DOI: 10.1111/j.1471-0528.2007.01294.

x
www.blackwellpublishing.com/bjog
General obstetrics

The diagnostic accuracy of external pelvimetry

and maternal height to predict dystocia in
nulliparous women: a study in Cameroon
AT Rozenholc,a SN Ako,b RJ Leke,b M Boulvaina
a Unité
de Développement en Obstétrique, Department of Gynecology and Obstetrics, University Hospital, Geneva, Switzerland
b Maternité
principale, Hôpital Central, Yaoundé, Cameroon
Correspondence: A Rozenholc, Unité de Développement en Obstétrique, Department of Gynecology and Obstetrics, University Hospital,
Bd de la Cluse 32, Geneva 14 – CH 1211, Switzerland. Email alexrozenholc@yahoo.com

Accepted 14 January 2007.

Objective In many developing countries, most women deliver at
home or in facilities without operative capability. Identification
before labour of women at risk of dystocia and timely referral to
a district hospital for delivery is one strategy to reduce maternal
and perinatal mortality and morbidity. Our objective was to
assess the prediction of dystocia by the combination of maternal
height with external pelvimetry, and with foot length and
symphysis-fundus height.
Design A prospective cohort study.
Setting Three maternity units in Yaoundé, Cameroon.
Population A total of 807 consecutive nulliparous women at term
who completed a trial of labour and delivered a single fetus in
vertex presentation.
Methods Anthropometric measurements were recorded at the
antenatal visit by a researcher and concealed from the staff
managing labour. After delivery, the accuracy of individual and
combined measurements in the prediction of dystocia was
analysed.
Main outcome measures Dystocia, defined as caesarean section
for dystocia; vacuum or forceps delivery after a prolonged labour
(>12 hours); or spontaneous delivery after a prolonged labour
associated with intrapartum death.
Results Ninety-eight women (12.1%) had dystocia. The
combination of a maternal height less than or equal to the 5th
percentile or a transverse diagonal of the Michaelis sacral
rhomboid area less than or equal to the 10th percentile resulted in
a sensitivity of 53.1% (95% CI 42.7–63.2), a specificity of 92.0%
(95% CI 89.7–93.9), a positive predictive value of 47.7% (95% CI
38.0–57.5) and a positive likelihood ratio of 6.6 (95% CI 4.8–9.0),
with 13.5% of all women presumed to be at risk. Other
combinations resulted in inferior prediction.
Conclusion The combination of the maternal height with the
transverse diagonal of the Michaelis sacral rhomboid area could
identify, before labour, more than half of the cases of dystocia in
nulliparous women.
Keywords Cephalopelvic disproportion, dystocia, height,
pelvimetry, sensitivity, specificity.

Please cite this paper as: Rozenholc A, Ako S, Leke R, Boulvain M. The diagnostic accuracy of external pelvimetry and maternal height to predict dystocia in
nulliparous women: a study in Cameroon. BJOG 2007;114:630–635.

Access to district hospitals to perform obstetrical inter-

Introduction
Maternal and perinatal mortality are very high in developing
countries. The worst figures show a maternal mortality 100
times1 and a perinatal mortality 10 times2 those of developed
countries. Dystocia is the underlying cause of about one-third
of maternal deaths, the immediate cause being haemorrhage
due to uterine rupture or atony following prolonged labour,
or sepsis following prolonged rupture of membranes.3 Dys-
tocia can also lead to severe maternal morbidity (e.g. genital
fistula), perinatal death or severe morbidity in the neonate
(e.g. cerebral damage).4,5
ventions when needed is essential to reduce maternal and
perinatal mortality.6 Caesarean section can be life-saving for
both the mother and the infant in case of severe dystocia. As
caesarean section can not be performed in peripheral health
centres, it is crucial to identify women at risk of dystocia
before labour, and to refer them for delivery in district
hospitals. This concerns mainly nulliparous women, as in
multiparous women, the best predictor of dystocia is poor
obstetrical history.7,8
Maternal height has been shown to be associated with dys-
tocia.9 This measurement is routinely used in most antenatal

