16 (80) 72 (69) infants ,31 weeks’ gestation to either 50% or 100% oxygen in
IPPV 14 (70) 80 (76) the DR. They hypothesised that cytokine concentration in
Endotracheal intubation 8 (40) 42 (40) bronchoalveolar lavage fluid 12 h after birth would be highest in
Surfactant administered 2 (10) 5 (5) infants treated with 100% oxygen, but found no significant
Oxygen administered 20 (100) 97 (92) difference. In infants randomised to receive 50% oxygen, one-
Time oxygen started (min 1 (0.86–1.2) 5.05 (4–5.5) third had an FiO2 above 50% during resuscitation. They did not
from birth)* measure SpO2 in the DR.
Each infant may have received several interventions. One approach to selecting the FiO2 to use in the DR is to use
Values are number (%) unless indicated otherwise. SpO2 measurements to adjust the FiO2.31 In the neonatal
*Median (interquartile range).
CPAP, continuous positive airway pressure; IPPV, intermittent intensive care unit, targeting a narrow range for SpO2 and
positive pressure ventilation. avoiding hyperoxia is associated with reduced morbidity in

Original article
Figure 2 SpO2 shown at each minute, for the first 10 min after birth, Figure 3 The heart rate shown at each minute, for the first 10 min after
from the group receiving 100% oxygen (OX100) and the group receiving birth, for infants from the 100% oxygen (OX100) group and the 21% oxyen
21% oxygen (OX21). The box plots show the median, interquartile range, (OX21) group. The box plots show the median, interquartile range, normal
normal range and outliers. Small circles indicate outliers. range and outliers. Small circles indicate outliers.
extremely-low-birthweight infants32 33 without a detrimental Escrig et al36 targeted 85%, Wang et al34 targeted 80–85% at
effect on developmental outcomes.32 It seems logical that a 5 min and 85–90% after 7 min, and Rabi et al37 targeted an SpO2
‘‘targeted oxygen delivery approach’’34 should be applied during range of 85–92%. Each group used slightly different targets
resuscitation. A small observational study35 titrated FiO2 against combined with different resuscitation protocols, which will
targeted SpO2 in 15 infants born at 24–29 weeks. They were have influenced the FiO2 used.
initially resuscitated with 100% oxygen and the FiO2 adjusted to Hyperoxia is common during resuscitation of preterm
maintain the SpO2 between 80% and 92%. The FiO2 was reduced infants. Tracy et al38 showed that, of 26 ventilated preterm
from 100% to ,40%. infants of mean gestation 28 weeks (range 23–34), 38% were
There are now three controlled studies on very preterm hyperoxic (PaO2 .100 mm Hg) 15 min after birth. Both Wang
infants where the FiO2 has been titrated to the SpO2 after birth. et al34 and our own study have shown that more of the infants
Escrig et al36 randomised 28 infants ,29 weeks’ gestation to who started in 100% oxygen had an SpO2 of .95% than those
receive 30% or 90% oxygen. The FiO2 was adjusted to achieve an who started in air. Rabi et al37 reported that infants
SpO2 of 85%. By 5 min, the FiO2 was just above 50%, with no resuscitated in 100% oxygen spent 49% of the time above
significant difference between the groups. Wang et al34 the SpO2 range of 85–92%. Therefore, in very preterm infants,
randomised infants ,32 weeks’ gestation to start resuscitation hyperoxia appears to be a problem after starting resuscitation
with 100% oxygen or air. In the 100% oxygen group, the FiO2 with a high FiO2 and indicates that starting with a lower FiO2
was weaned if the SpO2 was .95% at 5 min. In the air group, may be preferable.
the FiO2 was increased in 25% steps if the SpO2 was ,70% at Our results contribute to a growing body of evidence that it is
3 min or 85% at 5 min, or to 100% if the HR was ,100 beats/ possible to adjust the FiO2 to keep SpO2 measurements within a
min for 2 min or ,60 beats/min for 30 s at any time. All infants targeted range during resuscitation. 34 36 37 Our study shows that
in the air group received oxygen from ,3 min. Infants in the most of our very preterm infants received supplemental oxygen
100% oxygen group had significantly higher FiO2 from 1 to when the initial resuscitation was with air and their SpO2 was
7 min, but from 8 to 20 min it was similar in the two groups. ,70% at 5 min or ,90% at 10 min. However, with different
From 2 to 10 min, the SpO2 was higher in the group initially SpO2 targets, the use of supplemental oxygen would be
resuscitated with 100% oxygen. different. We have shown that, using oximetry, it is possible
In the only randomised study to mask clinicians to the SpO2, to titrate the FiO2 to the SpO2, with the SpO2 rising at a similar
Rabi et al37 randomised 106 infants ,33 weeks’ gestation to rate to that in term infants not receiving assistance in the
three groups. One received 100% oxygen throughout resuscita- delivery room.13 14 21
tion, the second received an initial concentration of 100%, When the initial resuscitation of very preterm infants has
which could than be changed, and the third group started with been started with .90% oxygen, it has generally been possible
air. In the last two groups, the FiO2 was changed by 20% every to reduce the FiO2 in response to the SpO2 to , 50% at 5 min
15 s until the SpO2 was between 85% and 92%. The mean time and 35% at 10 min. In comparison, when the initial resuscita-
that each group spent in the SpO2 target range was 11%, 21% tion gas was air or a low FiO2, the FiO2 at 5 min was similar to
and 16%, respectively (p,0.01). At the end of resuscitation, the when a high FiO2 was used. Therefore it now appears that,
FiO2 was similar in the two targeted groups. whatever the FiO2 used to initiate resuscitation, if the SpO2 is
The safe SpO2 range for very preterm infants during monitored within a few minutes an appropriate FiO2 is
resuscitation is undefined. We targeted an SpO2 of 80–90%. achieved.

Original article
HR is the most important indicator of an infant’s response to 12. Saugstad OD. Oxidative stress in the newborn: a 30-year perspective. Biol Neonate
2005;88:228–36.
resuscitation.18 31 Using pulse oximetry in the DR provides 13. Kamlin CO, O’Donnell CPF, Davis PG, et al. Oxygen saturation in healthy infants
clinicians with a continuous display of HR without having to immediately after birth. J Pediatr 2006;148:585–9.
interrupt resuscitation to listen to the HR intermittently. 14. Rabi Y, Yee W, Chen SY, et al. Oxygen saturation trends immediately after birth.
J Pediatr 2006;148:590–4.
Importantly we found that the HR increased after birth at a
15. O’Donnell CPF, Kamlin CO, Davis PG, et al. Feasibility of and delay in obtaining pulse
similar rate in this group of very preterm infants, regardless of oximetry during neonatal resuscitation. J Pediatr 2005;147:698–9.
whether the infant was resuscitated with air or 100% oxygen. 16. Malcolm G. Neonatal oxygen saturation download and analysis - user manual. 2007.
Our finding and those of other researchers34 36 have shown that, http://www.cs.nsw.gov.au/rpa/ (accessed 23 Oct 2008).
17. O’Donnell CPF, Davis PG, Morley CJ. Use of supplementary equipment for
even with a low SpO2 during the first few minutes after birth, resuscitation of newborn infants at tertiary perinatal centres in Australia and New
HR was similar to that achieved by healthy newborn term Zealand. Acta Paediatr 2005;94:1261–5.
infants not receiving assistance.13 14 21 18. Leone TA, Rich W, Finer NN. A survey of delivery room resuscitation practices in the
United States. Pediatrics 2006;117:164–75.
Two important questions cannot be answered by our study
19. Finer NN. Rich WD. Neonatal resuscitation: raising the bar. Curr Opin Pediatr
but could be addressed in future trials. Does pulse oximetry in 2004;16:157–62.
the DR improve outcomes for preterm infants? If pulse 20. Toth B, Becker A, Seelbach-Gobel B. Oxygen saturation in healthy newborn infants
oximetry is effective, what is the safe SpO2 range for preterm immediately after birth measured by pulse oximetry. Arch Gynecol Obstet
2002;266:105–7.
infants in the first minutes after birth? 21. Altuncu E, Ozek E, Bilgen H, et al. Percentiles of oxygen saturations in healthy term
newborns in the first minutes of life. Eur J Pediatr 2008 ;167 :687–8.
Acknowledgements: JAD and COFK are recipients of a RWH Postgraduate 22. Mariani G, Dik PB, Ezquer A, et al. Pre-ductal and post-ductal O2 saturation in
Scholarship. ATP is the recipient of a Ter Meulen Fund grant for working visits, Royal healthy term neonates after birth. J Pediatr 2007;150:418–21.
Netherlands Academy of Arts and Sciences, The Netherlands, and PGD is the recipient 23. Davis PG, Tan A, O’Donnell CPF, et al. Resuscitation of newborn infants with 100%
of a NHMRC Practitioner Fellowship. The work was supported by Australian National oxygen or air: a systematic review and meta-analysis. Lancet 2004;364:1329–33.
Health and Medical Research Council Program Grant No 384100. We thank Dr Girvan 24. Ramji S, Ahuja S, Thirupuram S, et al. Resuscitation of asphyxic newborn infants
Malcolm, Royal Prince Alfred Hospital, Sydney for his assistance with the NeO2M with room air or 100% oxygen. Pediatr Res 1993;34:809–12.
program. 25. Rao R,.Ramji S. Pulse oximetry in asphyxiated newborns in the delivery room. Indian
Pediatr 2001;38:762–6.
Competing interests: None. 26. Saugstad OD, Rootwelt T, Aalen O. Resuscitation of asphyxiated newborn infants
Patient consent: Parental consent obtained. with room air or oxygen: an international controlled trial: the Resair 2 study. Pediatrics
1998;102:e1.
27. Lundstrom KE, Pryds O, Greisen G. Oxygen at birth and prolonged cerebral
REFERENCES vasoconstriction in preterm infants. Arch Dis Child Fetal Neonatal Ed 1995;73:F81–6.
1. O’Donnell CPF, Gibson AT, Davis PG. Pinching, electrocution, ravens’ beaks, and 28. O’Donnell CPF, Kamlin CO, Davis PG, et al. Clinical assessment of infant colour at
positive pressure ventilation: a brief history of neonatal resuscitation. Arch Dis Child delivery. Arch Dis Child Fetal Neonatal Ed 2007;92:F465–7.
Fetal Neonatal Ed 2006;91:F369–73. 29. Kamlin COF, O’Donnell CPF, Everest N, et al. Accuracy of clinical assessment of
2. Saugstad OD. Optimal oxygenation at birth and in the neonatal period. Neonatology infant heart rate in the delivery room. Resuscitation 2006;71:319–21.
2007;91:319–22. 30. Harling AE, Beresford MW, Vince GS, et al. Does the use of 50% oxygen at birth in
3. Kondo M, Itoh S, Isobe K, et al. Chemiluminescence because of the production of preterm infants reduce lung injury? Arch Dis Child Fetal Neonatal Ed 2005;90:F401–5.
reactive oxygen species in the lungs of newborn piglets during resuscitation periods 31. American Heart Association. 2005 American Heart Association (AHA) Guidelines
after asphyxiation load. Pediatr Res 2000;47:524–7. for cardiopulmonary resuscitation (CPR) and Emergency cardiovascular care (ECC) of
4. Kutzsche S, Ilves P, Kirkeby OJ, et al. Hydrogen peroxide production in leukocytes pediatric and neonatal patients: neonatal resuscitation guidelines. Pediatrics
during cerebral hypoxia and reoxygenation with 100% or 21% oxygen in newborn 2006;117:1029–38.
piglets. Pediatr Res 2001;49:834–42. 32. Deulofeut R, Critz A, Adams-Chapman I, et al. Avoiding hyperoxia in infants
5. Saugstad OD, Ramji S, Vento M. Oxygen for newborn resuscitation: how much is (1250 g is associated with improved short- and long-term outcomes. J Perinatol
enough? Pediatrics 2006;118:789–92. 2006;26:700–5.
6. Vento M, Asensi M, Sastre J, et al. Resuscitation with room air instead of 100% 33. Chow LC, Wright KW, Sola A. Can changes in clinical practice decrease the
oxygen prevents oxidative stress in moderately asphyxiated term neonates. Pediatrics incidence of severe retinopathy of prematurity in very low birth weight infants?
2001;107:642–7. Pediatrics 2003;111:339–45.
7. Vento M, Asensi M, Sastre J, et al. Hyperoxemia caused by resuscitation with pure 34. Wang CL, Anderson C, Leone TA, et al. Resuscitation of preterm neonates by using
oxygen may alter intracellular redox status by increasing oxidized glutathione in room air or 100% oxygen. Pediatrics 2008;121:1083–9.
asphyxiated newly born infants. Semin.Perinatol 2002;26:406–10. 35. Kopotic RJ, Lindner W. Assessing high-risk infants in the delivery room with pulse
8. Vento M, Asensi M, Sastre J, et al. Oxidative stress in asphyxiated term infants oximetry. Anesth.Analg 2002;94:S31–6.
resuscitated with 100% oxygen. J Pediatr 2003;142:240–6. 36. Escrig R, Arruza L, Izquierdo I, et al. Achievement of targeted saturation values in
9. Vento M, Sastre J, Asensi MA, et al. Room-air resuscitation causes less damage to extremely low gestational age neonates, resuscitated with low or high oxygen
heart and kidney than 100% oxygen. Am J Respir.Crit Care Med 2005;172:1393–8. concentrations: a prospective randomized trial. Pediatrics 2008;121:875–81.
10. Australian Resuscitation Council. Guideline 13.1 Introduction to resuscitation of 37. Rabi Y, Nettel-Aguirre A, Singhal N. Room air versus oxygen administration during
the newborn infant. 2006. http://www.resus.org.au/ (accessed 17 Oct 2008). resuscitation of preterm infants (ROAR Study). Meeting of Pediatric Academic
11. Canadian Paediatric Society. Recommendations for specific treatment Societies, Hawaii, 2008. E-PAS2008: 5127.5
modifications in the Canadian context: addendum to the 2006 NRP Provider Textbook. 38. Tracy M, Downe L, Holberton J. How safe is intermittent positive pressure
Canadian Paediatric Society 2006. 9-9-0007. http://www.cps.ca/English/ProEdu/NRP/ ventilation in preterm babies ventilated from delivery to newborn intensive care unit?
addendum.pdf (accessed 23 Oct 2008). Arch Dis Child Fetal Neonatal Ed 2004;89:F84–7.

