Neonatology Section

Comparison of efficacy and safety of exchange transfusion through

different catheterizations: Femoral vein versus umbilical vein
versus umbilical artery/vein*
Yi-Hao Weng, MD; Ya-Wen Chiu, PhD

Objective: To compare the efficacy and safety of exchange
transfusion (ET) via three different catheterization methods:
femoral vein (FV); umbilical vein (UV); and umbilical artery/vein
(UA/V).
Design: A retrospective cohort of neonates who underwent ET
for hyperbilirubinemia between 1996 and 2007 was surveyed.
Subjects with gestational age <33 wks were excluded.
Setting: Neonatal intensive care units in a tertiary referral
hospital.
Patients: A total of 109 neonates with 128 ET procedures (33
via FV, 35 via UV, and 60 via UA/V routes) were analyzed.
Measurements and Main Results: There was no significant
difference in the decline of total serum bilirubin between each
group. When compared with the UA/V group, the transfusion rate
was slower in the FV and UV groups (p < .001). Adverse events
with clinical significance were more common in ET via the UA/V
route than ET via the FV and UV routes (p < .05; odds ratio, 2.4;
95% confidence interval, 1.2–5.0). Neonates with ET via the UA/V
route tended to have more asymptomatic laboratory aberrances
(p < .01; odds ratio, 2.5; 95% confidence interval, 1.3– 4.6). There
were no significant differences in the transfusion rate (p ⴝ .498)
and adverse events (p ⴝ .822) between the FV and UV groups.
Conclusions: ET through the FV route is an effective and secure
method for the treatment of neonatal hyperbilirubinemia when the
UV route is unavailable. Physicians should be cautious when
using UA/V catheterization for ET. (Pediatr Crit Care Med 2011; 12:
61– 64)
KEY WORDS: exchange transfusion; neonatal hyperbilirubinemia;
femoral vein; umbilical artery; umbilical vein

S evere neonatal hyperbiliru-
binemia carries a substantial
risk for long-term neurologic
sequelae and even death (1).
Exchange transfusion (ET) is useful in
rapidly lowering serum bilirubin level
(2). It requires vascular access to push
and pull blood. ET for neonatal hyperbi-
lirubinemia has been thoroughly investi-
gated (3–13). The standard catheteriza-
tion is through the umbilical vein (UV),
because the technique is simple and no
skin puncture is made (1, 13). ET via both
umbilical artery and vein (UA/V) catheter-
*See also p. 110.
From the Department of Pediatrics (YHW), Chang
Gung Memorial Hospital, Chang Gung University Col-
lege of Medicine, Taoyuan, Taiwan; and the Division of
Health Policy Research and Development (YWC), Insti-
tute of Population Health Sciences, National Health
Research Institutes, Miaoli, Taiwan.
This study was supported, in part, by the National
Health Research Institutes.
The authors have not disclosed any potential con-
flicts of interest.
For information regarding this article, E-mail:
yihaoweng@adm.cgmh.org.tw
Copyright © 2011 by the Society of Critical Care
Medicine and the World Federation of Pediatric Inten-
sive and Critical Care Societies
DOI: 10.1097/PCC.0b013e3181dbeb78
izations has also been used (9, 14). Um-
bilical catheterizations, however, may fail
because of a healed umbilicus or because
the pediatrician lacks sufficient experi-
ence in the procedure. Other alternatives
include peripheral and femoral vessels.
ET, using peripheral vessels, has been
shown to be as effective as ET, using
umbilical vein (5, 15). Nevertheless, con-
cerns remain regarding extravasation,
ischemia, and the level of skill required
for the procedure (16). Femoral vein (FV)
catheterization has also been used as a
substitute to umbilical and peripheral
routes (17, 18), but the efficacy and secu-
rity have not yet been studied. In addi-
tion, the effectiveness and adverse events
of ET through UA/V have received little
attention in recent decades (9). The cur-
rent study is the first to investigate the
efficiency and complication rate of ET via
FV and UA/V routes by contrasting ET via
UV route. This study will shed some light
on the use of catheterizations for ET in
the treatment of neonatal hyperbiliru-
binemia.
