Vol. 19, No. 2, pp. 156-161
Printed in Great Britain
Journal of Public Health Medicine
Social class, spoken language and pattern of
care as determinants of continuity of carer in
maternity services in east London
Harry Hemingway, Dawn Saunders and Luise Parsons
Abstract
Background The Government's policy of Changing childbirth
gives priority to user-oriented outcomes, such as continuity
of carer. It has been assumed that the organization (or
pattern) of maternity care is the main determinant of
continuity, with relatively little attention paid to sociodemographic factors. The aim of this study was to assess the
relative contribution of social class, spoken language and
pattern of care in determining continuity of carer.
Method Postal questionnaires were sent 14 days after
delivery to East London and the City Health Authority
residents delivering within a three-week period in May
1994. Bilingual interviews were carried out for non-Englishspeaking women. Pattern of care was assigned by the
midwife as either hospital or community (including team
based care, 'domino' and home births). The main outcome
measure was self-reported continuity of carer in antenatal,
delivery and postnatal care.
Results The response rate was 69 per cent (370/533). The
community pattern of care affected only antenatal continuity
(62 per cent community vs 50 per cent hospital, p < 0.05).
Women whose main spoken language was English or whose
social class was l-llln reported higher levels of continuity at
each phase of care, although this effect was largely confined
to the community pattern of care. The odds ratios (95 percent
confidence intervals) for the effect of social class (l-llln vs
other) on antenatal, labour and postnatal continuity within
the community pattern of care were 3.64 (1.09-12.18), 3.08
(1.09-8.74) and 4.93 (1.48-16.46), respectively.
Conclusion Spoken English and high social class were
associated with continuity of carer, although this effect was
mainly confined to women with a community pattern of care.
Achievement of national targets for continuity of carer may
not be possible in east London without explicit consideration
of sociodemographic factors.
Keywords: ethnicity, socioeconomic status, outcomes, consumer satisfaction
Introduction
The report of the Department of Health's Expert Maternity
Group, Changing childbirth1 set out a radical agenda for change
in maternity services based largely on the views and
experiences of women. Numerical targets have been set for
continuity of carer in labour. An enhanced role of the midwife is
proposed, with maternity care increasingly taking place within
the community. Midwifery has responded to these demands
with the establishment of teams of midwives with defined
clinical responsibilities spanning hospital and community
settings. It is assumed that team midwifery leads to an
improvement in continuity. However, there are few data to
say whether, or to what extent, this is true. Indeed, a recent
national survey found that only a third of midwifery units with
teams could identify the proportion of women delivered by a
midwife they knew.2
Perinatal mortality has long been considered the main
outcome measure of maternity services,3 although its use is limited
by strong socioeconomic determinants.4 The consequences of such
social factors have traditionally been held to lie largely outside
the remit of the maternity services. Continuity of carer has
emerged as an important outcome because of the longstanding
complaint women make about seeing a different professional at
each visit,5 being asked the same question by many different
professionals6 or being given conflicting advice. Such useroriented 'soft' outcomes could, at least in theory, be wholly
determined by service delivery. However, in practice, socioeconomic factors may also play a role.7 The aim therefore of
this study was to assess the relative contribution of social class,
spoken language and pattern of maternity care as determinants
of continuity of carer.
1
Department of Public Health, Kensington & Chelsea and Westminster Health
Authority, 50 Eastbourne Terrace, London W2 6LX.
2
Department of Epidemiology and Public Health, UCL Medical School, 1-19
Torrington Place, London WC1E 6BT.
3
Department of Public Health, East London and the City Health Authority,
Tredegar House, 97-99 Bow Road, London E3 2AN.
4
Directorate of Public Health, Bexley and Greenwich Health Authority, 221
Erith Road, Bexley Heath, Kent DA7 6HZ.
Harry Hemingway, Director of Research and Development,1 Senior
Lecturer2
Dawn Saunders, Research Midwife3
Lube Parsons, Director of Public Health4
Address correspondence to Dr H. Hemingway (e-mail: h.hemingwayOpublichcalth. ucl.ac.uk).