No change in developmental outcome with

incubator covers and nesting for very preterm
infants in a randomised controlled trial
C M Maguire, F J Walther, P H T van Zwieten, S Le Cessie, J M Wit, S Veen and
On behalf of the Leiden Developmental Care Project
14 Aug 2008;
doi:10.1136/adc.2008.141002

These include:

Notes


Original article
No change in developmental outcome with incubator

covers and nesting for very preterm infants in a
randomised controlled trial
C M Maguire,1 F J Walther,1 P H T van Zwieten,2 S Le Cessie,3 J M Wit,1 S Veen,1 On
behalf of the Leiden Developmental Care Project
1
Department of Pediatrics, ABSTRACT
Subdivision of Neonatology, Objective: To investigate in a randomised controlled trial What is already known on this topic
Leiden University Medical
Center, Leiden, The Netherlands;
the effect of basic elements of developmental care
2
Department of Pediatrics, (incubator covers and positioning aids) on growth and c Previous trials are based on the comprehensive
Subdivision of Neonatology, neurodevelopment in infants born at , 32 weeks. Newborn Individualised Developmental Care and
Haga Hospital, Juliana Children’s Method: Infants were randomised within 48 h of birth to Assessment Program (NIDCAP).
Hospital, The Hague, The
Netherlands; 3 Department of
a developmental care (DC) or standard care (C) group. c Outcomes of NIDCAP trials are based on small
Medical Statistics, Leiden Outcome measures at 1 and 2 years corrected age were sample sizes, and results are conflicting.
University Medical Center, growth, standardised neurological examinations, and c A Cochrane Review of developmental care
The Netherlands mental (MDI) and psychomotor (PDI) development (Dutch stated the need for larger trials and more follow-
Correspondence to:
version of the Bayley Scales of Infant Development II). up, and investigation of the effects of different
Dr S Veen, Department of Results: 192 infants were recruited (DC = 98; C = 94). aspects of developmental care was emphasised.
Pediatrics, J-6-S, Leiden Thirteen infants (DC = 7, C = 6) were excluded because
University Medical Center, PO they were admitted for ,5 days or died within the first 5
Box 9600, 2300 RC Leiden,
The Netherlands; s.veen@lumc.nl days. In total, 179 infants met the inclusion criteria. In-
hospital mortality was 12/91 (13.2%) in the DC group and
What this study adds
Accepted 27 July 2008 8/88 (9.1%) in the C group. Assessments were carried
Published Online First out on 147 children (DC = 74, C = 73) at 1 year and 142 c The effects of basic developmental care were
14 August 2008 children (DC = 72, C = 70) at 2 years. No significant investigated, thus answering the need for trials
difference in growth, neurological outcomes or MDI was of different aspects of developmental care as
found. A positive trend in PDI at 1 year (p = 0.05) did not stated in the Cochrane Review.
continue once the children reached 2 years. There was no c This was a large randomised controlled trial of
difference found when neurological and developmental basic development care in very preterm infants
scores were combined. with developmental, neurological and growth
Conclusions: Basic developmental care has no positive outcomes to 2 years corrected age.
effect on neurological and mental development or growth c It was shown that a less intensive, cost-saving
at 1 and 2 years of age in infants born at ,32 weeks. A form of developmental care has an effect on
positive effect on psychomotor development at 1 year did psychomotor development at 1 year of age but
not continue at 2 years of age. no significant effect on neurodevelopment of
preterm infants at 2 years of age.
Trial registration number: ISRCTN84995192.
the effects of different aspects of developmental

The care and survival rate of infants born preterm care was emphasised.18
has in recent years continued to improve.1–4 Even as The aim of this randomised controlled trial was
survival rates are improving, the risk of develop- to explore the effectiveness of basic developmental
mental disabilities remains high and increases as care on mental and psychomotor development,
the gestational age at birth decreases.1 5–7 Since the neurological outcome and growth at 1 and 2 years
1980s, developmental care programmes have been corrected age (CA) of preterm infants born at
created to support infant development in the ,32 weeks’ gestational age. We hypothesised that,
neonatal intensive care unit (NICU) while at the by reducing stress and promoting physiological
same time providing the necessary medical and stability through the use of incubator covers and
nursing interventions. Many are based on the nesting, the stability provided to the infants during
Newborn Individualised Developmental Care and their NICU stay would positively affect their later
Assessment Program (NIDCAP), an individual growth and development.
approach in which care giving is based on the
infant’s behaviour.8 9 The first studies of the
effectiveness of NIDCAP developmental care SUBJECTS AND METHODS
showed promising results.10–14 Follow-up studies Subjects
published to date up to preschool age have been The study was carried out from April 2000 to June
scarce, and the results are conflicting.15–17 In a 2004 at a tertiary NICU at two locations in the
recent Cochrane Review of developmental care, the Netherlands: Leiden University Medical Center in
need for larger trials, more follow-up, and study of Leiden and Juliana Children’s Hospital in The

