needs of prisoners and at avoiding giving spurious

labels to large numbers of "normal" criminals.

Criminality and substance misuse were recorded
independently and the diagnosis of personality
disorder reserved for those with other evidence of
disordered functioning.
We did not find the subcategories in the ICD
(ninth revision) useful in describing inmates with
personality disorder as most of the inmates (over
80%) showed pronounced features of two or more
subcategories. Axis II of DSM-III-R suffers from
a similar problem, assigning several labels to
one patient. This phenomenon has been called
comorbidity but can best be regarded as reflecting
the unsatisfactory state of existing classifications of
personality disorder.
Despair over the diagnosis of personality disorder has led some professionals to reject the
diagnosis' and others to reject patients who have
been given the diagnosis. We adopted Lewis's
view that the diagnosis is problematic but indispensable in referring to a group of patients who
show profound psychiatric disturbance but do not
fit readily into other categories of mental illness.4
The inmates we identified stood out from their
peers by virtue of their mental state or behaviour.
Usually the interviewee, other inmates, prison
officers, and doctors shared our view that their
personality problems were of a nature and severity
that warranted psychiatric attention. More time or
information may have yielded more cases, but
we would claim a degree of face validity. A
comprehensive psychiatric service for prisoners
would have to take these inmates into account.
Deciding which diagnosis is primary depends on
the purpose for which the question is asked. Our
criterion in compiling table II was the provision
of services: which problem would dictate the
immediate management of the patient? It represents an oversimplification of the reality of

psychiatric practice.
A MADEN

Department of Forensic Psychiatry,

Institute of Psychiatry,
Loindon SE5 8AF
I Guze SB. Criminalliv and psychiatric disorders. New York: Oxford
University Press, 1976.
2 1)ell S, Robertson G. Sentenced io hospital: offenders in Broadmoor.
Oxford: Oxtord University Press, 1988. (Maudslev monograph
No 32.)
3 Lewis G, Appleby L. Personality disorder: the patients psychiatrists dislike. Brj Psych 1988;153:44-9.
4 Lewis A. Psychopathic personality: a most elusive category.

PsycholMfed 1974;4:133-40.

involved other specialists; and one involved a

general practitioner. It must be emphasised,
however, that some persons reported the opposite
-for example, that general practitioners took care
of them in a more heedful way than previously.
Different kinds of episodes of discrimination were
reported, but particularly common was the refusal
to give the requested health service (37 episodes).
If these findings are confirmed in other prospective
studies the health authorities should consider
intervening with practitioners who are not following the ethical rules of their profession.
UMBERTO TIRELLI
VINCENZO ACCURSO
MICHELE SPINA
EMANUELA VACCHER
AIDS Unit,
Centro di Riferimento Oncologico,
33081 Aviano,
Italy

Delamothe T. America worries about contagion. BMJ 1991302:

1418. (15 June.)

General practitioners' access to

x ray services

SHON LEWIS

SIR,-Dr N E Early states, "It might be pertinent

to ask radiologists how many referrals they reject
(as a proportion of the total) from junior hospital
staff, consultant hospital staff, non-fundholding
general practitioners, and fundholding general
practitioners."'
I do not have any figures for the number of
referrals rejected but can assure him that general
practitioners are not the only group of doctors
being asked to reduce the number of requests
for examinations. We have achieved a considerable
reduction in the numbers of preoperative chest
radiographic examinations and of contrast examinations of the urinary tract that we perform.
Hospital doctors as well as general practitioners are
being asked to reduce their requests for x ray

Department of Psychiatry,
Charing Cross and Westminster Medical School,

examinations.

Child sexual abuse

T S BROWN

Bradford Royal Infirmary,

Bradford,
West Yorkshire BD9 6RJ
I Early NE. General practitioners' access to x ray services.

1991;303:122. (13 July.)

BM_

HIV and discrimination

Complications of pregnancy and

delivery and psychosis in adult
life

SIR,-Although there is much emphasis on the

discrimination carried out by the United States
with the travel restrictions on foreigners with HIV
infection,' very little is officially known on the
discrimination carried out by health workers
against people with HIV infection either in or
outside of the United States. In Italy we conducted
a prospective study to evaluate this phenomenon.
A coded questionnaire was distributed to all
outpatients and inpatients of the AIDS unit of our
institution. Informed consent had been obtained
and questionnaires were anonymous. Between
30 May and 7 August 1991, 86 subjects filled in the
questionnaire. Sixty three used intravenous drugs
(48 men and 15 women), 11 were heterosexuals
(five men and six women), nine were homosexual
men, and three were men without known risk
factors for HIV infection. Among these persons,
84 were HIV positive; the two others were HIV
negative but at high risk of HIV infection. Of the
34 reporting episodes of discrimination by health
workers in public or private Italian institutions,
eight reported more than one episode. Twenty
three of the 42 episodes involved dentists; seven
involved surgeons; six involved internists; four