©Oxford University Press 1997
SOCIAL CLASS AND CONTINUITY OF CARER
Method
The sample population was all mothers resident in East London
and the City Health Authority delivering during a 21-day period
from 21 April 1994. Mothers were identified by the birth
notification forms generated in the seven maternity units at
which 99 per cent of births take place. Residency in the health
authority was established from the postcode on the birth
notification form. There were 558 health authority residents
delivering during this three-week period, 96 per cent (533) of
whom gave birth in one of the three maternity units within the
health authority. This sample size was chosen so as to have an
80 per cent power to detect at the p < 0.05 significance level a
10 per cent difference in proportion of women reporting
continuity in labour in the hospital and community patterns of
care, assuming a 70 per cent response rate. Women whose child
died in the perinatal period were excluded from the survey.
Approval from the three local research ethical committees was
obtained and women gave informed consent to participate.
The Office of Population Censuses and Surveys (OPCS)
maternity services questionnaire was modified to include
questions on antenatal, labour and postnatal continuity, main
spoken language, ethnicity (OPCS 1991 Census categories) and
the occupation of the woman and her partner. Social class was
coded for both the woman's own occupation and that of her
partner according to the Registrar General classification.9 For
analytical purposes, social class was dichotomized into social
class I-UJ non-manual and other. Housewives were assigned to
the latter group. The modified questionnaire was piloted on four
women, and minor changes were made. Area deprivation was
assigned according to the Jarman UPA8 score.10
Continuity of carer was assessed with the following
questions: 'During your antenatal care, did you get most of
your care from one or two people that you got to know, or did
you tend to see different people each time?' 'Had you met any
of the midwives who looked after you in labour?' (Care from
midwives since leaving hospital) 'Had you met any of the
midwives at any time during your pregnancy or labour?' A
summed continuity score was calculated (one point for each of
three phases of care, to give a maximum score of three).
The conduct of this study did not influence the selection (self
or other) of women to any particular pattern of care. The
definition of pattern of care was defined by a Steering Group
(which included senior midwives from each provider) before
the start of the study. Pattern of care was identified by the
midwife on the birth notification form as representing
predominantly 'hospital' or 'community' based care. Community maternity care included team based care, 'domino'
and home births (n = 8). Dichotomizing care in this way
represented a necessary simplification for analytical purposes,
and its validity was supported by the observation that a community
pattern of care was associated with a higher proportion of women
reporting having a named midwife, midwife-only antenatal
care, a midwife delivery and shorter lengths of stay.
157
The questionnaire was sent daily to arrive 14 days after
delivery. Each questionnaire was accompanied by a personally
addressed letter on health authority letterhead. Women who had
not responded by 24 days post partum were given a telephone
reminder. The delivering midwife identified those women who
did not speak English and they were visited at home by
bilingual interviewers, with prior telephone arrangement where
possible. Nine interviewers covering six languages (Sylheti,
Urdhu, Gujarati, Hindi, Turkish and French) were trained to
administer the questionnaire by the study co-ordinator. The
questionnaire included items which were also recorded on the
birth notification form, such as ethnicity, birthweight and
intervention details. There was excellent agreement between
the birth notification and both interviewer and postally
administered questionnaires on these items.
All analyses were carried out using the statistical software
SAS. Proportions were compared using the x 2 statistic.
Multiple logistic regression (PROC LOGISTIC) was used to
calculate adjusted odds ratios and their 95 per cent confidence
intervals (CIs); the significance of interaction terms was
assessed using PROC GLM.
Results
The overall response rate was 370/533 (69 per cent).
Questionnaires were administered postally for 83 per cent of
respondents and by bilingual interviewer for 17 per cent
Questionnaires were completed at a median of 3 weeks (range
1-16) after delivery. The distribution of deprivation scores as
well as the proportion of women living in areas with a score
above an arbitrary cutpoint showed no differences between
responders and non-responders (Table 1). Nor were there any
differences by age, parity, pattern of care or type of delivery.
Twenty-nine per cent of the responders and 23 per cent of the
non-responders were booked for a community pattern of care.
There was a higher proportion of Bangladeshi women among
responders (20 per cent vs 13 per cent, p < 0.05), possibly
reflecting the availability of Bangladeshi (Sylheti) interviewers.
As telephone reminders were given to non-responders, the
higher proportion of women without a telephone among nonresponders (32 per cent vs 15 per cent, p < 0.05) was not
surprising. Item non-response for social class and spoken
language was not related to continuity.