Original article
Hague. Inclusion criteria were: infants born with a gestational intensive care days, short-term growth or neuromotor develop-
age ,32 (31+6) weeks. Exclusion criteria included: infants with mental outcome.19
major congenital anomalies, infants needing major surgery and The children were assessed at 1 and 2 years CA by psychology
infants of drug-addicted mothers. After parental informed interns supervised by a clinical psychologist, who were blinded
consent had been obtained by the resident or staff member on to whether the child was in the DC or C group. All ages
call, infants were randomised within 48 h of birth to the mentioned hereafter are corrected for prematurity. Mental and
developmental care (DC) group or the control standard care (C) psychomotor development was assessed using the Dutch
group using sealed envelopes made in groups of six using a version of the Bayley Scales of Infant Development II (BSID-
computer-generated randomisation allocation. Infants in both II).20 21 The mean score of the mental developmental index
groups who were admitted for ,5 days were excluded from (MDI) and the psychomotor developmental index (PDI) is 100,
follow-up because the duration of the basic developmental care with 1 SD of 15 points. An MDI or PDI > 85 (> 21 SD) is
intervention was hypothesised not to be long enough to obtain considered normal, an MDI or PDI between 70 and 84 (22 to
an effect. A power analysis performed before the study showed 21 SD) is considered mildly delayed, and index scores ( 69 (,
that a sample size of 140 infants was needed to show a 22 SD) severely delayed. The Dutch norms, which became
significant difference (p,0.05) with a power of 80%, based on available during our research, were used.
the expected difference of half a standard deviation (7.5) on the A standardised neurological examination at 1 year, as
developmental test scores at 1 and 2 years CA. described by Touwen,22 23 and at 2 years, as described by
Hempel,24 was administered by neonatologists experienced in
developmental assessments and blinded to the group assign-
Methods
ment of the child. They were classified as definitely abnormal
The intervention included reducing light and sound through the
(DA) when there was definite neurological dysfunction such as
use of standardised incubator covers, and supporting motor
cerebral palsy, mildly abnormal (MA) in the presence of mild
development and physiological stability by positioning the
deviations in muscle tone regulation, reflexes, fine or gross
infant in ways that encourage flexion and containment through
motor performance, or cranial nerve function, or normal (N). To
the use of standardised nests and positioning aids. Infants in the
obtain a single outcome measure, neurological outcome, PDI
C group received standard care, which at that time consisted of
and MDI were combined. When at least one of these three
no covers or nesting. The ethics committees of both locations
outcome measures was DA, children were considered to be DA,
approved the study. The short-term findings to term age
and when at least one outcome measure was MA, children were
showed no difference in number of days of respiratory support, considered to be MA.
Weight was measured on a paediatric digital scale, length was
measured from crown to heel with the child lying supine on a
standard measurement board, and head circumference was
measured around the largest area of the head (occipital–frontal
circumference) using a non-stretch tape measure.
Data were analysed using SPSS V12.0 for Windows. Outcome
parameters and infant and parental characteristics were
compared with the t test, Mann–Whitney test or x2 test (for
trend) where appropriate. p,0.05 was considered significant.
Linear regression was used to evaluate the influence of the
duration of the intervention on 1 and 2 year outcomes by
testing if there was an interaction effect between the interven-
tion duration and the treatment groups.
RESULTS
In total, 192 infants were recruited for the study; 98 in the DC
group and 94 in the C group. Thirteen infants (DC = 7, C = 6)
were excluded because they were admitted for less than 5 days
or died within the first 5 days. One of the six infants in the C
group was taken out of the study on day 3 at the parents’
request. A total of 179 infants met the inclusion criteria: 12/91
(13.2%) in the DC group and 8/88 (9.1%) in the C group died
during hospitalisation, with the main cause of death being
cerebral or pulmonary complications. Two infants in each group
died from necrotising enterocolitis. There was no significant
difference in the in-hospital mortality between the DC and C
group (p = 0.40). This left 159 infants (DC = 79, C = 80) for
follow-up. At 1 year, four infants were lost to follow-up in the
DC group and seven infants in the C group because they were
transferred to hospitals out of the health region or parents did
Figure 1 Infants in the developmental care study. DC, developmental not come back for follow-up. One infant in the DC group died
care; C, control; CA, corrected age. between term age and 1 year. Between 1 and 2 years, two

Original article
Table 1 Medical background variables of children seen at 1 and 2 year follow-up

1 year follow-up 2 year follow-up
Birth characteristic DC (n = 74) C (n = 73) DC (n = 72) C (n = 70)
Gestational age (weeks)* 29.5 (1.6) 29.1 (1.9) 29.5 (1.6) 29.1 (1.9)
Range 25.9–31.9 25.0–31.9 25.9–31.9 25.0–31.9
Birth weight (g)* 1248.4 (338.1) 1238.5 (337.2) 1266.3 (329.6) 1236.6 (338.5)
Range 585–2155 640–2080 585–2155 640–2080
Male gender 39/74 (52.7) 46/73 (63.0) 38/72 (52.8) 44/70 (62.9)
SGA
SGA P,10 and P>3 8/74 (10.8) 6/73 (8.2) 8/72 (11.1) 5/70 (7.1)
SGA P,3 6/74 (8.1) 4/73 (5.5) 4/72 (5.6) 4/70 (5.7)
Inborn{ 46/74 (62.2) 46/72 (63.9) 45/72 (62.5) 44/70 (62.9)
Apgar scores at 5 min n = 74 n = 72{ n = 72 n = 69{
Median (range) 9.0 (2–10) 8.0 (5–10) 9.0 (2–10) 8.0 (5–10)
CRIB score* 3.2 (2.9) 3.7 (2.9) 3.0 (2.7) 3.8 (3.0)
Range 0–13 0–11 0–10 0–11
Data are number (%), unless otherwise indicated. Comparisons were performed using x2 test or t tests where appropriate.
*Mean (SD).
{Infants born in the participating tertiary neonatal centre.
{Correct number is shown in table if there are missing values.
C, control group; CRIB, clinical risk index for babies26; DC, developmental care group; SGA, small for gestational age (P,10, less
than the 10th centile, etc).
children in the DC and three children in the C group were lost (DC = 4, C = 4) tested were 26–27 months at 2 years, but their
to follow-up because of parents moving or not wanting to index scores were based on the norms for that age, so we
continue with the follow-up. The baseline data from the infants included them in the analysis. There was no difference in the
who were lost to follow-up were comparable to those from the mean age of all children assessed at the 1 and 2 year follow-up.
infants who were assessed (data not shown). There were 147 Two children had no developmental test because of illness or
children (DC = 74/79 (93.7%), C = 73/80 (91.3%)) at 1 year and because they were uncooperative. At 1 year, the children in the
142 children (DC = 72/79 (91.1%), C = 70/80 (87.5%)) at 2 years DC group showed a trend of a higher PDI than those in the C
seen at the follow-up clinic (fig 1). No significant differences in group (p = 0.05) but no significant difference (p = 0.56) in their
infant and parental characteristics were found (tables 1 and 2). MDI. At 2 years, this difference was no longer evident, as the
MDI and PDI scores were comparable (table 3).
Developmental outcomes
Our primary outcome in which the power analysis was Neurological outcomes
calculated was developmental outcome at 1 and 2 years. At 1 There were 147 (DC = 74, C = 73) children who were assessed
year, 145 children (DC = 73, C = 72) of the 147 children seen at with a neurological examination at 1 year, and 140 (DC = 71,
follow-up were tested with the Bayley Scales-II–NL, and at 2 C = 69) children at 2 years. Two children (DC = 1, C = 1) were
years of age 140 (DC = 70, C = 70) of the 142 children were not tested at 2 years because they were uncooperative. No
tested. Three children (DC = 1, C = 2) were 13–14 months old differences in neuromotor development at 1 year and 2 years of
at the 1 year developmental follow-up, and eight children age were found between the DC and C group (table 4).
Table 2 Parental background variables

1 year follow-up 2 year follow-up
DC C DC C
Maternal age (years) n = 74 n = 73 n = 72 n = 73

mean (SD) 31.3 (5.1) 31.4 (4.9) 32.5 (5.1) 31.4 (4.9)
Paternal age (years) n = 70 n = 69 n = 67 n = 69
mean (SD) 34.3 (5.3) 35.0 (5.7) 35.0 (5.2) 35.0 (5.7)
Mother caucasian 48/74 (64.9) 53/73 (72.6) 48/74 (64.9) 53/73 (72.6)
Father caucasian 52/74 (70.3) 56/73 (76.7) 52/74 (70.3) 56/73 (76.7)
Education level of mother*
Low 34/74 (46.0) 23/72 (32.0) 32/71 (45.1) 23/72 (32.0)
Intermediate 24/74 (32.4) 33/72 (45.8) 23/71 (32.4) 33/72 (45.8)
High 16/74 (21.6) 16/72 (22.2) 16/71 (22.5) 16/72 (22.2)
Education level of father*
Low 26/74 (35.2) 19/71 (26.8) 26/71 (36.6) 19/71 (26.8)
Intermediate 30/74 (40.5) 29/71 (40.8) 28/71 (39.4) 29/71 (40.8)
High 18/74 (24.3) 23/71 (32.4) 17/71 (23.9) 23/71 (32.4)
Data are number (%) unless otherwise indicated. Comparisons were performed using x2 test (for linear trend) or t tests where
appropriate.
*Low = vocational training, intermediate = high school, high = college/university.
C, control group; DC, developmental care group.

Original article
Table 3 Mental and psychomotor development at 1 and 2 years corrected age

1 year 2 years
DC (n = 73) C (n = 72) p Value DC (n = 70) C (n = 70) p Value
Age at test 12.14 (0.34) 12.14 (0.40) 0.99 24.3 (0.68) 24.1 (0.47) 0.12
(months)*
Range 11.2–13.2 11.4–12.1 23.2–26.4 23.5–27.4
MDI* 102.3 (15.1) 101.2 (15.7) 0.66 100.9 (14.9) 102.3 (16.2) 0.58
Range (57–138) (55–132) (55–130) (56–132)
PDI* 99.2 (17.0) 93.7 (16.1) 0.05 96.0 (14.6) 92.3 (17.0) 0.18
Range (55–135) (55–124) (55–121) (55–145)
MDI classification
scores{
>85 64 (87.7) 62 (86.1) 0.69 61 (87.1) 60 (85.7) 1.00
70–84 7 (9.6) 7 (9.7) 7 (10.0) 9 (12.9)
(69 2 (2.7) 3 (4.2) 2 (2.9) 1 (1.4)
PDI classification
scores{
>85 62 (84.9) 56 (77.8) 0.27 54 (77.1) 48 (68.6) 0.20
70–84 6 (8.2) 8 (11.1) 13 (18.6) 16 (22.9)
(69 5 (6.8) 8 (11.1) 3 (4.3) 6 (8.6)
Data are number (%) unless otherwise indicated. Comparisons were performed using the x2 test (for linear trend) or t tests where
appropriate (p,0.05 is significant).
*Mean (SD).
{> 85 = normal or above normal, 70–84 = mildly delayed, ( 69 = significantly delayed.
C, control group; DC, developmental care group; MDI, mental developmental index; PDI, psychomotor developmental index.
Total outcome score Growth