SIR,-Do obstetric complications constitute a risk

factor for later schizophrenia? From the results of
their national follow up study Dr D John Done and
colleagues conclude that they do not.' Their study
is impressive, with large numbers and elegant
statistical analyses. But is their conclusion justified?
Two aspects of the study give cause for concern:
the statistical power and the clinical factors used to
define risk. With an overall sample size of some
16000 the issue of statistical power might seem
irrelevent. Yet, as the authors themselves hint, the
final number of under 50 cases of schizophrenia
may well be too small to test adequately the
hypothesis in question. Obstetric complications in
general probably confer in the order of a twofold
increased risk of later schizophrenia.2 This is not a
large effect in comparison with that of familial risk
factors, for example. The analyses of the subgroup
of patients with high risk showed that all the
groups with psychosis, including schizophrenia,
but not those with neurosis had an odds ratio of
stillbirth or neonatal death of between 1-4 and 2 4.
The wide confidence intervals on these figures
testify to the small sample sizes and may well

582

explain the inability to show a significant effect.

The second problem is the choice of clinical
variables used to define obstetric risk. A model that
includes only one variable, the prescription of
drugs, as indicating the condition of the baby must
be interpreted with the greatest caution. Understanding of neonatal physiology at that time was
poor, and drug treatments were empirical-for
instance, the main indication for treatment with
nikethamide (Coramine) was impending death
from any cause, and most babies dying in hospital
would have received this as a last resort. It thus
makes little sense to include this as an independent
predictor of neonatal death.
Increased rates of obstetric complications in the
histories of patients with affective psychosis, as
well as schizophrenia, compared with neurotic
patients have been shown before.' One way in
which the authors might examine further the issue
of obstetric risk and later schizophrenia is to look
for an inverse correlation between calculated risk
and age at onset of the illness. Several previous
studies have reported that obstetric complications
predict an earlier onset, and if this can be shown
not to be the case in the reported sample it will
strengthen the authors' conclusions that no link
between obstetric risk and later schizophrenia has
been shown,
London W6 8RP
ANN STEWART
Department in Paediatrics,
University College and Mliddlesex School of Medicine,
London V'C1 E 6JJ
1 Donc DJ, Johnstone EC, Frithi CI), Giolding J, Shepherd PM,
Crow TJr. Complications of pregnancy and delivery its relation
to psychosis in adult life: data from the British perinatal
mortality survey sample. B.MJ 1991;302:1576-80. ( 29 June.:
2 Lewis SW. Congenital risk factors for schizophrenia. I'syvchol
Mtd 1989;19:5-13.
3 Lewis SW, M\urray RMi. Obstetric complication, neurodevelopmental deviance, and risk of schizophrenia. .7 Psvchiatr Res
1987;21:413-2 1.

SIR,-Dr Brendan McCormack's editorial on

sexual abuse and learning disabilities drew attention to an important problem affecting the lives of
children and adults who are unable to speak for
themselves because of severe communication or
cognitive impairments, or both.' The inadequate
protection provided by current law is particularly
worrying.

There is one aspect of the author's discussion of

diagnosis, however, that needs clarification.
Dr McCormack emphasised the difficulty of
recognising that sexual abuse is occurring and
mentioned the presence of "sexualised behaviour,
temper tantrums, and challenging behaviour" as
pointers. Unfortunately, the latter two features are
very common in conditions in the spectrum of
autistic disorders; they arise from the characteristic
severe impairments of understanding of social
interaction and the rules governing social behaviour. "Sexualised behaviour" in the form of
masturbation in public is also frequently seen. If
the author's recommendation that "such behaviour
should always give rise to suspicion of abuse in
an adult with learning disabilities" is adopted
uncritically then this would involve a large proportion of people with autistic conditions. If it is
assumed that the parents or other carers are at
fault, this would add immeasurably to the stress of
looking after those with communication problems
and socially inappropriate behaviour. I have
recent experience of three such cases in which
parents were unjustly accused solely because of
the behaviour of their autistic children-classic
examples of the difficulty of proving a negative.
Autistic children and adults are, of course,
potentially vulnerable to all kinds of abuse. The

BMJ

VOLUME

303

7 SEPTEMBER 1991

fact that they have disturbed behaviour whether or

not they have been abused makes diagnosis even
more difficult. The point I wish to make is that
special care should be taken when assessing the
significance of the behaviour pattern in people
with impairments of social interaction and communication.
LORNA WING
Nationial Autistic Society,
London NW2 SRB
I MN1cCormack B. Sexual abisec and lcarning disabilities. B13M7
1991;303:143-4. (20 Julv.)