Table 2 shows the proportion of women who reported
continuity by phase of care. There were 264 women who had a
hospital pattern of care and 106 women who had a community
pattern of care. A higher proportion of women reported
antenatal continuity among a community pattern of care than
among hospital care; 62 per cent vs 50 per cent, p < 0.05. The
percentage of women reporting continuity in labour care was
low in relation to the Cumberlege target of 75 per cent (28 per
cent and 33 per cent in hospital and community care,
respectively). Few women reported continuity at all three
phases of care (10 per cent hospital vs 16 per cent community).
JOURNAL OF PUBLIC HEALTH MEDICINE
158
Table 1 Comparison of responders and non-responders
Age (years)
mean (range)
Parity
median (range)
Responders
(n = 370)
Non-responders
(n = 163)
28.1 (15-47)
27.3(14-39)
2.0(1-12)
2.0(1-12)
Ethnicity (%)
White
Black Caribbean
Black African
Bangladeshi
Other
40
5
17
20
18
36
11
21
13*
19
No telephone (%)
15
32*
High area deprivation
score (Jarman score
66
60
Hospital pattern of care (%)
Normal vaginal delivery (%)
71
75
76
72
*P<0.05.
There was a direct relationship between continuity score
and the proportion of women with a community pattern of care
(p for trend = 0.03), in social class I-Dln (p for trend <0.05),
the proportion of women whose partners were in social class I Dln (p for trend <0.05) and the proportion of women who had
English as a first language (p for trend <0.0006).
To address potential confounding or interaction between
pattern of care and social class and spoken language, stratified
analyses were performed (Tables 3, 4 and 5). Women of
social class I-Illn compared with those of other social class
reported higher antenatal, labour and postnatal continuity
within the community pattern of care only (p = 0.06, 0.06
and 0.01, respectively). Similarly, women whose main spoken
language was English compared with those whose main spoken
language was not English reported higher antenatal, labour and
postnatal continuity within the community pattern of care
(p = 0.01, 0.15 and 0.0001, respectively). There were 183
women for whom English was their first language and who selfassigned ethnicity. Of these, women whose self-reported
ethnicity was white reported similar continuity to women of
other ethnic groups: antenatal 78/121 vs 32/62 (p = 0.09),
Table 2 Proportion of women reporting continuity of carer
by phase and pattern of care
% reporting continuity
Hospital
in = 264)
Community
(n=106)
p value
Antenatal
Labour
Postnatal
All 3 phases of care
50
28
52
10
62
33
56
16
<0.05
0.40
0.49
0.58
delivery 38/119 vs 20/61 (p = 1.00), post-natal 77/121 vs 32/
59 (p = 0.22).
A total of 125 women responded to the question 'if English
is not yourfirstlanguage, how often during your pregnancy was
there a health advocate present who could translate for you
when you were with the midwives and the doctors?' Of these
women, 56 (45 per cent) never had an advocate present during
antenatal care; 34/69 (49 per cent) of women with an advocate
reported antenatal continuity compared with 23/56 (41 per cent)
without an advocate [odds ratio 1.39 (0.64-3.02)].
Table 5 shows the odds ratios (95 per cent CIs) for the effect
of social class and spoken language on continuity of carer.
Social class and main spoken language tended to have stronger
effects in the community pattern of care, although the
interaction terms were only significant for spoken language.
Spoken language and social class were strongly related. The
odds of having English as a main spoken language was 8.68 for
women of social class I-IIIn compared with women of other
social class. To assess the independent contribution of each
variable, multiple logistic regression was performed. Adjustment for spoken language attenuated the odds ratios for social
class in the community pattern of care to 2.24 (0.53-9.51), 3.03
(0.81-11.4) and 2.72 (0.64-11.54) for antenatal, labour and
postnatal continuity, respectively. Adjustment for social class
attenuated the odds ratios for spoken language in the community
pattern of care to 2.01 (0.62-6.58), 1.02 (0.31-3.38) and 2.36
(0.72-7.75) for antenatal, labour and postnatal continuity,
respectively. Further adjustments for age, parity and the number
of antenatal visits did not attenuate these associations. Using
partner's social class instead of the woman's own social class
made no difference to the estimates.
Overall, the proportion of women who rated continuity as
'important' was 92 per cent for antenatal, 44 per cent for delivery
and 21 per cent for postnatal continuity. Neither spoken language
nor social class influenced the importance attached to continuity.