PATIENTS AND METHODS
Approval to collect the data was granted by
the Institutional Review Board of Chang Gung
Memorial Hospital. Medical charts of neonates
who had received ET through FV, UV, or UA/V
catheterizations for hyperbilirubinemia in the
neonatal intensive care units of Chang Gung
Children’s Hospital from 1996 to 2007 were
reviewed. To help eliminate confounding fac-
tors, neonates with gestational age of ⬍33 wks
were not included, because they are vulnera-
ble to ET (8 –10). In addition, subjects who
underwent ET via other catheterization routes
(such as peripheral vessels) or ET for condi-
tions other than hyperbilirubinemia (such as
polycythemia) were excluded. Any possible eti-
ologies causing neonatal hyperbilirubinemia—
such as glucose-6-phosphate dehydrogenase
(G6PD) deficiency, infants of diabetic mothers,
polycythemia, congenital hypothyroidism,
spherocytosis, bacterial infection (sepsis, urinary
tract infection, omphalitis), gastrointestinal ob-
struction, breast milk feeding, extravascular
hemorrhage (cephalohematoma, bruise), ABO
incompatibility (defined as any blood group A or
B neonate of group O mother), and Rh incom-
patibility (defined as Rh-positive infants born to
Rh-negative mothers)—were recorded.
Any abnormality that occurred within 7
days after ET was classified as an adverse
event. A clinical event was defined as any per-
manent serious sequela (such as death), tran-
sient or prolonged complication (such as ne-
crotizing enterocolitis, leg ischemia, catheter
malfunction, sepsis, omphalitis, apnea, brady-

Pediatr Crit Care Med 2011 Vol. 12, No. 1 61

cardia, cyanosis, respiratory distress, hypoten- Table 1. Demographic and clinical characteristics
sion, hypertension, seizure, or renal failure),
or laboratory abnormality requiring treatment FV (27),a UV (33),a UA/V (49),a
(sodium, ⬍120 mEq/L or ⬎160 mEq/L; potas- n (%) n (%) n (%) pb
sium, ⬍2.5 mEq/L or ⬎9.0 mEq/L; calcium,
Frequency of ET procedure .518
⬍6.5 mg/dL (term)/6 mg/dL (late preterm), or
Once 23 (85.6) 31 (93.9) 40 (81.6)
⬎13 mg/dL; platelet count, ⬍10,000/␮L; glu- Twice 3 (11.1) 2 (6.1) 8 (16.3)
cose ⬍50 mg/dL) (10). A subclinical event was Thrice 0 (0.0) 0 (0.0) 0 (0.0)
defined as any laboratory aberrance without Four times 1 (3.7) 0 (0.0) 1 (2.1)
clinical manifestation and treatment (sodium, Gender .203
120 –130 mEq/L or 150 –160 mEq/L; potas- Male 20 (74.1) 17 (51.5) 30 (61.2)
Female 7 (25.9) 16 (48.5) 19 (38.8)
sium, 2.5–3 mEq/L or 7–9 mEq/L; calcium,
Birth place .787
6.5– 8 mg/dL (term)/6 –7 mg/dL (late pre- Inborn 7 (25.9) 9 (27.3) 16 (32.7)
term), or ⬎11–13 mg/dL; platelet count, Outborn 20 (74.1) 24 (72.7) 33 (67.3)
10,000 –50,000/␮L). Delivery mode .422
The indications for ET were based on es- Cesarean section 6 (22.2) 10 (30.3) 18 (36.7)
tablished guidelines set up by our institution: Vaginal delivery 21 (77.8) 23 (69.7) 31 (63.3)
Gestational age, wks .423
peak total serum bilirubin (TSB) value ⬎10
Mean ⫾ SD 38.3 ⫾ 1.8 37.7 ⫾ 2.2 37.9 ⫾ 1.6
mg/dL at ⬍24 hours old, ⱖ15 mg/dL at 24 – 47 Birth weight, g .138