630 ª 2007 The Authors Journal compilation ª RCOG 2007 BJOG An International Journal of Obstetrics and Gynaecology

clinics, despite a limited prediction. Symphysis-fundus
height,10 shoe size11,12 and clinical internal pelvimetry13,14
result in a prediction inferior to that of maternal height.
Some authors reported that external pelvimetry has a
limited value to identify women at risk of dystocia.15,16 In
contrast, Liselele et al.17 showed that the addition of the
measurement of the transverse diagonal of the Michaelis
sacral rhomboid area (in short the Michaelis transverse,
Figure 1) to the maternal height could increase the sensitiv-
ity in predicting dystocia from 21% to 52%, with a positive
predictive value of 24%.
Our primary objective was to assess the accuracy of external
pelvimetry (specifically the addition of the measurement of
the Michaelis transverse to the maternal height) in the pre-
diction of dystocia in a different population. Our secondary
objective was to compare combinations of maternal height
with other external pelvic measurements, with the foot length
and the symphysis-fundus height, in order to identify nul-
liparous women at risk of dystocia.
Methods
Data were collected in one peripheral urban and the two referral
maternity units of Yaoundé, the capital of Cameroon. All
centres offered antenatal and delivery care, including caesarean
section. Consecutive nulliparous women presenting at the ante-
Figure 1. Intertrochanteric diameter (A); antero-posterior diameter (B);
blue bar: transverse diagonal of the Michaelis sacral rhomboid area (C).
Modified from Liselele HB, et al. BJOG 2000;107:947–52.17
ª 2007 The Authors Journal compilation ª RCOG 2007 BJOG An International Journal of Obstetrics and Gynaecology
Pelvimetry and height to predict dystocia
natal clinics for a third trimester visit were included. A few
women with an obviously abnormal pelvis and women with
twin pregnancy were not included (exact number not recorded).
One research assistant (doctor or midwife) was trained to
perform the measurements in each centre. Maternal height,
pelvic and foot length measurements were performed at the
antenatal visit. Foot gauges were specially designed, fixing
a measuring tape on a wooden plank. Pelvic measurements
consisted of the antero-posterior diameter (also named Baude-
locque or external conjugate), the intertrochanteric diameter
and the Michaelis transverse (Figure 1). The Michaelis trans-
verse is defined by the distance between the two visible depres-
sions in the skin overfacing the sacro-iliac joints.
The antero-posterior and intertrochanteric diameters were
measured using a Breisky pelvimeter, while the Michaelis
transverse was measured using a tape measure. All measure-
ments were recorded to the nearest 0.5-cm interval. Results
were kept in a closed envelope attached to the antenatal file to
allow collection after delivery. These measurements were not
available to the clinician in charge of the delivery and thus
were not used for decision making during labour. Moreover,
the research assistants who performed the measurements were
not involved in the delivery. Symphysis-fundus height and
abdominal circumference were measured in the last 426
included women, at the admission for labour.
Information on mode of delivery and outcome was
obtained from the delivery room register. Exclusion criteria
at delivery were nonvertex presentation, birthweight less than
2500 g, elective caesarean section and caesarean section for
reasons other than dystocia.
Dystocia was defined as caesarean section for dystocia, as
assessed by the clinician in charge based on the partograph;
vacuum or forceps delivery after a prolonged labour (more
than 12 hours) or spontaneous delivery after a prolonged
labour associated with intrapartum death.
During the first phase of the study, data were collected
in the three centres, while during the second phase data
were collected only in one centre. During the first phase, the
antenatal measurements were performed by several observers
trained to perform the measurements by the principal inves-
tigator (A.R.) (phase 1, 467 women included), while during
the second phase, the antenatal measurements were per-
formed by a single observer who did not participate in this
training (phase 2, 340 women included).
Means were compared using the t-test. Cutoff values for all
the measurements were defined as the values closest to the 5th
and 10th percentiles of our population. These cutoffs were
chosen according to the results of the study by Liselele et al.17
Sensitivity, specificity, positive predictive value and the posi-
tive likelihood ratio (sensitivity divided by [1 – specificity])
with their 95% confidence intervals (CI) were calculated
using these thresholds. Various combinations of maternal
height with pelvic, foot length and symphysis-fundus height
631
 Rozenholc et al.