When we combined the developmental and neurological At 1 year, one child from the C group was not measured, and at
outcome measures into one score, the percentage of children 2 years one child from the DC group and one child from the C
in the C group who were definitely delayed at both 1 and 2 group were not measured. No significant differences in growth
years was much higher than in the DC group (1 year: (weight in grams, height and head circumference in centimetres)
DC = 12.2%, C = 23.3%; 2 years: DC = 5.6%, C = 18.3%). were found between the DC and C group at 1 or 2 years. When
However, the difference did not reach the level of significance we calculated the standard deviation scores (SDS) using the
(table 4). Dutch growth charts, we found no differences in weight and
We then carried out a linear regression analysis to see if the head circumference between the groups, and the DC group
number of days on which infants received the DC intervention showed a significantly (p = 0.04) better SDS for length than the
influenced the neurological outcomes at 1 and 2 years by testing C group. There were, however, more infants in the C group
if there was an interaction effect between the intervention (11.3%) than in the DC group (4.4%) who required postnatal
duration and the two treatment groups. No significant effect on corticosteroids (p = 0.08), the usual dosage being 0.20 mg/kg/
the neurological outcome at 1 year (p = 0.79) or 2 years day in two doses with tapering of the dosage over a period of
(p = 0.67) or on the combined neurological and developmental 16 days. As this may influence growth, we corrected for use of
scores at 1 year (p = 0.86) and 2 years (p = 0.60) was found. postnatal steroids and then found no significant difference in
Table 4 Neurological outcomes and combined score of neurological outcomes, MDI and PDI at 1 and 2
years corrected age
1 year 2 years
DC C p Value DC C p Value
Neurological n = 74 n = 73 n = 71 n = 69
outcome
N 56 (75.7) 52 (71.2) 0.25 50 (70.4) 46 (66.7) 0.18
MA 13 (17.5) 10 (13.7) 17 (24.0) 11 (15.9)
DA 5 (6.8) 11 (15.1) 4 (5.6) 12 (17.4)
Combined n = 74 n = 73 n = 72* n = 71{
neurological score
(MDI and PDI)
N 48 (64.8) 45 (61.6) 0.26 38 (52.8) 37 (52.1) 0.25
MA 17 (23.0) 11 (15.1) 30 (41.7) 21 (29.6)
DA 9 (12.2) 17 (23.3) 4 (5.6) 13 (18.3)
Neurological examination at 1 year was as described by Touwen22 23 and that at 2 years as described by Hempel.24 Data are
number (%). Comparisons were performed using x2 test (for linear trend) where appropriate.
*One child’s combined score was derived from the PDI and MDI.
{The combined scores of two children were derived from the PDI and MDI.
C, control group; DA, definitely abnormal (MDI/PDI score (69); DC, developmental care group; MDI, mental developmental index;
N, normal (MDI/PDI score >85); MA, mildly abnormal (MDI/PDI score 70–84); PDI, psychomotor developmental index.

Original article
Table 5 Growth outcomes at 1 and 2 years corrected age

1 year 2 years
Growth outcome DC C p Value DC* C p Value p Value{
Weight (kg) n = 74 n = 72 n = 72 n = 69
Mean 9.31 (1.38) 9.11 (1.28) 0.37 11.9 (1.5) 11.5 (1.4) 0.10
SDS 20.72 (1.27) 20.94 (1.28) 0.31 20.69 (1.12) 21.03 (1.1) 0.08 0.18
Head circumference n = 72* n = 71 n = 71 n = 68
(cm)
Mean 46.4 (1.7) 46.2 (1.7) 0.42 48.6 (1.7) 48.2 (1.7) 0.14
SDS 20.15 (1.10) 20.38 (1.11) 0.21 20.06 (1.03) 20.41 (1.09) 0.06 0.10
Length (cm) n = 74 n = 70 n = 72 n = 69
Mean 74.7 (3.7) 74.1 (3.6) 0.36 87.3 (3.6) 86.0 (4.0) 0.06
SDS 20.54 (1.27) 20.77 (1.31) 0.29 20.36 (1.06) 20.75 (1.24) 0.04 0.10
Data are mean (SD). Comparisons were performed using t tests (p,0.05 was considered significant).
*Correct number is shown in table if there are missing values.
{After correction for postnatal steroid use.
C, control group; DC, developmental care group; SDS, standard deviation scores.27
length SDS between the two groups (p = 0.10) as shown in to 2 years of age. The infants were randomised in an appropriate
table 5. No differences were found in the incidence of manner; however, there could be no blinding of the interven-
bronchopulmonary dysplasia (oxygen dependent at .36 weeks’ tion, as the infants in the DC group had incubator covers and
gestational age) or oxygen requirement at .28 days of life nesting. This did make it easier to ensure a strict control group
between the two groups.19 in which control infants were not provided with any nesting or
incubator covers, as this was the standard method of care when
this trial began and so was easy to maintain during the study
DISCUSSION
period. The percentage of infants lost to follow-up was low, the
This randomised controlled trial showed that a less intensive,
assessors were blinded to the treatment group, and neurological
cost-saving form of developmental care had no positive effect on
outcomes were obtained using a standardised neurological
neurological and mental development or growth at 1 and 2 years
examination.
in infants born at ,32 weeks. A positive effect on psychomotor
We hypothesised that, by reducing stress and promoting
development at 1 year did not continue at 2 years of age.
physiological stability through the use of incubator covers and
Some differences were seen between the neuromotor and
nesting, the stability provided to the infants during their NICU
developmental scores, which may be explained by the fact that
hospitalisation would positively affect their later growth and
the Touwen and Hempel techniques measure minor qualitative
development. However, this was not the case. For future
neuromotor dysfunction, whereas the BSID-II PDI measures
research it is important to note that this commonly used form
motor skills, identifies motor delays, and gives a quantitative
of developmental care showed no long-lasting positive effects up
score. When we combined the scores into a single ‘‘mildly
to 2 years of age. Perhaps a more intensive, individualised
abnormal’’ or ‘‘definitely abnormal’’ score in order to obtain a
developmental care programme such, as NIDCAP, for a longer
clearer picture of the outcome, the difference between the
duration would show improved outcomes.
groups remained insignificant.
We also looked at the number of days in which infants had Acknowledgements: We thank Sylvia M van der Pal, PhD, Monique de Haan, MD,
received developmental care when hospitalised to see if that Monique Rijken, MD, PhD, Shirley Martens, MD, Jan Feenstra, Clinical Psychologist,
positively influenced neurological and developmental outcomes Tanja Kooij, BA, Psychological Assistant, Annelijn Kruger, MSc and the psychology
at 1 and 2 years, but found no interaction effect. In addition, interns, Department of Pediatrics, Leiden University Medical Center, for their
contribution to this research project.
when growth SDS were corrected for postnatal steroid use, we
found no difference in growth outcomes. Funding: This study was funded by the ZONMW (grant 2100.0072) and the Health
To date, there has been no large randomised controlled trial Care Efficiency Research Fund LUMC.
examining growth and neurodevelopmental outcome of a basic Competing interests: None.
developmental care programme. Therefore comparison with Ethics approval: Obtained.
other studies is not possible. Most studies have examined the Patient consent: Parental consent obtained.
more intensive individualised NIDCAP developmental care
programme and had smaller sample sizes and mixed results.15–
18
Most outcomes of developmental care studies have focused on
short-term morbidity and growth or neurodevelopment up to REFERENCES
1. Hack M, Fanaroff AA. Outcomes of children of extremely low birthweight and
9–12 months,11 15 17 with one study that followed infants’ gestational age in the 1990s. Semin Neonatol 2000;5:89–106.
development to 3 years showing no significant difference in 2. Stoelhorst GM, Rijken M, Martens SE, et al. Changes in neonatology: comparison of
development between the two groups16 and at preschool age a two cohorts of very preterm infants (gestational age ,32 weeks): the Project On
possible positive effect on behaviour. However, these outcomes Preterm and Small for Gestational Age Infants 1983 and the Leiden Follow-Up Project
on Prematurity 1996–1997. Pediatrics 2005;115:396–405.
should be interpreted with caution because of the small sample 3. Marlow N. Neurocognitive outcome after very preterm birth. Arch Dis Child Fetal
size.25 No studies have been reported in the Cochrane meta- Neonatal Ed 2004;89:F224–8.
analysis examining the effect of basic developmental care 4. Luke B, Brown MB. The changing risk of infant mortality by gestation, plurality, and
programmes such as ours on neurodevelopment.18 race: 1989–1991 versus 1999–2001. Pediatrics 2006;118:2488–97.
5. Stoelhorst GM, Rijken M, Martens SE, et al. Developmental outcome at 18 and
We have tried with this study to answer some of the 24 months of age in very preterm children: a cohort study from 1996 to 1997. Early
questions posed concerning developmental care and follow-up Hum Dev 2003;72:83–95.