hours old, ⱖ20 mg/dL at 48 –95 hours old and Mean ⫾ SD 3116 ⫾ 479 2930 ⫾ 475 3149 ⫾ 525
ⱖ25 mg/dL at ⱖ96 hours old. Catheterization Symptoms except jaundice .035
procedures were performed by senior pediatric With 12 (44.4) 5 (15.2) 12 (24.5)
residents or fellows. In the FV method, a 4F Without 15 (55.6) 28 (84.8) 37 (75.5)
Age at ET, days ⬍.001
150-mm radioopaque polyurethane catheter
⬍7 7 (25.9) 29 (87.9) 42 (85.7)
(Becton Dickenson, Franklin Lakes, NJ) was 7–30 20 (74.1) 4 (12.1) 7 (14.3)
inserted, using the Seldinger procedure (17). Hemoglobin, g/dL .401
All catheterizations were performed under ⬍13 17 (63.0) 15 (45.5) 26 (53.1)
sterile conditions. Location of catheters was 13–22 10 (37.0) 18 (54.5) 23 (46.9)
confirmed by radiologic survey before ET. Etiology
Blood group incompatibility 6 (22.2) 16 (48.4) 30 (61.2) .005
Catheters were removed when it was apparent
G6PD deficiency 18 (66.7) 12 (36.4) 9 (18.4) ⬍.001
that another ET would not be needed. Breast feeding 8 (29.7) 2 (6.1) 7 (14.2) .041
Donor blood used for ET was from either Infection 1 (3.7) 2 (6.1) 1 (2.0) .637
Rh negative or O type red cells mixed with AB Extravascular hemorrhage 1 (3.7) 0 (0.0) 4 (8.2) .216
type plasma. Isovolemic double-volume proce- Other 3 (11.1) 2 (6.1) 4 (8.2) .778
dure was carried out by residents and medical Unknown 1 (3.7) 6 (18.2) 3 (6.1) .094
interns. A cycle of ET was completed by with-
FV, femoral vein; UV, umbilical vein; UA/V, umbilical artery/vein; ET, exchange transfusion; G6PD,
drawal and infusion with an aliquot of 5 mL/
glucose-6-phosphate dehydrogenase.
kg. Calcium gluconate (0.1 g) was infused ev- a
Number of neonates; bp values were by ␹2 test, Fisher’s exact test, and analysis of variance when
ery five cycles of ET. In the UA/V group, blood appropriate.
was withdrawn from the UA catheter and in-
fused into the UV catheter. As for FV and UV
groups, blood was withdrawn from and in- UA/V routes. Table 1 summarizes the de- three quarters of neonates with ET via FV
fused into the same catheter. All infants re- mographic data and clinical presenta- route were ⱖ7 days old. Age was signifi-
ceived prophylactic antibiotics and photother- tions by the three different catheteriza- cantly older in the FV group than the UV
apy during the period of ET. tions. Ninety-four (86.2%) neonates and UA/V groups. Furthermore, no signif-
The statistical analyses were conducted, received ET once; 13 (11.9%) required a icant difference was noticed in anemia
using a commercially available program second procedure; and two (1.9%) under- between each group. The most common
(SPSS for Windows, version 12.0, SPSS, went ET four times (Table 1). There was factor leading to neonatal hyperbiliru-
Inc., Chicago, IL). Categorical variables no significant difference in the frequency
were analyzed, using the chi-square test or
binemia was blood group incompatibility
of ET among the FV, UV, and UA/V groups (47.7%), followed by G6PD deficiency
Fisher’s exact test. For comparison between (p ⫽ .518). Birth place, gender, birth
groups with quantitative variables, the null (35.8%) and breast milk feeding (15.6%).