measurements were assessed. As an example, in the combina-

893 nulliparous
tion of the maternal height with the Michaelis transverse, women
women at risk were either those with a maternal height infe-
rior or equal to the cutoff, or those with a Michaelis transverse
inferior or equal to the cutoff. The prediction of dystocia by
the different measurements and combinations was compared
when one or several observers performed the measurements.
Antenatal anthropometric
Data management and analysis were performed using EpiInfo measurements
version 6 (CDC, Atlanta, GA, USA) and Medcalc version 7.4
(MedCalc software, Mariakerke, Belgium).
Assuming a prevalence of dystocia of 10% and a proportion
of positive test results of 10%, we calculated that a sample size Excluded after delivery*
n = 86
of 610–960 women was needed to obtain a precision of ±10%
in the evaluation of sensitivity and predictive value of the test
ranging between 20% and 80%. 807 completed trials of labour

All participants gave oral informed consent. The study pro-

tocol was approved by the ethics committee of the Yaoundé
University and by the authorities of the hospitals involved in
the study.
Test positive†, n = 109 Test negative‡, n = 698

Results
Between March 2002 and April 2004, we included 893 women
at the antenatal clinics. After delivery, 86 women were Dystocia, No dystocia, Dystocia, No dystocia,
n = 52 n = 57 n = 46 n = 652
excluded for nonvertex presentation (n = 22); birthweight less
than 2500 g (n = 38); elective caesarean section (n = 2) and Figure 2. Stard flow diagram. *Excluded after delivery for: non-vertex
caesarean section for reasons other than dystocia (n = 24). presentation (n = 22); birthweight less than 2500 g (n = 38);
Thus, the analysis included 807 nulliparous women who com- elective caesarean section (n = 2); caesarean section for reasons
other than dystocia (n = 24). †Test positive if maternal heigtht = 5th
pleted a trial of labour and delivered a single fetus in vertex
percentile or Michaelis transverse = 10th percentile. ‡Test negative
presentation weighing at least 2500 g (Figure 2). if maternal heigtht > 5th percentile and Michaelis transverse > 10th
The proportion of deliveries complicated by dystocia was percentile.
12.1% (98/807). There were 7.7% (62/807) caesarean section
for dystocia, 2.1% (17/807) vacuum or forceps after a pro-
longed labour and 2.3% (19/807) spontaneous deliveries after
a prolonged labour associated with intrapartum death. Over- Table 1. Comparison of maternal characteristics and birthweight
all, there were 62 perinatal deaths (77 per 1000 births) of between groups
which 40 were associated with dystocia.
Maternal height, all pelvic measurements and foot length Variables Normal Dystocia P value*
delivery (n 5 98)
were smaller in the dystocia group than in the normal delivery
(n 5 709)
group. Conversely, symphysis-fundus height and birthweight
were higher in the dystocia group. Abdominal circumference
Height 162.2 (5.7) 155.4 (6.3) , 0.001
was similar in the two groups (Table 1). Michaelis transverse 10.9 (1.1) 10.1 (1.6) , 0.001
There was no significant difference in the distribution of Intertrochanteric diameter 25.1 (2.9) 23.9 (2.9) , 0.001
maternal height between women included during phase 1 Antero-posterior diameter 21.2 (3.4) 19.4 (2.3) , 0.001
(measurements performed by several observers) or phase 2 Foot length 22.9 (2.4) 21.4 (2.0) , 0.001
(measurements performed by a single observer). The 5th per- Symphysis-fundus height** 33.5 (2.7) 34.9 (2.9) , 0.001
centile was 150 cm and the 10th percentile was 153 cm. In Abdominal circumference** 94.5 (5.9) 94.5 (5.2) 0.997
contrast, there was a significant difference in the distribution Birthweight 3173 (404) 3463 (400) , 0.001
of the other measurements. The values, in centimetres, cor-
responding to the 10th percentile in phase 1 and phase 2 were, All measurements in centimetres, except birthweight in grams.
respectively: Michaelis transverse 9.0 and 10.0; intertrochan- Values are given as means (SD).
*Computed by t-test.
teric diameter 20.0 and 23.0; antero-posterior diameter 18.0
**Measurements were performed in 426 women.
and 17.0 and foot length 20.5 and 19.5. The different values