Original article
6. Rijken M, Stoelhorst GM, Martens SE, et al. Mortality and neurologic, mental, and 16. Kleberg A, Westrup B, Stjernqvist K. Developmental outcome, child behaviour and
psychomotor development at 2 years in infants born less than 27 weeks’ gestation: mother–child interaction at 3 years of age following Newborn Individualized
the Leiden follow-up project on prematurity. Pediatrics 2003;112:351–8. Developmental Care and Intervention Program (NIDCAP) intervention. Early Hum Dev
7. Bhutta AT, Cleves MA, Casey PH, et al. Cognitive and behavioral outcomes of 2000;60:123–35.
school-aged children who were born preterm: a meta-analysis. JAMA 17. Kleberg A, Westrup B, Stjernqvist K, et al. Indications of improved cognitive
2002;288:728–37. development at one year of age among infants born very prematurely who received
8. Als H, Gibes R. Newborn Individualized Developental Care and Asessment Program care based on the Newborn Individualized Developmental Care and Assessment
(NIDCAP) training guide. Boston: Children’s Hospital, 1990. Program (NIDCAP). Early Hum Dev 2002;68:83–91.
9. Als H. A synactive model of neonatal behavioral organization: framework for the 18. Symington A, Pinelli J. Developmental care for promoting development and preventing
assessment of neurobehavioral development in the premature infant and for support morbidity in preterm infants. Cochrane Database Syst Rev 2006;CD001814.
of infants and parents in the neonatal intensive care environment. In: Sweeney JK, 19. Maguire CM, Veen S, Sprij AJ, et al. Effects of basic developmental care on
ed. The high-risk neonate: developmental therapy perspectives London: Routledge, neonatal morbidity, neuromotor development, and growth at term age of infants who
1986:3–55. were born at ,32 weeks. Pediatrics 2008;121:e239–45.
10. Als H, Lawhon G, Brown E, et al. Individualized behavioral and environmental care for the 20. Bayley N. Bayley Scales of Infant Development. 2nd edn. San Antonio: The
very low birth weight preterm infant at high risk for bronchopulmonary dysplasia: Psychological Corporation, Harcourt Brace & Company, 1993.
21. Meulen BF, Ruiter SAJ, Spelberg HC, et al. BSID-II-NL, deel I: praktische
neonatal intensive care unit and developmental outcome. Pediatrics 1986;78:1123–32.
handleiding, Nederlandse versie. Lisse: Swets Testpublishers, 2002.
11. Als H, Lawhon G, Duffy FH, et al. Individualized developmental care for the very low-
22. Touwen BCL. Neurological development in infancy. London: Heinemann, 1976.
birth-weight preterm infant. Medical and neurofunctional effects. JAMA
23. Touwen BCL. Development of neurological functions in the infant period.
1994;272:853–8.
Eur J Morphol 1995;33:320–1.
12. Westrup B, Kleberg A, von Eichwald K, et al. A randomized, controlled trial to 24. Hempel MS. The neurological examination technique for toddler-age. Groningen:
evaluate the effects of the newborn individualized developmental care and University of Groningen, 1993.
assessment program in a Swedish setting. Pediatrics 2000;105:66–72. 25. Westrup B, Bohm B, Lagercrantz H, et al. Preschool outcome in children born very
13. Buehler DM, Als H, Duffy FH, et al. Effectiveness of individualized developmental prematurely and cared for according to the Newborn Individualized Developmental
care for low-risk preterm infants: behavioral and electrophysiologic evidence. Care and Assessment Program (NIDCAP). Acta Paediatr 2004;93:498–507.
Pediatrics 1995;96:923–32. 26. The International Neonatal Network. The CRIB (clinical risk index for babies)
14. Fleisher BE, VandenBerg K, Constantinou J, et al. Individualized developmental care score: a tool for assessing initial neonatal risk and comparing performance of neonatal
for very-low-birth-weight premature infants. Clin Pediatr (Phila) 1995;34:523–9. intensive care units. Lancet 1993;342:193–8.
15. Ariagno RL, Thoman EB, Boeddiker MA, et al. Developmental care does not alter 27. Fredriks AM, van Buuren S, Burgmeijer RJ, et al. Continuing positive secular growth
sleep and development of premature infants. Pediatrics 1997;100:E9. change in The Netherlands 1955–1997. Pediatr Res 2000;47:316–23.
BMJ Masterclasses
BMJ Masterclasses are educational meetings designed specifically to meet the learning needs of
doctors. They help doctors keep up to date with the latest evidence and recent guidelines in major
clinical areas, enabling them to use the latest evidence to make better decisions. The latest evidence,
recent guidelines and best practice are delivered in an interactive and informative manner by leading
experts. The speakers are specifically chosen as highly-skilled communicators who can authoritatively
enthuse the audience and interpret the latest research and guidelines into practical tips for busy
doctors. BMJ Masterclasses have proved a huge hit with clinicians, with many saying they have
influenced their clinical practice.
http://masterclasses.bmj.com/

Psychological stress of parents of preterm infants

enrolled in an early discharge programme from
the neonatal intensive care unit: a prospective
randomised trial
P Sáenz, M Cerdá, J L Díaz, P Yi, M Gorba, N Boronat, P Barreto and M Vento

online 17 Jul 2008;
doi:10.1136/adc.2007.135921

These include:

Notes


Original article
Psychological stress of parents of preterm infants

enrolled in an early discharge programme from the
neonatal intensive care unit: a prospective
randomised trial
P Sáenz,1 M Cerdá,1 J L Dı́az,2 P Yi,2 M Gorba,1 N Boronat,1 P Barreto,2 M Vento1
1
Servicio de Neonatologı́a, ABSTRACT
Hospital Universitario Materno Background: Psychological stress of parents of preterm What is already known on this topic
Infantil La Fe, Valencia, Spain;
2
Departamento de Psicologı́a de
infants is aggravated by prolonged hospitalisation. Early
la Personalidad, Facultad de discharge programmes (EDPs) have been implemented to c Prolonged hospitalisation of premature babies
Psicologı́a, Universidad de alleviate this situation. causes significant parental stress in the form of
Valencia, Valencia, Spain Objective: To evaluate parental psychological stress in anxiety and depression.
an EDP for the first 3 months after neonatal intensive care c In addition, preterm babies experience
Correspondence to:
Máximo Vento, Neonatal unit (NICU) discharge. difficulties in establishing an adequate bonding
Research Unit, Servicio de Design/methods: Prospective randomised trial compar- with their parents.
Neonatologı́a, Hospital ing parents of preterm infants assigned to EDP (n = 72) or
Universitario Materno Infantil La
Fe, Avenida de Campanar, 21: standard discharge programme (SDP) (standard dis-
E46009 Valencia, Spain; charge) (n = 68). At discharge, parents were evaluated
maximo.vento@uv.es using the Hospital Anxiety and Depression Scale (HAD), What this study adds
and the Likert Scale for well-being every 10 days for
Accepted 16 June 2008 3 months. Parental narrative of Worrying and Helping c The early discharge programme, which included
17 July 2008 issues was assessed using a semi-structured interview. interaction with parents and backup with the
Results: Length of stay was greater in the SDP group primary care paediatrician, substantially
(p,0.01). HAD showed no differences in anxiety, but SDP shortened the length of hospitalisation and did
mothers scored higher in depression (p,0.05). not increase utilisation of public health
Altogether, parents reported a worrisome emotional resources.
condition (EDP 87.2%; SDP 80%), which decreased at the c In addition, it improved early parental-preterm
end of the study (EDP 45.2%; SDP 34.5%). Their baby’s interaction, thus helping to normalise effective
physical well-being was the most relevant issue in the parenthood earlier.
narrative for Worrying and Helping issues at discharge
(EDP 69.2%; SDP 67.5%); however, it decreased at the
end of the study (EDP 22.6%; SDP 24.1%). At discharge, immigrant population. However, survival of extre-
the paediatrician’s support was more for the SDP group. mely premature infants is inevitably associated
No differences on the Well-Being Scale were found, but with prolonging hospitalisation. As a consequence,
the EDP group always scored better. neonates are separated from their families for long
Conclusions: Vulnerability of parents enrolled in an EDP periods of time and immersed in an environment
did not increase after hospital discharge. Physical well- prone to causing medical and psychosocial prob-
being of the baby was the most important issue for both lems.5 6 Hence, babies admitted to the neonatal
groups. EDP parents requested less paediatric support intensive care unit (NICU) will commonly experi-
and scored higher in the Well-being verbatim. ence difficulties in establishing an adequate bond-
Trial registration number: Registered at the Clinical ing with their parents.7 Parents have to acquaint
Trial Government Protocol Registration System no themselves with the ‘‘technological atmosphere’’
NCT00569608. of the NICU, and adapt to the emotional and
physical separation from their child.8 9 Early
discharge programmes (EDPs) aimed at reducing
Over the past two decades the rate of survival of hospitalisation substantially are meant to diminish
extremely premature infants has substantially parental emotional stress,10–14 in addition to redu-
increased in industrialised countries.1 cing costs.15
Socioeconomic improvement, generalised obstet- Parental stress associated with admission to the
rical care, and advances in neonatal care, including NICU of a preterm infant has been extensively
regionalisation of neonatal intensive care, have studied.16 However, very little information regard-
drastically contributed to reducing neonatal mor- ing the psychological impact of EDP upon parental
tality.1–4 Concomitantly, the rate of prematurity stress is available.12 13 17 18
has increased due to socio-medical factors, such as The primary objective of our study was to
women working outside the home, a delay in the determine the degree of psychological stress asso-
first pregnancy, a more widespread use of assisted ciated with EDP from the NICU as compared with
reproductive techniques, and an increased number the standard discharge programme (SDP). We
of uncontrolled pregnancies, especially among the hypothesised that parental stress associated with

Original article
an EDP would be significantly higher when compared with an enteral feeding, no procedures or medication needing hospita-
SDP. To test our hypothesis, we performed a prospective clinical lisation; and (iii) discharge planned at approximately 2 weeks.
trial assigning parents of premature infants born in our At this stage, parents were informed about the study by the
Maternity Unit randomly to either an EDP or an SDP, and study coordinator and consent was requested. Twins were
compared the degree of anxiety, depression, general well-being, always assigned to the same study group because of the
and buffer mechanisms used as protectors to overcome stress. characteristics of the intervention.
Exclusion criteria included: (i) severe congenital malforma-
MATERIAL AND METHODS tions; (ii) severity of clinical condition (eg, intraventricular
haemorrhage grade III or IV, moderate or severe bronchopul-
Methods
monary dysplasia according to the NIH consensus definition,19
Design severe congenital infection); (iii) parents diagnosed with a
This was a prospective, randomised, clinical trial performed at a psychological condition; (iv) non-Spanish-speaking parents; (v)
tertiary level NICU (Hospital La Fe, Valencia, Spain) including refusal to participate by the assigned PCP.
premature infants with birth weight ,2000 g or ,36 weeks’
gestation, admitted to the NICU between October 2005 and
October 2006. The primary care paediatrician (PCP) responsible REFERENCES
for the patients enrolled after discharge from the hospital agreed SDP and EDP characteristics
to participate in the study. The SDP at our NICU includes the following at discharge:
A flow diagram of the study is shown in fig 1. Randomisation clinical stability as defined above; weight .2000 g; corrected
was performed using computer-generated random numbers. gestational age >36 weeks; training of parents to fulfil their
Allocation was performed by using sealed opaque envelopes infant’s requirements adequately.
opened at the time of recruitment. This was performed when The EDP differed at discharge in weight (1600–2000 g);
eligible infants fulfilled the inclusion criteria which were: (i) corrected age ((36 weeks); requirement of a stable social
reaching a postnatal weight of >1600 g; (ii) being clinically background assessed by a sociological score >12 (see table 1)
stable defined as having normal arterial oxygen saturation and programmed clinical visits with the PCP.
(SpO2), heart rate, blood pressure, stools and urine output in the To evaluate psychological well-being in both groups, blinded
week previous to discharge; absence of apnoeic episodes, full psychologist-performed phone interviews were held every
10 days for 3 months post-discharge.
Perinatal variables and clinical follow-up