weight, gestational age, and delivery
hypothesis that there was no difference be- The other causes included bacterial infec-
mode carried no significant differences
tween each group was tested by analysis of tion (3.7%), extravascular hemorrhage
between each group. Before ET, 29
variance. Significance was defined as p ⬍ (4.6%), hereditary spherocytosis (2.8%),
(26.6%) neonates had clinical manifesta-
.05. The adverse event related to ET was polycythemia (1.8%), gastrointestinal ob-
calculated by the number of procedures,
tions other than jaundice, including irri-
tability, poor appetite, hypotonia, opis- struction (0.9%), infant of diabetic
rather than the number of neonates. Logis- mother (0.9%), herb intake (0.9%), and
tic regression was used to examine relation- thotonos, seizure, fever, lethargy, apnea,
respiratory distress, and high-pitched cry. congenital hypothyroidism (0.9%). Only
ships among variables.
These manifestations were more com- 9.2% of neonates did not have known
mon in the FV group than the UV and underlying cause. Blood group incompat-
RESULTS ibility was less common in the FV group
UA/V groups. In the UV and UA/V groups,
A total of 109 neonates were enrolled, most neonates were ⬍7 days old at the than the UV and UA/V groups. By con-
including 27 via FV, 33 via UV, and 49 via time of ET. In contrast, approximately trast, G6PD deficiency and breast milk

62 Pediatr Crit Care Med 2011 Vol. 12, No. 1

Table 2. Levels of total serum bilirubin before exchange transfusion subclinical events. Three neonates died
within 3 days after ET, including two in
Total Serum Bilirubin, FV, UV, UA/V,
the FV group and one in the UA/V group.
mg/dL n ⫽ 33a n ⫽ 35a n ⫽ 60a p
All three had a peak TSB level of ⬎45
Mean ⫾ SD 29.8 ⫾ 7.4 26.0 ⫾ 6.2 23.4 ⫾ 6.2 ⬍.001 mg/dL. There were significant differences
Range 12.0–47.6 14.9–47.2 11.6–42.1 in both clinical and subclinical adverse
events among the three groups, especially
FV, femoral vein; UV, umbilical vein; UA/V, umbilical artery/vein. thrombocytopenia. Specifically, the adverse
a
Number of procedures. events with clinical significance were more
common in ET via UA/V route than ET via
Table 3. Efficacy and rate of exchange transfusion FV and UV routes (p ⫽ .015; odds ratio, 2.4;
95% confidence interval, 1.2–5.0). Neo-
FV (30) UV (35) UA/V (57) p
nates with ET via UA/V route tended to have
Total serum bilirubin .394
asymptomatic laboratory aberrances (p ⫽
reduction, % .003; odds ratio, 2.5; 95% confidence inter-
Mean ⫾ SD 34.9 ⫾ 13.7 33.4 ⫾ 9.6 31.4 ⫾ 12.1 val, 1.3– 4.6), mainly in thrombocytopenia
Range 7.7–60.3 9.5–48.5 4.4–53.5 (p ⫽ .001; odds ratio, 4.3; 95% confidence
Rate, mL/kg/min ⬍.001 interval, 1.8 –10.7). In addition, the overall
Mean ⫾ SD 1.02 ⫾ 0.33 0.97 ⫾ 0.26 1.33 ⫾ 0.47
Range 0.52–1.98 0.43–1.79 0.66–2.57
adverse events were not significantly
different between the FV and UV groups
FV, femoral vein; UV, umbilical vein; UA/V, umbilical artery/vein. (p ⫽ .822).