632 ª 2007 The Authors Journal compilation ª RCOG 2007 BJOG An International Journal of Obstetrics and Gynaecology
 Pelvimetry and height to predict dystocia

Table 2. Prediction of dystocia by maternal height, external pelvimetry, foot length and symphysis-fundus height: univariate analysis

Sensitivity Specificity Positive predictive value Positive likelihood ratio

Height  5th percentile 28.6 (19.9–38.6) 98.4 (97.2–99.2) 71.8 (55.1–85.0) 18.4 (9.6–35.3)
Michaelis transverse  10th percentile 45.9 (35.8–56.3) 92.7 (90.5–94.5) 46.4 (36.2–56.8) 6.3 (4.4–8.7)
Intertrochanteric diameter  10th percentile 26.5 (18.1–36.4) 88.9 (86.3–91.1) 24.8 (16.9–34.1) 2.4 (1.6–3.5)
Antero-posterior diameter  10th percentile 16.3 (9.6–25.2) 88.7 (86.1–90.9) 16.7 (9.8–25.6) 1.4 (0.9–2.3)
Foot length  10th percentile 24.5 (16.4–34.2) 92.1 (89.9–94.0) 30.0 (20.3–41.3) 3.1 (2.0–4.7)
Symphysis-fundus height  90th percentile 28.3 (17.4–41.4) 89.1 (85.4–92.1) 29.8 (18.4–43.4) 2.6 (1.6–4.2)

Values are given as % (95% CI).

corresponding to the tenth percentile in each phase were used
in the overall analysis. Therefore, cutoffs are reported as per-
centiles instead of centimetres.
Maternal height and the Michaelis transverse had the
highest sensitivity, specificity, positive predictive value and
positive likelihood ratio (Table 2). The intertrochanteric
diameter, the antero-posterior diameter, the foot length and
symphysis-fundus height did not predict as well. The combi-
nation of a maternal height less than or equal to the 5th
percentile or a Michaelis transverse less than or equal to the
10th percentile resulted in the best sensitivity, specificity, posi-
tive predictive value and positive likelihood ratio (Table 3).
The addition of a symphysis-fundus height superior or equal
to the 90th percentile to the above combination increased the
sensitivity, at the cost of an increased proportion of women
presumed to be at risk.
The prediction of perinatal death associated with dystocia
by the combination of maternal height with the Michaelis
transverse (sensitivity of 55.0% and specificity of 88.7%)
was similar to the prediction of all cases of dystocia.
The prediction by all individual and combined measure-
ments was within the same range in the two phases of the
study (Table 4).
Discussion
This study confirms that the combination of the measure-
ments of the maternal height with the transverse diagonal
of the Michaelis sacral rhomboid area is a valuable method
to screen nulliparous women during pregnancy for the occur-
rence of dystocia at delivery.
The proportion of dystocia was 12.1%, within the range of
4.0–22.0% reported in sub-Saharan Africa.8,18–21 The propor-
tion of caesarean section for dystocia was 7.7%, within the
range of 1.5–8.5% reported in the same countries.22 The pres-
ent work considered not only caesarean section, but other
outcomes of labour likely associated with dystocia and
focused on nulliparous women. These two factors contributed
to a relatively high percentage of dystocia. The proportion of
perinatal death among all deliveries and the fraction due to

Table 3. Prediction of dystocia by combinations of maternal height with the Michaelis transverse, intertrochanteric diameter, foot length and symphysis-
fundus height