Women who reported continuity of delivery care were more
likely to consider it important than those who did not (77 per cent
vs 22 per cent, p < 0.0001) but there were no comparable effects
for antenatal or postnatal continuity.
Discussion
The community pattern of care was associated with higher
antenatal but not delivery or postnatal continuity. Women
whose main spoken language was English or whose social class
was I-IIln reported higher levels of continuity at each phase of
care; this effect was largely confined to the community pattern
of care. There was some evidence - not statistically significant
- that the effect of spoken language may be mitigated by the
presence of a health advocate. However, before discussing the
interpretation and implications of thesefindings,the limitations
of the data require consideration.
The response rate of 69 per cent and the similarity of
responders and non-responders make a serious non-response
SOCIAL CLASS AND CONTINUITY OF CARER
159
Table 3 Percentage of women reporting continuity by pattern of care and own social class
Hospital
Community
%
Reporting
continuity
Social
class l-llln
(n=100)
Other
social class
(n=111)
Antenatal
Labour
Postnatal
All 3 phases
of care
54
29
56
16
48
28
51
7
bias unlikely. The data from postal and bilingual interviewer
administration may not be wholly comparable; however, postal
administration alone would have resulted in a poor response
rate among non-English-speaking women. 11 ' 12 Women's perceptions of their maternity care change in the post-natal
period. 13 ' 14 However, the questionnaire was completed in a
relatively narrow time period postnatally (median three weeks),
which should minimize a timing bias. It is not known how selfreported continuity relates to objective assessments of continuity, although current policy emphasizes the former.
Although the mutually adjusted odds ratios for spoken
language and social class were consistent with independent
effects on continuity, the result should be interpreted with
caution. The two variables were strongly related and are likely
to have been measured with differing degrees of precision. This
can be shown to lead to spurious inferences in logistic
regression analyses.15 The measurement of social class in
men may be less imprecise, but using partner's social class
made no difference to the results.
These data do not allow a causal link between social class
and continuity of carer to be inferred; social class is likely to be
a marker of one or more specific mediating factors. Furthermore, the specific components of community vs hospital care
merit further consideration. In east London, the number of
deliveries per midwife was 38 per annum compared with the
p value
Social
class l-llln
(n = 23)
Other social
class
(n = 53)
p value
0.4
0.9
0.6
0.05
83
50
83
41
57
25
49
12
0.06
0.06
0.01
0.01
national average of 30. 16 A survey of midwives was carried out
as part of this study and, despite the low response rate [40 per
cent (121/301)], only 10 per cent of midwives felt that they had
enough time to give women the attention they required; this did
differ between community and hospital midwives.
The finding that social class and spoken language were
associated with the experience of continuity mainly within a
community pattern of care is consistent with a self-selection
explanation. Educated women, who are articulate in the
language used by service providers, may seek the pattern of
care which suits their needs. This explanation is supported by
the similarity of continuity between hospital and community
patterns of care seen within women of other social class or
without English as their main spoken language. For these
women, pattern of care has no effect on continuity.
There are no other published studies investigating the effects
of social class and spoken language on continuity of carer,
therefore generalization of the findings to other inner city areas
must be cautious. 'Middle class models of birth' based on
natural childbirth, feminism and consumerism 17 have been
criticized for having little appeal to working class mothers. This
study found no evidence that the importance attached to
continuity was associated with social class. This is consistent
with the observation that the ideal of being in control during
labour is subscribed to by women of all social classes.7 Women
Table 4 Percentage of women reporting continuity by pattern of care and main spoken language
Hospital
%
Reporting
continuity
Antenatal
Labour
Postnatal
All 3 phases
of care
Community
English
as main
spoken
language
(n=149)
Other main
spoken
language
(n=100)
56
29
55
12
40
30
46
8
p value
English
as main
spoken
language
(n = 35)
Other main
spoken
language
(n = 67|
p value
0.02
09
0.18
0.53
80
43
85
33
53
27
43
8
0.01
0.15
0.0001
0.004
160
JOURNAL OF PUBLIC HEALTH MEDICINE
Table 5 Odds ratios (95% confidence intervals) for the effect of social class and main spoken language on continuity
Social class (l-llln vs other)
Antenatal
Labour
Postnatal
Main spoken language (English vs other)
Hospital
Community
Interaction
(pattern of
care x social class)
1.21 (0.71-2.06)
1.06(0.59-1.93)
1.26(0.73-2.18)
3.64(1.09-12.18)
3.08(1.09-8.74)
4 93(1.48-16.46)
0.09
0.60
0.59
want more information to make informed choices,5'18 and in
one study lower social class women wanted more information
but obtained less than their higher status counterparts.19
However, other studies provide conflicting evidence, describing
social class influences on expectations and priorities,20 with
working class women thinking less about procedures,17
attributing outcomes more to chance21 and being less interested
in discussing issues.22*23
Women for whom English was not their main spoken
language valued continuity as highly as English-speaking
women, but consistently reported less of it, even within
community patterns of care. Recent qualitative work suggests
that religion among Asian and non-Asian women is not a
determinant of women's ideas and experiences of pregnancy
and childbirth, and that Asian women show a strong commitment to at least some aspects of Western maternity care.24 We
found that among women whose first language is English,
ethnicity has no effect on continuity. This is consistent with the
prime importance of verbal negotiation in achieving continuity
rather than a lack of cultural sensitivity on the part of service
providers. Women reporting the presence of a health advocate
did have an increased odds (although not statistically
significant) of experiencing continuity of antenatal care.