Major perinatal variables were obtained from the medical
records. Socio-economic variables were obtained using a scoring
system specifically designed for this study (see table 1). Clinical
variables including weight, length, head circumference, and type
of feeding were recorded by the PCP during the follow-up visits.
When two or more of the programmed visits to the PCP were
missed, the case was considered lost to follow-up and
consequently excluded from the study.
Psychological assessment
Evaluation at discharge
Level of psychological adjustment and emotional well-being was
evaluated using the Hospital Anxiety and Depression Scale
(HAD), a self-evaluating scale used as an instrument to
Table 1 Sociological scoring applied to parents of premature infants

discharged from the neonatal intensive care unit with an early or
standard discharge programme
Parameters scored Mother Father
Age Adolescent 2 points

Personal status Single 2 points
Pregnancy Not desired 2 points
Poorly monitored 1 point
Unmonitored 4 points
Residence Frequently changing 6 points 6 points
Homeless 8 points 8 points
Level of instruction Attended school .15 y 2 points 2 points
No school attendance 4 points 4 points
Economic resources Jobless at present 4 points 4 points
Without resources at all 8 points 8 points
Figure 1 Flow diagram of the study design. *Clinical stability was
Addictions Tobacco .5/d 2 points 2 points
defined as having normal SpO2, heart rate, blood pressure, stools, and
Alcohol .2 drinks/d 4 points 4 points
urine output in the week previous to discharge. BW, birth weight; GA,
gestational age; PCP, primary care paediatrician; SpO2, arterial oxygen Other drugs 8 points 8 points
saturation. Parents whose scoring was >12 points were not randomised.

Original article
determine anxiety and depression in a hospital context.20 It Figure 2 depicts the number of eligible, disregarded and
comprises 14 items with a range from 0 to 3. Subscales of included infants who completed the study. Thus, from a total of
depression and anxiety are also valid measures of the severity of 199 eligible infants, 171 patients were randomised. However, 59
emotional changes. Items composing the subscale of depression were excluded, so 140 completed the study. Twenty-eight
are widely based on the core symptoms of the psychopathology infants were excluded because they were not randomised or
of depression. Items composing the subscale of anxiety are incorrectly randomised. After randomisation was completed,
based on clinical manifestations of situational anxiety. Each two infants (EDP group) were lost because they were
scale has a scoring range of 0 to 21. Scores above 10 are transferred to another hospital, and 13 infants (EDP group: 8;
considered clinically significant. SDP group: 5) were disregarded because their parents refused to
continue participating in the trial. In addition, 16 infants did
Evaluation during follow-up not complete the follow-up and were, therefore, excluded. From
Follow-up was performed by a blinded psychologist using a the 140 babies who completed the study, 72 were assigned to
Likert-type Well-Being Scale20 and a semi-structured interview the EDP group and 68 to the SDP group.
focusing on Worrying and Helping issues. Phone calls were No significant differences regarding major clinical character-
made every 10 days (nine calls per family) for 3 months post- istics were found between neonates in the control (SDP) and
discharge. experimental group (EDP) as shown in table 2, nor in their
parents’ characteristics (table 3). However, table 2 shows there
A) Worrying and Helping issues were significant differences at discharge regarding length of
A semi-structured interview was used to assess difficulties and hospital stay, weight, and length and head circumference.
resources. Both aspects were evaluated using two questions: (a)
what has worried or disturbed you the most in the last week? Parents’ vulnerability at discharge
And (b) what has helped you the most or made you feel better Comparison of anxiety and depression between both groups of
in the last week? For each question, parents had three options to parents at discharge using the HAD is shown in fig 3 (A and B).
rank responses. Answers were scored according to the order of No differences regarding anxiety were found between mothers
response (first, second or third place). and fathers in either group. No differences in depression
between fathers in either group were found. However, mothers
B) Well-Being Scale in the SDP group were significantly more depressed than
In order to achieve a comprehensive scoring of parental mothers in the EDP group (p,0.05).
emotional well-being we made a continuous register using a
Likert-type scale. The scoring range for the answers was from 0
to 10 (0 was the minimum, 10 the maximum). Scoring was
performed using the following question: taking into considera-
tion what has worried and helped you in the last days, how
would you score your feeling of well-being in the last days on a
scale from 0 to 10?
This study is part of a more comprehensive EDP including
3 months’ follow-up post-discharge by PCP. The study size was
originally calculated considering approximately 25% lost to
follow-up of initially enrolled patients. The power analysis
indicated that 70 preterm infants were needed in the interven-
tion and control groups, respectively, to achieve a reduction of
10% in the psychological variables studied with a level of
significance a of 0.05.
Subjects were analysed in the study group to which they were
originally assigned (intention-to-treat analysis) Descriptive
statistics were calculated for all the variables determining the
characteristics of the sample. We checked for normal distribu-
tion and ensured an adequate variability. In addition, univariate
analyses were performed using student’s t test for variables with
normal distribution and non-parametric tests (Mann–Whitney
U test) for variables with a non-normal distribution. An analysis
of variance was performed for variables having more than two
possible values. Calculations were made using SPSS 13.0
software.
RESULTS
Population
The EDP differed at discharge in weight (1600–2000 g);
corrected age ((36 weeks); requirement of a stable social
background assessed by a sociological score >12 (see table 1) Figure 2 Flow diagram illustrating enrolment of patients in the study.
and programmed clinical visits to the PCP. EDP, early discharge programme; SDP, standard discharge programme.

Original article
Table 2 Major characteristics of the population enrolled in the early Table 3 Parental characteristics of infants enrolled in the study
discharge programme and standard discharge programme from the NICU Standard discharge Early discharge
Standard Early discharge p (n = 77) (n = 94) p Value
discharge (n = 72) (n = 84) Value
Mother’s age (years) 30.96 32.29 0.095
Gestational age (wks)* 32 (29–35) 33 (30–35) 0.103 Firstborn child (%) 69 55.4 0.134
Mean (SD) birth weight (g) 1603 (373.9) 1629 (319.0) 0.651 Monoparental family (%) 1.7 3.6 0.615
Birth weight ,1500 g (%) 39.7 27.8 0.135 Mother smoker (%) 12.1 5.4 0.205
Mean (SD) birth height (cm) 41.5 (3.3) 41.3 (3.3) 0.761 Father smoker (%) 24.1 19.6 0.572
Mean (SD) birth head 29.1 (2.0) 29.5 (1.9) 0.315 Mother’s school attendance 5.2 3.6 0.517
circumference (cm) ,15 years of age (%)
Male (%) 58.8 51.4 0.377 Father’s school attendance 8.6 5.4 0.378
Multiples (%) 36.2 37.5 0.886 ,15 years of age (%)
Admission to the NICU (%) 45.6 38.9 0.422 Non-desired pregnancy (%) 0 3.6 0.239
Neonatal length of stay (days)* 26 (11.6–57.2) 15.5 (6.0–47.5) 0.001 Teenage mothers (%) 1.7 0 0.509
Mean (SD) discharge weight (g) 2169.1 (177.3) 1872.3 (93.1) 0.001 Previous abortions (%) 32.8 23.2 0.257
Mean (SD) discharge height (cm) 44.4 (1.7) 43.4 (1.5) 0.001 No differences were found for hospital emergency readmission, although the SDP
Mean (SD) discharge head 32.0 (1.2) 31 (1.0) 0.001 exhibited a higher incidence (SDP: 10.3% vs EDP: 4.2%; non-significant).
circumference (cm) EDP, early discharge programme; SDP, standard discharge programme.
*Expressed as median with 5–95% centiles in parentheses.
NICU, neonatal intensive care unit. by the end of the study, it represented 48.4% in the EDP and
60% in the SDP.
Emotional well-being during follow-up Support of the PCP was an important helping issue for 53.8% of
Figure 4 shows the results of the Likert-type scale performed parents in the EDP and for 42.5% in the SDP at discharge from
every 10 days for 3 months post-discharge from hospital. the NICU. At the end of the follow-up, 33.3% in the EDP group
Although there are no significant differences between parents of and 35.5% in the SDP group still considered it a relevant helping
the EDP and SDP groups, the former group consistently issue. In addition, perception of an adequate psychomotor
obtained better scores throughout the study. development of the baby, although not perceived as relevant
immediately post-discharge, acquired more importance during
Worrying in parents the follow-up, and at 3 months after discharge, 64.5% in the
Figure 5A shows the results of the narrative of parental EDP group and 63.3% in the SDP group scored this item as
worrying for premature infants throughout the 3-month highly relevant (see fig 5B).
follow-up. As depicted in fig 5A, immediately after hospital Other factors like coordination of tasks between the couple,
discharge the majority of parents expressed concern. Thus, leisure activities, relationship among siblings, etc, never scored
87.2% in the EDP group and 80% in the SDP group reported a highly enough to be considered relevant.
worrisome emotional condition. However, in the following
weeks, the percentage of parents expressing a perception of DISCUSSION
being worried decreased significantly in both groups. Moreover, Psychological stress in parents of extremely premature infants
at the end of the 3-month follow-up, 45.2% of parents in the due to prolonged hospitalisation has been studied using
EDP group and 34.5% in the SDP group expressed their different perspectives.9 10 16 21 22 Separation of a newly born
emotional situation as worrisome. Figure 5A shows that there infant from its mother is considered the most stressful and
were no differences between the two groups at any point during negative experience for both mother and child. Moreover, the
the study. quality of mother–infant relationship during the first days of
When answers for parental concern were specifically ana- life has been reported as one of the most relevant factors capable
lysed, physical well-being of their children was the most frequent of exacerbating or softening the adverse impact of preterm
response immediately after discharge (EDP group: 69.2%; SDP birth, particularly in relation to subsequent competencies and
group: 67.5%). However, this item rendered less significant and development.23 However, early discharge may be associated
towards the end of the study it was substantially reduced (EDP with medical risks (eg, nutrition, haematology, infections),
group: 22.6%; SDP group: 24.1%). Another relevant worrying familial psychological stress (eg, anxiety, overprotection, fear)
issue immediately after discharge was achieving coordination of and/or overuse of medical facilities (eg, visits to the paedia-
the tasks between the couple (EDP group: 15.3%; SDP group: trician, visits to emergency wards, re-hospitalisation). In order
22.5%). However, at the end of the study only 3% of parents in to avoid these negative consequences, hospital discharge of
both groups expressed worry in relation to this item. This result premature infants needs to be planned in advance following ad
reflects achievement of good coordination of tasks between the hoc established checklists and guidelines.24–26
couple, which helped to cope with the new situation. Other On the other hand, although many studies have analysed the
answers, such as difficulties in maintaining their social life, fear of medical and economical impact of early discharge pro-
re-hospitalisation, jealousy of other siblings, children’s character etc., grammes1 14 15 27–29 few have acknowledged the influence of
never represented a significant percentage of responses in the EDP on normalising preterm infants’ environment and thus
interviews throughout the study and tended to disappear. contributing to provide the social support that may effectively
buffer family distress.12 13 17 30
Helping issues in parents We hypothesised that an EDP, which transferred responsi-
Undoubtedly, physical well-being of their children was perceived as bility from professional caregivers to parents when their baby
the most important item, helping parents overcome the was still relatively immature, could increase parental psycholo-
stressful situation after hospital discharge. This item scored gical distress. Thus, it would be after hospital discharge when
90% in both groups, but it lost relevance during follow-up and parental psychological vulnerability would emerge. However,