Table 4. Adverse events of exchange transfusion DISCUSSION

FV (33), UV (35), UA/V (60), To date, there are limited data com-
n (%) n (%) n (%) p paring the efficacy and safety of ET via FV
method versus ET via UV methods. Our
Total clinical events 8 (1.9) 3 (0.7) 23 (2.9) .012 study has revealed, for the first time, that
Death 2 (6.1) 0 (0.0) 1 (1.7) .229
Leg ischemia 0 (0.0) 0 (0.0) 1 (1.7) .565 ET via FV route is as effective and secure
Catheter malfunction 2 (6.1) 0 (0.0) 2 (3.3) .354 as ET via UV route. The reason for this
Infection 0 (0.0) 1 (2.9) 2 (3.3) .580 effectiveness is that the tips of catheters
Apnea 0 (0.0) 0 (0.0) 3 (5.0) .175 in both FV and UV routes are all located
Bradycardia 0 (0.0) 0 (0.0) 2 (3.3) .316
Cyanosis 0 (0.0) 0 (0.0) 4 (6.7) .096
in the inferior vena cava. There is in-
Hypotension 0 (0.0) 0 (0.0) 3 (5.0) .175 creasing evidence that percutaneous in-
Renal failure 0 (0.0) 0 (0.0) 1 (1.7) .565 sertion of femoral venous catheters is a
Sodium imbalancea 1 (3.0) 1 (2.9) 1 (1.7) .890 safe procedure with few complications for
Potassium imbalanceb 1 (3.0) 0 (0.0) 1 (1.7) .600 neonates (17, 18). Although bacterial in-
Hypercalcemiac 1 (3.0) 0 (0.0) 0 (0.0) .255
Hypocalcemiad 1 (3.0) 1 (2.9) 1 (1.7) .890 fection after femoral venous insertion has
Platelet ⬍10,000/␮L 0 (0.0) 0 (0.0) 1 (1.7) .565 been reported (19), our study did not find
Total subclinical events 5 (3.8) 12 (8.6) 34 (14.2) .003 any condition of bacterial infection in ne-
Sodium imbalancea 0 (0.0) 1 (2.9) 4 (6.7) .264 onates with ET, using FV catheteriza-
Potassium imbalanceb 0 (0.0) 2 (5.7) 1 (1.7) .266
Hypocalcemiad 0 (0.0) 6 (17.1) 7 (11.7) .056
tions. Sterile procedure, removal of cath-
Platelet 10,000–50,000/␮L 5 (15.2) 3 (8.6) 22 (36.7) .002 eterizations as soon as ET was not
indicated, and prophylactic use of antibi-
FV, femoral vein; UV, umbilical vein; UA/V, umbilical artery/vein. otics may decrease the risk of infection.
a
Sodium, ⬍120 mEq/L or ⬎160 mEq/L in clinical events, 120 –130 or 150 –160 mEq/L in Even though neonates undergoing ET via
subclinical events; bpotassium, ⬍2.5 mEq/L or ⬎9.0 mEq/L in clinical events, 2.5–3 mEq/L or 7–9 FV route had higher TSB levels and more
mEq/L in subclinical events; ccalcium, ⬎13 mg/dL; dcalcium, ⬍6.5 mg/dL (term)/6 mg/dL (late clinical manifestations than those under-
preterm) in clinical events, 6.5– 8 mg/dL (term)/6 –7 mg/dL (late preterm) in subclinical events.
going ET via UV route, the effectiveness
and adverse events of ET were not differ-
feeding were more common in the FV catheter malfunctions. There was no sig- ent between these two groups.
group. nificant difference in the reduction of In our study, neonates with ET via
The TSB levels before ET are pre- TSB level at 1 hr after ET between each UA/V route experienced more adverse
sented in Table 2. There were 128 ET group. However, the speed of ET was sig- events than those via FV and UV routes.
procedures: 33 via FV, 35 via UV, and 60 nificantly faster in the UA/V group than This finding is similar to the results of a
via UA/V routes. The TSB levels were sig- the other two groups (p ⬍ .001). Further- study (9) showing that the complication
nificantly higher in the FV group than more, there were no significant differ- rate in ET through UA/V catheterization
the UV and UA/V groups. Table 3 illus- ences in the transfusion rate between the was higher than that of ET through other
trates the efficacy and rate of ET. Six FV and UV groups (p ⫽ .498). catheterizations. Our study has further
procedures were excluded due to termi- The adverse events attributed to ET extended their inquiry by identifying the
nation in the middle of ET; these involved are summarized in Table 4. There were transfusion rate of ET via FV route was
two required resuscitations and four 85 abnormal episodes: 34 clinical and 51 distinct from ET via FV and UV routes. ET

Pediatr Crit Care Med 2011 Vol. 12, No. 1 63

carried out by simultaneous pull and
push of blood could speed up the proce-
dure (14). In addition, our study found
neonates with ET via UA/V route tended
to have platelets ⱕ50,000/␮L, which is
probably due to platelet consumption or
insufficient platelet production during
fast ET in these patients. Except throm-
bocytopenia, there was no significant dif-
ference in each adverse event among the
three groups as a result of low incidence.