SIR,-Dr Andrew J Wiener,' in commenting on

my editorial on the long term effects of child sexual
abuse,2 first constructs a straw man of his own
making by claiming that my argument against
child sexual abuse depends entirely on its propensity to inflict long term harm, then demolishes
this thesis to his own satisfaction.
I agree that child sexual abuse is an evil irrespective of its long term consequences. In condemning
a social evil there is no need to medicalise it by
claiming that it causes illness. On the other hand,
some social ills-for example, unemploymentprobably do contribute to morbidity and mortality
and surely it is permissible to point this out.
Dr Wiener rightly raises the possibility that the
same social and family factors that predict adult
psychopathology may also predict vulnerability to
sexual abuse. We are preparing for publication
data showing that the sexual abuse itself contributes to adult psychopathology independently of
the disrupted and deprived background which
enhances the risks of being abused.
The law on sexual contact with children could be
changed tomorrow and some moral convulsion
could conceivably alter the public's attitude to the
sexual use of children, but such contact would
almost certainly go on producing immediate and
long term psychological and emotional damage to
those children. Child sexual abuse is in my opinion
a medical problem as well as a moral and legal
problem, and the clear recognition of this, far from
"stunting" our view of the rights of children, as
asserted by Dr Wiener, provides information
essential for their protection.
PAUL MULLEN

Department of Psychological Medicine,

University of Otago,

Dunedin,
New Zealand
1 Wiener AJ. Consequences of child sexual abuse. BMJf 1991;303:
415. (17 August.)
2 Mullen PE. The consequences of child sexual abuse. BM7
1991;303:144-5. (20 July.)

Availability of cadaver organs

for transplantation
SIR,-In discussing the availability of cadaver
organs for transplantation Dr David J Hill and
colleagues question whether current clinical criteria
are sufficiently exhaustive to ensure that all brain
stem function has permanently ceased. They
further claim that in North America theatre staff
are concerned that ventilated, beating heart donors
are not truly dead when operations to remove
organs begin and that the same anxieties exist
in the United Kingdom, including among anaesthetists. In support of this contention they quote
our review of the management of multiple organ
donors as stating that anaesthetists both anaesthetise and paralyse brain stem dead donors. This
is misleading. Hypertension and tachycardia are
not unusual during organ retrieval, and because of
the importance of maintaining haemodynamic
stability at this time, we suggested that glyceryl

BMJ VOLUME 303

7 SEPTEMBER 1991

trinitrate, nitroprusside, or isoflurane (a volatile

anaesthetic agent that is a potent vasodilator) could
be used if necessary to control these responses.Explanations for these haemodynamic changes
have included intact spinal reflex arcs between
afferent pain fibres and sympathetic efferent
nerves, humoral responses after adrenal stimulation, and residual brain stem function.' Although
apparently some anaesthetists are more comfortable if volatile anaesthetic agents are administered
during organ retrieval, the vast majority consider
that the use of such agents other than to control
potentially harmful tachycardia and hypertension
is illogical for brain stem dead donors.4 It is
also recommended that somatic motor reflexes,
similarly mediated at a spinal level, should be
controlled with a muscle relaxant simply to facilitate
surgery.
The presence of these reflex haemodynamic and
motor responses to surgical stimulation and the
need to control them during organ retrieval do not
in our view invalidate the current clinical criteria
for diagnosing irreversible damage to the brain
stem and thereby establishing that there is no
prospect of the patient recovering.
A C TIMMINS
C J HINDS

Anacsthetic laboratory,
St Bartholomew's Hospital,
London EC IA 7BE
I Hill D)J, Evans 1)W, (Gresham GA. Availability of cadaver organs
for transplantation. BMJ 1991303:312. (3 August.)
2 Timmins AC, Hinds CJ. M\lanagement of the multiple-organ
donor. Current OpiniOn in Anaestheszology 1991;4:287-92.
3 Wetzel RC, Setzer N, Stiff JL, Roberts MC. Hemodynamic
responses in brain dead organ donor patients. Anesth Analg
1985;64: 125-8.
4 Bodenham A, Park GR. Care of the multiple organ donor.
Intensive Care Med 1989;15:340- 8.