Original article
Figure 4 Likert-type scale representing well-being scores of parents of

preterm infants randomised to the early (EDP) and standard (SDP)
discharge programmes, respectively. Scoring was obtained using a
weekly phone interview during a 3-month period after hospital discharge.
for parents, we found no significant differences in the level of

anxiety with the standard discharge group, which reveals a
successful adaptation. Hence, EDP facilitated prompter normal-
isation of family life. In studies that have compared early versus
standard discharge it has been found that maternal anxiety is
increased in mothers randomised to the standard group.17 In our
study, we found greater depression in mothers of the SDP than
in the EDP. It is possible that having been pronounced as
efficient caregivers increased self-assurance and satisfaction in
mothers of the EDP group.17 This finding is relevant because
previous studies on very low birth weight infants have described
a significant relationship between the severity of maternal
depression and children’s mental outcome,22 and altered
methods of family coping.31 32 Preterm-birth-derived depression
Figure 3 (A) Graph representing mean and standard deviation for

parental anxiety at discharge from the neonatal intensive care unit
(NICU) as measured by the Hospital Anxiety and Depression Scale
(HAD). (B) Graph representing mean and standard deviation for parental
depression at discharge from the NICU as measured with the HAD Scale.
Values within the box-plots graph represent median and quartiles.
Outliers have been included. *p,0.05 mother’s depression standard
versus early discharge.
our results show that levels of anxiety in parents of EDPs are

not significantly different from SDPs.13 In Open Neonatal Units,
where parents have free access to their babies 24 h a day, and
where training in basic skills and in fulfilment of special needs
required by hospitalised neonates is promoted, post-discharge
adaptation should seemingly be easier.29 In our study, although
parents in the standard discharge group were allowed to enter
the NICU freely, they tended to rely more on the caregivers,
while parents in the early discharge group were more willing to
take over the responsibility of parenthood. Allen et al found that
Figure 5 (A) Graphic representation of for verbatim (narrative) scores
maternal anxiety, together with an increased parental percep-
expressing worries of parents of preterm infants randomised to an early
tion of child vulnerability, were the most significant variables discharge (EDP) or a standard discharge (SDP) programme. Scoring was
influencing an infant’s subsequent competencies and develop- performed weekly during a 3-month period after hospital discharge. (B)
ment.30 However, our results show that early discharge did not Evolution of the psychomotor development expressed as helping issue to
increase the level of anxiety in mothers or fathers.12 13 Thus, overcome worrying situations. Scoring was performed weekly during a
although early discharge was seemingly an additional stressor 3-month period after hospital discharge.

Original article
may be the result of a lack of interaction between parents and SDP was the baby’s physical well-being and neurodevelopment.
their baby since they are excluded from its care while in the Therefore, early discharge did not modify parental worrying
NICU.33 Therefore, our EDP, promoting an early transfer of issues or requests for help. However, it helped to normalise
responsibility to parents, may have minimised hospitalisation- parenthood earlier.
derived depression. Future studies comparing different strategies Close follow-up of premature infants performed by the PCP
related to the different roles of parents in the NICU (family after an early discharge can prevent and/or detect medical
coping) will help us to identify the main factors influencing the problems early on in the babies as well as psychological stress in
parental adaptation process. parents, which could interfere with effective parenting.
After hospital discharge, parents in both groups expressed
their concern about their infant’s care. However, serial inter- Acknowledgements: This research work was financed by a grant (AP015/06) from
views in the following weeks revealed that the percentage of the Consellerı́a de Sanitat & Bienestar Social (Generalitat Valenciana) to MV and PS.
Thanks are extended to the participating families and primary care paediatricians. We
parents expressing a perception of worry decreased significantly would like to express our special gratitude to Professor Avroy A Fanaroff (Division of
and no differences between the two groups were found. Taking Neonatology; Rainbow Babies & Children’s Hospital; Cleveland; USA) for reviewing and
into consideration that babies in the early discharge group were editing the manuscript.
smaller and younger at discharge, their perception of normal Competing interests: None.
family life was achieved at an earlier post-conceptional age. Ethics approval: The trial protocol was approved by both the Ethics and Research
Consequently, although the Well-Being Scale showed no Committees of our hospital.
significant differences between the two groups, parents in the
EDP always scored better34; this could translate to a greater
confidence in their efficacy as caregivers. However, a possible
REFERENCES
1. Fanaroff AA, Stoll BJ, Wright LL, et al and NICHD Neonatal Research Network.
limitation in the follow-up was reliance on parental perception Trends in neonatal morbidity and mortality for very low birth weight infants.
of the impact of the experience instead of having used objective Am J Obstet Gynecol 2007;196:147e1–e8.
tools for the evaluation of parental stress. Notwithstanding, 2. Hintz SR, Van Meurs KP, Perritt R, et al; NICHD Neonatal Research Network.
Neurodevelopmental outcomes of premature infants with severe respiratory failure
previous studies have shown that subjective reports on well- enrolled in a randomized controlled trial of inhaled nitric oxide. J Pediatr
being are a valuable tool for measuring life satisfaction.21 2007;151:16–22.
In our study, after hospital discharge, family life and worries 3. Broitman E, Ambalavanan N, Higgins RD, et al; National Institute of Child Health and
Human Development Neonatal Research Network. Clinical data predict
were focused on the baby’s physical well-being in both groups neurodevelopmental outcome better than head ultrasound in extremely low birth
as expressed in the verbatim narrative for parental worrying weight infants. J Pediatr 2007;151:500–5.
issues.34 Our findings are in accordance with the EPIPAGE 4. Wilson-Costello D, Friedman H, Minich N, et al. Improved neurodevelopmental
study, which aimed at assessing parents’ psychological health outcomes for extremely low birth weight infants in 2000–2002. Pediatrics
2007;119:37–45.
2 months after discharge of very preterm infants.34 These 5. Ment LR, Vohr B, Allan W, et al. Change in cognitive function over time in very low
researchers used a semi-structured interview to unravel role birth-weight infants. JAMA 2003;289:705–53.
factors in each of the parents. They too found feelings of 6. Whitfield MF, Grunau RE. Teenagers born at extremely low birth weight. Paediatr
anxiety and depression in mothers. However, fathers seemed Child Health 2006;11:275–7.
7. Feldman R, Eidelman AI. Maternal postpartum behaviour and the emergence of
more able to cope and overcome the traumatic event of infant-mother and infant-father synchrony in the preterm and full-term infants: the
prematurity.34 Another key helping issue that contributed to role of neonatal vagal tone. Dev Psychobiol 2007;49:290–302.
overcoming parental distress was their perception of an 8. Fowlie PW, Mc Haffie H. ABC of preterm birth. Supporting parents in the neonatal
unit. BMJ 2004;329:1336–8.
improvement in their babies’ neurodevelopment. Therefore,
9. Jotzo M, Poets CF. Helping parents cope with the trauma of premature birth: an
physical well-being and adequate neurodevelopment seem to be evaluation of a trauma-preventive psychological intervention. Pediatrics
the most important factors that contribute to overcoming 2005;115:915–19.
parental stress after hospital discharge. 10. Charpak N, Ruiz JG, Figueroa de Calume Z. What is the issue when discharging
‘‘premies’’: early discharge from hospital or early integration with the family? Acta
Support given by the primary care paediatrician was also a Paediatr 2001;90:1105–6.
relevant helping issue. Our National Health System provides 11. Saénz P, Diaz Cordobés JL, Cerdá M, et al. Psychological follow up of parents of
universal free healthcare. Therefore, all babies included in the preterm infants enrolled in an early discharge programme from the NICU. EPAS
study had the possibility of free access to paediatric care. During 2007;616280.27.
12. Saénz P, Cerdá M, Gorba M, et al. Psychological distress in parents of premature
follow-up, 10.3% of babies from the SDP as compared with infants at discharge from the NICU. EPAS 2007;616289.26.
4.2% in the EDP were readmitted to hospital as an emergency. 13. Merritt TA, Pillers D, Prows SL. Early NICU discharge of very low birth weight
Seemingly, the closer follow-up programme established by the infants: a critical review and analysis. Semin Neonatol 2003;8:95–115.
14. Merritt TA, Raddish M. A review of guidelines for discharge of premature infants:
primary care paediatricians for patients in the EDP contributed opportunities for improving cost effectiveness. J Perinatol 1998;18:S27–37.
to reducing the need for re-hospitalisation. 15. Yee W, Ross S. Communicating with parents of high-risk infants in neonatal intensive
In order to reduce parental stress, a future approach should care. Paediatr Child Health 2006;11:291–4.
include psychological intervention. It has been shown that 16. Ortenstrand A, Winblandh B, Nordström G, et al. Early discharge of preterm infants
followed by domiciliary nursing care parents’ anxiety assessment of infant health and
maternal emotional responses deeply influence parenting of breastfeeding. Acta Paediatr 2001;90:1190–5.
premature infants.23 Early individualised family-based interven- 17. Kaaresen PI, Ronning JA, Ulvund SE, et al. A randomized, controlled trial of the
tion during hospitalisation and after discharge has proved to effectiveness of an early-intervention programme in reducing parenting stress after
reduce maternal stress and depression and increase maternal preterm birth. Pediatrics 2006;118:9–19.
18. Ehrenkranz RA, Walsh MC, Vohr BR, et al for the National Institutes of Child Health
self-esteem. Moreover, such intervention has improved early and Human Development Neonatal Research Network. Validation of the National
parental-preterm interactions and reduced the length of Institutes of Health Consensus Definition of Bronchopulmonary Dysplasia. Pediatrics
hospitalisation.23 30 2005;116:1353–60.
19. Zigmond AS, Snaith RP. Anxiety and depression scale. Acta Psychiatrica Scand
1983;67:361–70.
CONCLUSIONS 20. Likert R. A technique for the measurement of attitudes. Arch Psychol
1932;140:1–55.
Our EDP significantly reduced the length of hospitalisation of
21. Singer LT, Salvator A, Guo S, et al. Maternal psychological distress and
preterm babies without increasing family stress. After discharge, parenting stress after the birth of a very low-birth-weight infant. JAMA
the most important concern of families in both the EDP and the 1999;281:799–805.