Taken together, we suggest that ET pro-
cedure via UA/V route may increase the
risks of adverse events.
Our study has two important clinical
implications. First, FV route serves as a
safe and efficient alternative to UV cath-
eterization. The umbilical route may be
not available for ET in some circum-
stances, such as closure of umbilical ves-
sels or inexperienced hands. Further-
more, pediatricians have often used FV
catheterization (20). Thus, FV catheter-
ization is a necessary substitute for UV
catheterization. Second, adverse events
were more common in neonates with ET
via UA/V route. A number of studies have
shown a link between umbilical arterial
catheterization and serious complica-
tions (9, 21, 22). As reported in those
studies, our findings suggest that ET, us-
ing UA/V catheterization, should not be
recommended even though it can save
time.
The mortality rate of ET has been es-
timated to be around 2% (10). In our
study, three neonates with maximum
TSB value of ⬎45 mg/dL died (mortality
rate, 2.3%). As none of the survivors had
peak TSB levels exceeding those of the
three fatalities, we assumed that death
was due, at least in part, to acute biliru-
bin encephalopathy. In addition, the
three infants who died were unstable be-
fore ET. Several publications (5, 9, 10)
showed that unstable neonates are more
susceptible to complications in ET. These
findings indicate that physicians should
be cautious when employing ET for ill
neonates.
In this study, the most common etiol-
ogy for hyperbilirubinemia was blood
group incompatibility, followed by G6PD
deficiency. Blood group incompatibility
and G6PD deficiency have been charac-
terized as two major causes of neonatal
hyperbilirubinemia in Taiwan (2, 23, 24).
Neonates with blood group incompatibil-
ity often present with early-onset hyper-
bilirubinemia (2). In contrast, the devel-
opment of hyperbilirubinemia at age ⱖ7
64
days old is more common in G6PD-
deficient newborns (23). Furthermore,
attempts at umbilical catheterization
usually fail in neonates ⱖ7 days of age.
These differences account for the distinct
characteristics between FV catheteriza-
tion and umbilical catheterization
groups.
There were several limitations to this
study. First, our study was not a random-
ized clinical trial. To the best of our
knowledge, a randomized trial comparing
these three models of ET has yet to be
published. Second, we did not compare
our three methods with other catheter-
izations, such as single umbilical artery
and peripheral vessels. Third, the differ-
ences in clinical characteristics among
the three groups may lead to bias regard-
ing the safety and efficiency of ET. But
the regression analysis revealed that dif-
ferences in the adverse events were not
influenced by other factors, including ini-
tial symptoms, bilirubin levels, age at ET,
hemoglobin value, and causes of neonatal
hyperbilirubinemia (data not shown).
In conclusion, ET via UA/V method is
associated with more adverse events than
ET via FV and UV methods. Physicians
should be cautious when using the UA/V
route for ET. As ET via FV and ET via UV
have similar efficacy and risk of adverse
events, we conclude that FV is a useful
alternative route for ET. This study pro-
vides support for the use of FV catheter-
ization for ET in the treatment of neona-
tal hyperbilirubinemia when UV
catheterization is not available.
REFERENCES
1. Dennery PA, Seidman DS, Stevenson DK:
Neonatal hyperbilirubinemia. N Engl J Med
2001; 344:581–590
2. Weng YH, Chiu YW: Spectrum and outcome
analysis of marked neonatal hyperbiliru-
binemia with blood group incompatibility.
Chang Gung Med J 2009; 32:400 – 408
3. Behjati S, Sagheb S, Aryasepehr S, et al:
Adverse events associated with neonatal ex-
change transfusion for hyperbilirubinemia.