Nearly half of the women who did not have English as a first
language never had an advocate, and it is possible that
improving the availability of advocacy services would improve
continuity among these women.25
The Cumberlege report is the first UK policy document to
recommend the use of continuity as an important measure of the
quality of maternity services. Continuity of carer may constitute
a form of social support, an intervention with only beneficial
effects on pregnancy outcomes.26 Providing continuity of care
- for example, provision of consistent rather than conflicting
advice and adherence to evidence-based guidelines - may be
facilitated by ensuring continuity of carer. Furthermore, it is
plausible that higher numbers of professionals involved in
maternity care lead to widening practice variations and reduced
cost-effectiveness. It has been argued that a policy of continuity
of carer implies that each midwife should be responsible for a
defined caseload.27 Team midwifery has been the main service
development designed to deliver continuity of carer.2 Although
in some cases team midwifery may improve continuity,28 in
many cases continuity is not systematically monitored.
Hospital
Community
Interaction (pattern of
carex spoken
language)
1.78(1.15-2.76)
0 78(0.48-1.26)
1 46 (0.94-2 26)
2.02 (0.92-4.45)
1.97(0.90-4.29)
3.43(1.46-8.04)
0.01
0.96
0.02
Validation and reliability studies (including in ethnic minority
groups) using a short form questionnaire are currently under
way which may help to redress this.
The Cumberlege report states that, by 1998, 75 per cent of
women should know the midwife who delivered them. Even
though the question used in this survey was less stringent ('had
you ever met any of the midwives who delivered you?'), only
29 per cent of women reported labour continuity. None of the
factors examined was significantly associated with labour
continuity. This mayreflectthe unpredictability of the timing of
delivery. Furthermore, 92 per cent of women rated antenatal
continuity as very important compared with 44 per cent for
delivery and 21 per cent for postnatal continuity. These figures
are lower than previously reported30 and suggest limitations to
policies which emphasize labour continuity.
Spoken English and high social class were associated with
continuity of carer, although this effect was mainly confined to
women with a community pattern of care. Achievement of
national targets for continuity of carer may not be possible in
east London without explicit consideration of sociodemographic factors. Although communication may be a unifying
explanation of the class and language differences, the
possibility of other mediating factors should not be ignored.
Further studies are required to investigate (1) how social class,
spoken language and ethnicity influence continuity of carer, and
therefore (2) interventions to achieve an equitable distribution
of this newly emphasized outcome measure.
Acknowledgements
We gratefully acknowledge: Donna Lamping and Anne Fleissig in
the design of the questionnaire; Margaret Anthony, Irene Davies,
Ruby Edwards, Lynne Thomas and Pearl Welch for their generous
support in carrying out this survey; thereviewers'comments on an
earlier draft of this manuscript The major acknowledgement goes
to the women of east London who participated in this research,
which continues to shape local policy.
References
1 Department of Health. Changing childbirth. Report of the
expert maternity group (Cumberlege report). London:
HMSO, 1993.
SOCIAL CLASS AND CONTINUITY OF CARER
2 Institute of Manpower Studies. Mapping team midwifery: a
report to the Department of Health. London: IMS, 1993.