Combinations Women at risk Sensitivity Specificity Positive Positive

predictive value likelihood ratio

Height  5th percentile or Michaelis 13.5 53.1 (42.7–63.2) 92.0 (89.7–93.9) 47.7 (38.0–57.5) 6.6 (4.8–9.0)
transverse  10th percentile
Height  5th percentile or 16.6 46.9 (36.8–57.3) 87.6 (84.9–89.9) 34.3 (26.3–43.0) 3.8 (2.8–5.0)
intertrochanteric diameter  10th percentile
Height  5th percentile or 12.8 42.9 (32.9–53.2) 91.4 (89.1–93.4) 40.8 (31.2–50.9) 5.0 (3.5–6.9)
foot length  10th percentile
Height  5th percentile or 11.6 43.9 (33.9–54.3) 92.8 (90.6–94.6) 45.7 (35.4–56.3) 6.1 (4.3–8.6)
symphysis-fundus height  90th percentile
Height  5th percentile or 19.7 64.3 (54.0–73.7) 86.5 (83.7–88.9) 39.6 (32.0–47.7) 4.7 (3.7–6.0)
Michaelis transverse  10th percentile or
symphysis-fundus height  90th percentile

Values are given as % (95% CI).

ª 2007 The Authors Journal compilation ª RCOG 2007 BJOG An International Journal of Obstetrics and Gynaecology 633
 Rozenholc et al.

dystocia were comparable to those usually reported in sub-

Saharan Africa.4,8

35.4 (22.2–50.5) 22.0 (11.5–36.0) 98.1 (96.3–99.2) 99.0 (97.0–99.8) 68.0 (46.5–85.0) 78.6 (49.2–95.3) 18.5 (8.5–39.8) 21.3 (6.6–68.6)
7.6 (4.3–13.4)
8.2 (4.7–14.0)
A meta-analysis of the value of maternal height as a risk

Positive likelihood ratio

Phase 2
factor for dystocia showed the 5th percentile to have a sensi-
tivity of 21.0%, a specificity of 95% and a positive likelihood
ratio of 4.2.9 The prediction by the maternal height obtained
in our study was slightly higher. In our setting, clinicians may

50.0 (35.2–64.8) 42.0 (28.2–56.8) 91.4 (88.3–93.9) 94.5 (91.2–96.8) 40.0 (27.6–53.5) 56.8 (39.5–72.9) 5.8 (3.8–8.7)
58.3 (43.2–72.4) 48.0 (33.7–62.6) 90.5 (87.2–93.1) 94.1 (90.8–96.5) 41.2 (29.4–53.8) 58.5 (42.1–73.7) 6.1 (4.1–8.8)
have had a tendency to diagnose dystocia excessively when
Phase 1

caring for short women.

The prediction by the Michaelis transverse that we found
was very close to that by Liselele et al.,17 who reported a
sensitivity of 42.9%, a specificity of 91.1% and a positive
likelihood ratio of 4.8 for the 10th percentile. Likewise, the
Phase 2
Positive predictive value

sensitivity, specificity and likelihood ratio of the combina-

tion of a maternal height less than or equal to the 5th per-
centile or a Michaelis transverse less than or equal to the 10th
percentile obtained here were similar to those obtained by
Liselele (respectively, 52.4%, 87.0% and 4.0). The positive
Phase 1

predictive value was higher in our study, because of a higher

Table 4. Comparison of prediction by measurements performed during phase 1 (several observers) or phase 2 (one observer)

percentage of dystocia and possibly because of the overesti-

mation of the positive predictive value of the maternal
height. The high specificity would result in a limited per-
centage of unnecessary referrals, minimising the burden on
Phase 2

district hospitals.
Significant variations in the distribution of the Michaelis
Specificity

transverse measurement between phase 1 and 2 question the

reproducibility of this measurement. Agreement was not
evaluated in this study. Nevertheless, this measurement had
Phase 1

similar prediction when performed by several observers or by

a single observer, provided that the cutoff was determined as
a percentile of the distribution in each phase, instead of a sin-
gle value in centimetres in the whole population. The observ-
ers performing the measurements during phase 1 were
Phase 2

instructed to measure the distance between the middle points

of the two depressions defining the Michaelis transverse. Dur-
Sensitivity

ing phase 2, the observer, who was not instructed specifically,

measured the distance between the lateral edges of the depres-
sion. This difference in the measurement technique likely
Phase 1

corresponds to the overestimation noticed during phase 2.