Antepartum haemorrhage and

cervical cancer
SIR,-I was interested in the comment from the
National Maternity Hospital consultants about
bleeding in pregnancy among women with carcinoma of the cervix.' The experience of clinical
colleagues must always be respected when discussing clinical matters, but so often a small
figure variation may be misleading when unusual
conditions are being considered.
Mr John M Stronge and colleagues quote their
experiences of four women with carcinoma of the
cervix presenting in pregnancy at 40, 34, 32,
and 36 weeks of gestation. All had "substantial
haemorrhage." They do not quote any women with
carcinoma of the cervix in this time who presented
with lesser bleeding. This is against my clinical
experience. While preparing the ABC series I
searched my memory and could think of eight to 10
women with invasive carcinoma of the cervix in
pregnancy, of whom two had moderate bleeding;
the rest had considerably less, some having no real
bleeding at all but only spotting.
When such small series are examined we must
turn to published reports; these confirm that
many women with invasive carcinoma of the cervix
discovered in pregnancy have little bleeding. Many
cases, even of invasive cancer, are diagnosed from
cytological screening and subsequent colposcopy
of women with an abnormal smear; these women
do not bleed much and so would weight the figures.
In a series collected by Moore and Gusberg of 22
women with invasive carcinoma of the cervix in
pregnancy "most had no symptoms"; two had
contact bleeding and three had "some bleeding."2
Hence less than 20% of their series could have been
scored as having substantial bleeding.
Cromer and Hawkin reviewed 20 women with
carcinoma of the cervix in pregnancy; four of these
had carcinoma in situ and 16 invasive carcinoma.

Seven of those with invasive carcinoma had some

bleeding but only one (6%) had severe bleeding.'
A bigger series was collected by Hacker et al,
who summarised the findings from other sources
and ended with 263 women with invasive carcinoma
of the cervix out of 579 795 pregnancies. Of these
cancers 168 were diagnosed either in the first
trimester or in the postpartum period and are not
relevant to this discussion. In the remainder, 35%
of the women presented with a discharge or no
symptoms and 65% had some bleeding. But the
range of the duration of the symptom was 2-5-6-1
months (mean 4 5 months)-the bleeding could
not have been substantial or someone would have
taken some action.
In 1906 Bernard Shaw (born as a breech presentation as a district case in The National Hospital)
presented Doctor's Dilemma to the world in
London.' In the preface to that book he states,
"Even trained statisticians often fail to appreciate
the extent to which statistics are vitiated by the
unrecorded assumptions of their interpreters."
GEOFFREY CHAMBERLAIN

St George's Hospital Mledical School,

London SW17 ORE
1 Stronge JM, Boyd W, Rasmussen MiJ. Antepartum haemorrhage
and cervical cancer. BAIJ 1991;303:249-50. (27 July.)
2 Moore D, Gusberg S. Cancer precursors in pregnancy. Obstet
Gvnecol 1959;13:530-8.
3 Cromer J, Hawkin S. Cancer of the cervix and pregnancy. Obstet
Gvnecol 1963;22:346-51.
4 Hacker NF, Berek JS, Lagasse LD, Charles EH, Savage EW,
Mioore JG. Carcinoma of the cervix associated with pregnancy.

Obstet Gynecol 1982;59:735-46.

5 Shaw B. Ihe doctor's dilemma. London: I'enguin, 1946:61.

Breast carcinomas diagnosed in

the 1980s
SIR, -Drs H Joensuu and S Toikkanen compared
breast carcinomas diagnosed in the 1980s with
those diagnosed in earlier decades.' The histological factors compared in the study included
a measurement of the mitotic count and an assessment of the overall tumour grade.
We have recently shown that a delay in fixation
of six hours produces a reduction of about 50% in
the number of observable mitotic figures in breast
carcinomas.2' In some tumours the decreased
mitotic count can result in change in the overall
Bloom and Richardson tumour grade and hence
prognostic group.' We believe that it is now
essential to include details of tissue fixation in any
study comparing breast carcinomas on the basis of
mitotic counts, particularly if tumour grade and
subsequent prognosis are to be discussed.
R D START
S S CROSS
J H F SMITH
Department of Histopathology,
Northern General Hospital,
Sheffield S5 7AU

1 Joensuu H, Toikkanen S. Comparison of breast carcinomas

diagnosed in the 1980s with those diagnosed in the 1940s to
1960s. BMJ 1991;303:155-8. (20 July.)
2 Start RD, Flynn MS, Rogers K, Smith JHF. Delayed fixation
significantly decreases observed mitotic figures in breast
carcinoma. J Pathol 19911163:1 54A.
3 Start RD, Flynn MS, Cross SS, Rogers K, Smith JHF. Is the
grading of breast carcinoma affected by a delay in fixation?
Virchows Arch [A] (in press).

Mental health needs of Asians

SIR,-Dr Tony Dixon reported the results of a
survey, conducted by the Confederation of Indian
Organisations, of Asians living in Haringey.'
I have analysed preliminary findings of a follow
up study of a cohort of first generation immigrants
from the Indian subcontinent, all of whom were

583