3 Maternity Services Advisory Committee. Maternity care in
action: part 2. Care during childbirth. London: HMSO, 1984.
4 Knox EG, Marshall T, Kane S, Green A. Social and health
care determinants of area variations in perinatal mortality.
Community Med 1980; 2: 282-290.
5 Reid ME, Mcllwaine GM. Consumer opinion of a hospital
antenatal clinic. Social Sci Med 1980; 14A: 363-368.
6 Maclntyre S. Expectations and experiences of first pregnancy.
Aberdeen: MRC Sociology Unit, 1981.
7 Green JM, Coupland VA, Kitzinger JV. Expectations,
experiences, and psychological outcomes of childbirth: a
prospective study of 825 women. Birth 1990; 17: 15-24.
8 Mason V. Women's experience of maternity care - a survey
manual. London: HMSO, 1989.
9 Employment Department Group Oo. Standard occupational
classification. London: HMSO, 1990.
10 Jarman B. Underprivileged areas: validation and distribution
of scores. Br Med J Clin Res Ed 1984; 289: 1587-1592.
11 Cartwright A. Some experiments with factors that might
affect the response of mothers to a postal questionnaire.
Statist Med 1986; 5: 607-617.
12 Cartwright A. Who responds to postal questionnaires?
J Epidemiol Commun Hlth 1986; 40: 267-273.
13 Bennett A. The birth of a first child: do women's reports
change over time? Birth 1985; 12: 153-158.
14 Erb L, Hill G, Houston D. A survey of parents' attitudes to
their caesarean births in Manitoba hospitals. Birth 1983; 10:
85-91.
15 Davey Smith G, Phillips AN. Declaring independence: why
we should be cautious. J Epidemiol Commun Hlth 1990; 44:
257-258.
16 National Audit Office. Maternity services. Report by the
Comptroller and Auditor General. London: HMSO, 1990.
161
19 Shapiro MC, Najman JM, Chang A, Keeping JD, Morrison
J, Western JS. Information control and the exercise of
power in the obstetrical encounter. Social Sci Med 1983; 17:
139-146.
20 Mcintosh J, Robinson S, Thomson A, eds. Models of
childbirth and social class: a study of 80 working class
primigravidae. In: Midwives, research and childbirth.
London: Chapman and Hall, 1989.
21 Quine L, Rutter DR, Gowen S. Women's satisfaction with
the quality of the birth experience: a prospective study of
social and psychological predictors. J Reprod Infant
Psychol 1993; 11: 107-113.
22 Canadian Medical Association. Obstetrics '87: a report of
the Canadian Medical Association on obstetrical care in
Canada. Can Med Assoc J 1986; 136(Suppl.).
23 Jacoby A. Mothers' views about information and advice in
pregnancy and childbirth: findings from a national study.
Midwifery 1988; 4: 103-110.
24 Woollett A, Dosanjh N, Nicolson P, Marshall H,
Djhanbakhch O, Hadlow J. The ideas and experiences of
pregnancy and childbirth of Asian and non-Asian women in
east London. Br J Med Psychol 1995; 68: 65-84.
25 Parsons L, Day S. Improving obstetric outcomes in ethnic
minorities: an evaluation of health advocacy in Hackney.
J Publ Hlth Med 1992; 2: 183-191.
26 Oakley A. Measuring the effectiveness of psychosocial
interventions in pregnancy. Int J Technol Assessment Hlth
Care 1992; 1: 129-138.
27 Sandall J. Continuity of care? Recent developments in
maternity care in Britain: towards a sociological perspective.
London: University of London, 1991.
28 Flint C, Poulengeris P, Grant A. The 'Know your midwife'
scheme - randomised trial of continuity of care by a team of
midwives. Midwifery 1989; 5: 11-16.
17 Nelson M. Working class women, middle class women
and models of childbirth. Social Problems 1983; 30:
284-297.
29 Lamping DL. Survey of women's experience of maternity
services (short form). London School of Hygiene, 1996.
30 Melia RJ. Consumers' views of the maternity services:
implications for change and quality assurance. J Publ Hlth
Med \99\; tt: 120-126.
18 Maclntyre S. Communications between pregnant women
and their medical and midwifery attendants. Midwives
Chron 1982; November 387-394.
Accepted on 5 November 1996