Original article
22. Forcada-Guex M, Pierrehumbert B, Borghini A, et al. Early dyadic patterns of mother-infant 29. Melnyk BM, Feinstein NF, Alpert-Gillis L, et al. Reducing premature infants’ length of
interactions and outcomes of prematurity at 18 months. Pediatrics 2006;118:e107–14. stay and improving parents’ mental health outcomes with the creating opportunities
23. Arnaud F. Discharge of very preterm infants from neonatology: check list. J Gynecol for parent empowerment (COPE) neonatal intensive care unit programme: a
Obstet Biol Reprod 2004:S108–10. randomized, controlled trial. Pediatrics 2006;118:e1414–27.
24. Sauve R, Lee SK. Neonatal follow-up programmes and follow-up studies: Historical 30. Allen EC, Manuel JC, Legault C, et al. Perception of child vulnerability among
and current perspectives. Paediatr Child Health 2006;11:267–70. mothers of former premature infants. Pediatrics 2004;113:267–73.
25. Executive Committee of the Connecticut Chapter of the AAP. Neonatal 31. Spear ML, Leef K, Epps S, et al. Family reactions during infants’ hospitalization in the
Intensive care Unit (NICU). Discharge Guidelines 2005. neonatal intensive care unit. Am J Perinat 2002;19:205–13.
26. Kotagal UR, Perlstein PH, Gamblian V, et al. Description and evaluation of a 32. Wigert H, Johansson R, Berg M, et al. Mothers’ experiences of having
programme for the early discharge of infants from a neonatal intensive care unit. their newborn child in a neonatal intensive care unit. Scand J Caring Sci
J Pediatr 1995;127:285–90. 2006;20:35–41.
27. Casiro OG, McKenzie Me, McFadyen L, et al. Earlier discharge with community- 33. Vento M, Saénz P, Valle S, et al. Early discharge programme from the NICU with co-
based intervention for low birth weight infants: a randomized trial. Pediatrics operation of the Primary Care Paediatrician. EPAS 2006;595535.204.
1993;92:128–34. 34. Garel M, Bahaud M, Blondel B. Consequences for the family of a very preterm birth
28. Raddish M, Merrit TA. Early discharge of premature infants. A critical analysis. Clin two months after discharge. Results of EPIPAGE qualitative study. Arch Pediatr
Perinatol 1998;25:499–520. 2004;11:1299–307.

In springtime, as crocuses emerge, daffodils blossom and lambs
gambol in the fields, paediatricians’ thoughts naturally turn to
orchidometers. A seminal paper in 2001 elegantly demonstrated
how a Teaser sweet could be substituted for an 8 ml
orchidometer bead.1 This provided grateful paediatricians with
a readily available means to assess mid-puberty in adolescent
boys as well as a source of sustenance at the end of a busy clinic.
To the dismay of many paediatricians, by 2007 the testicular
Teaser had mutated into a flat-bottomed dome that retained its
edible qualities but was quite useless as an orchidometer
substitute.2 The manufacturer was urged to reinstate Teasers
to their former aesthetic and functional glory. Fortunately, as
shown in the illustration, the Teaser has been refashioned into
its celebrated orchidometer shape, although somewhat dimin-
ished as a 6 ml rather than 8 ml orchidometer bead, along with
Galaxy, Mars and Milky Way mini eggs. Furthermore, the
Eastertide appearance of the Cadbury mini creme egg provides
an excellent substitute for the 10 ml orchidometer bead with all
the tactile and edible qualities so well described in relation to
the Teaser. The standard Cadbury creme egg, although slightly
too large to substitute as an orchidometer bead, is a useful
indicator of a pathologically enlarged testis. Each spring,
paediatricians should grasp this seasonal opportunity of an
edible orchidometer with both hands. We’ve got the balls—let’s Figure 1 The seasonal orchidometer.
not be afraid to use them.
C Durand, J Gibbs Competing interests: None.
Department of Paediatrics, Countess of Chester Hospital, Chester, Cheshire, UK Arch Dis Child Fetal Neonatal Ed 2009;94:F104. doi:10.1136/adc.2008.140921
Correspondence to: J Gibbs, Department of Paediatrics, Countess of Chester REFERENCES

Hospital, Liverpool Road, Chester CH2 1UL, Cheshire, UK ; john.gibbs@coch.nhs.uk 1. Bhalla P, Sally, Pippa, et al. An inexpensive and edible aid for the diagnosis of puberty
in the male: multispecies evaluation of an alternative orchidometer. BMJ
Acknowledgements: Thanks to R Cooke for his photographic expertise and for 2001;323:1486.
refraining from eating the artwork. 2. Williams G, Dharmaraj P. Dissent of the testis. BMJ 2007;335:1287.

ADC - Fetal & Neonatal - March2009

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

ADC - Fetal & Neonatal - March2009

Enviado por

Direitos autorais:

Formatos disponíveis

Egyptian_Pediatric yahoo group Egyptian_Pediatric yahoo group

F138 How common are rib fractures in extremely low

Cambrian Printers Limited, Aberystwyth, UK. J Cassidy, J Smith, H Pandya, R Firmin,

The seasonal orchidometer

Arch. Dis. Child. Fetal Neonatal Ed. 2009;94;F104

Updated information and services can be found at:

To order reprints of this article go to:

To subscribe to Archives of Disease in Childhood - Fetal and Neonatal Edition go to:

Images in neonatal medicine

The seasonal orchidometer

C Durand, J Gibbs Competing interests: None.

Correspondence to: J Gibbs, Department of Paediatrics, Countess of Chester REFERENCES

F104 Arch Dis Child Fetal Neonatal Ed March 2009 Vol 94 No 2

Risk of stillbirth and neonatal death linked with

Arch. Dis. Child. Fetal Neonatal Ed. 2009;94;F105-F110; originally published

Updated information and services can be found at:

Rapid responses You can respond to this article at:

To order reprints of this article go to:

To subscribe to Archives of Disease in Childhood - Fetal and Neonatal Edition go to:

Risk of stillbirth and neonatal death linked with

c Additional tables are ABSTRACT

Arch Dis Child Fetal Neonatal Ed 2009;94:F105–F110. doi:10.1136/adc.2007.135459 F105

individual between different registers to a high degree of Cause-of-death classifications

Figure 1 Relative risks of stillbirth to

F106 Arch Dis Child Fetal Neonatal Ed 2009;94:F105–F110. doi:10.1136/adc.2007.135459

Figure 2 Neonatal deaths: relative risks

Arch Dis Child Fetal Neonatal Ed 2009;94:F105–F110. doi:10.1136/adc.2007.135459 F107

F108 Arch Dis Child Fetal Neonatal Ed 2009;94:F105–F110. doi:10.1136/adc.2007.135459

Table A1 Classification of causes of stillbirth and neonatal death*

Arch Dis Child Fetal Neonatal Ed 2009;94:F105–F110. doi:10.1136/adc.2007.135459 F109

Drug and Therapeutics Bulletin (DTB)

Your key source of unbiased, independent advice

DTB is now also available online at http://dtb.bmj.com:

To subscribe, or for further information, please visit http://dtb.bmj.com

F110 Arch Dis Child Fetal Neonatal Ed 2009;94:F105–F110. doi:10.1136/adc.2007.135459

Impact of shielding parenteral nutrition from light

Arch. Dis. Child. Fetal Neonatal Ed. 2009;94;F111-F115; originally published

Updated information and services can be found at:

Rapid responses You can respond to this article at:

To order reprints of this article go to:

To subscribe to Archives of Disease in Childhood - Fetal and Neonatal Edition go to:

Impact of shielding parenteral nutrition from light on

Arch Dis Child Fetal Neonatal Ed 2009;94:F111–F115. doi:10.1136/adc.2007.135327 F111

pharmacy prior to commencement of the multivitamin/amino RESULTS

F112 Arch Dis Child Fetal Neonatal Ed 2009;94:F111–F115. doi:10.1136/adc.2007.135327

higher in the group receiving LE TPN (1.5 (0.3) vs 0.9 (0.1),

Effect of light exposure on blood glucose concentration

Figure 4 Blood glucose concentrations as a function of postnatal age.

Arch Dis Child Fetal Neonatal Ed 2009;94:F111–F115. doi:10.1136/adc.2007.135327 F113

F114 Arch Dis Child Fetal Neonatal Ed 2009;94:F111–F115. doi:10.1136/adc.2007.135327

Quality & Safety in Health Care

Arch Dis Child Fetal Neonatal Ed 2009;94:F111–F115. doi:10.1136/adc.2007.135327 F115

Random safety audits in the neonatal unit

Arch. Dis. Child. Fetal Neonatal Ed. 2009;94;F116-F119; originally published

Updated information and services can be found at:

Rapid responses You can respond to this article at:

Editor's choice (371 articles)

To order reprints of this article go to:

To subscribe to Archives of Disease in Childhood - Fetal and Neonatal Edition go to:

Random safety audits in the neonatal unit

Neonatal Unit, Southmead ABSTRACT

F116 Arch Dis Child Fetal Neonatal Ed 2009;94:F116–F119. doi:10.1136/adc.2007.131052

Arch Dis Child Fetal Neonatal Ed 2009;94:F116–F119. doi:10.1136/adc.2007.131052 F117