Indian J Pediatr 2009; 76:83– 85
4. Salas AA, Mazzi E: Exchange transfusion in
infants with extreme hyperbilirubinemia: An
experience from a developing country. Acta
Paediatr 2008; 97:754 –758
5. Chen HN, Lee ML, Tsao LY: Exchange trans-
fusion using peripheral vessels is safe and
effective in newborn infants. Pediatrics 2008;
122:e905– e910
6. Badiee Z: Exchange transfusion in neonatal
hyperbilirubinaemia: Experience in Isfahan,
Iran. Singapore Med J 2007; 48:421– 423
7. Al-Hiali S, Al-Diwan JK, Al-Janabi M, et al:
Exchange transfusion in a neonatal unit in
western Iraq. Ann Trop Paediatr 2007; 27:
155–156
8. Sanpavat S: Exchange transfusion and its
morbidity in ten-year period at King Chula-
longkorn Hospital. J Med Assoc Thai 2005;
88:588 –592
9. Patra K, Storfer-Isser A, Siner B, et al: Ad-
verse events associated with neonatal ex-
change transfusion in the 1990s. J Pediatr
2004; 144:626 – 631
10. Jackson JC: Adverse events associated with
exchange transfusion in healthy and ill new-
borns. Pediatrics 1997; 99:E7
11. Keenan WJ, Novak KK, Sutherland JM, et al:
Morbidity and mortality associated with ex-
change transfusion. Pediatrics 1985; 75:
417– 421
12. Tan KL, Phua KB, Ang PL: The mortality of
exchange transfusions. Med J Aust 1976;
1:473– 476
13. Farquhar JW, Smith H: Clinical and bio-
chemical changes during exchange transfu-
sion. Arch Dis Child 1958; 33:142–159
14. Ata M, Holman CA: Simultaneous umbilical
arteriovenous exchange transfusion. Br
Med J 1966; 2:743–745
15. Fok TF, So LY, Leung KW, et al: Use of
peripheral vessels for exchange transfusion.
Arch Dis Child 1990; 65:676 – 678
16. Scheer B, Perel A, Pfeiffer UJ: Clinical review:
Complications and risk factors of peripheral
arterial catheters used for haemodynamic
monitoring in anaesthesia and intensive care
medicine. Crit Care 2002; 6:199 –204
17. Wardle SP, Kelsall AW, Yoxall CW, et al:
Percutaneous femoral arterial and venous
catheterisation during neonatal intensive
care. Arch Dis Child Fetal Neonatal Ed 2001;
85:F119 –F122
18. Serrao PR, Jean-Louis J, Godoy J, et al: Infe-
rior vena cava catheterization in the neonate
by the percutaneous femoral vein method.
J Perinatol 1996; 16:129 –132
19. Asnes RS, Arendar GM: Septic arthritis of the
hip: a complication of femoral venipuncture.
Pediatrics 1966; 38:837– 841
20. Kanter RK, Zimmerman JJ, Strauss RH, et al:
Central venous catheter insertion by femoral
vein: Safety and effectiveness for the pediat-
ric patient. Pediatrics 1986; 77:842– 847
21. Hilliard J, Schreiner RL, Priest J: Hemoperi-
toneum associated with exchange transfu-
sion through an umbilical arterial catheter.
Am J Dis Child 1979; 133:216
22. Malloy MH, Nichols MM: False abdominal
aortic aneurysm: An unusual complication of
umbilical arterial catheterization for ex-
change transfusion. J Pediatr 1977; 90:
285–286
23. Weng YH, Chiu YW: Clinical characteristics
of G6PD deficiency in infants with marked
hyperbilirubinemia. J Pediatr Hematol On-
col 2010; 32:11–14
24. Weng YH, Chou YH, Lien RI: Hyperbiliru-
binemia in healthy neonates with glucose-6-
phosphate dehydrogenase deficiency. Early
Hum Dev 2003; 71:129 –136
Pediatr Crit Care Med 2011 Vol. 12, No. 1