This emphasises the need for standardisation of the measure-
ment technique, which would allow the determination of
a single cutoff for a population measured by several observers,
Michaelis transverse  10th percentile
Michaelis transverse  10th percentile

as in phase 1.
The measurements of the intertrochanteric and antero-pos-
Values are given as % (95% CI).

terior diameters, and the foot length either separately or in any

combination did not result in improved prediction. The sym-
Height  5th percentile or
Height  5th percentile

physis-fundus height, which was the only measurement related

to the fetal component of dystocia, was also unhelpful.
The value in centimetre corresponding to the 10th per-
centile of the Michaelis transverse in phase 1 was the same in
our population than in the study by Liselele et al.17 in Zaire.
This suggested that, as for maternal height (less than or
equal to 150 cm), a cutoff for the Michaelis transverse (less

634 ª 2007 The Authors Journal compilation ª RCOG 2007 BJOG An International Journal of Obstetrics and Gynaecology

than or equal to 9.0 cm) may be applicable in different
populations.
The size of the Michaelis transverse is associated with the
transverse pelvic capacity. Among black women, the propor-
tion of anthropoid pelvises, characterised by a reduction in
the pelvic transverse diameters, is twice that in white
women.23 Therefore, in black women, the transverse pelvic
capacity may be more critical during labour, and the Michae-
lis transverse may be more associated with dystocia than in
white women. Anyhow, most white women live in the West-
ern world where caesarean section during labour is readily
available, limiting the interest of screening for dystocia. In
Chinese women, the proportion of anthropoid pelvises is
intermediate between white and black women,24 and this
method could be useful.
The effects of the implementation of this screening method
should be evaluated in a randomised controlled trial.
Conclusion
This simple antenatal screening method could be imple-
mented in centres without operative capability, for timely
referral of women at risk, for delivery in district hospitals. It
must be underlined that all referred women should be allowed
a trial of labour, as this is a screening and not a diagnostic test.
Acknowledgements
We thank Professor Fritz Baumann, for his help throughout
the study and Prof. Guillaume Atchou for his support during
the visits in Cameroon.
Funding
Fondation Suisse pour la Santé Mondiale, Thônex,
Switzerland. j
References
1 AbouZahr C, Wardlaw T. Maternal mortality in 2000: estimates devel-
oped by WHO, UNICEF and UNFPA. Geneva: World health Organi-
zation, 2004 [http://www.who.int/reproductivehealth/publications/
maternal_mortality_2000/]. Accessed 8 July 2005.
2 Lawn JE, Cousens S, Zupan J. 4 million neonatal deaths: when? Where?
Why? Lancet 2005;365:891–900.
3 Hartfield VJ. Maternal mortality in Nigeria compared with earlier inter-
national experience. Int J Gynaecol Obstet 1980;18:70–5.
4 Mati JK, Aggarwal VP, Sanghvi HC, Lucas S, Corkhill R. The Nairobi
birth survey. IV. Early perinatal mortality rate. J Obstet Gynaecol East
Cent Africa 1983;2:129–33.
ª 2007 The Authors Journal compilation ª RCOG 2007 BJOG An International Journal of Obstetrics and Gynaecology
Pelvimetry and height to predict dystocia
5 Tsu VD. Maternal height and age: risk factors for cephalopelvic dispro-
portion in Zimbabwe. Int J Epidemiol 1992;21:941–6.
6 Thaddeus S, Maine D. Too far to walk: maternal mortality in context.
Soc Sci Med 1994;38:1091–110.
7 The Kasongo Project Team. Antenatal screening for fetopelvic dysto-
cias. A cost-effectiveness approach to the choice of simple indi-
cators for use by auxiliary personnel. J Trop Med Hyg 1984;87:
173–83.
8 Harrison KA. Child-bearing, health and social priorities: a survey of 22
774 consecutive hospital births in Zaria, Northern Nigeria. Br J Obstet
Gynaecol 1985;92(Suppl 5):1–119.
9 Dujardin B, Van Cutsem R, Lambrechts T. The value of maternal height
as a risk factor of dystocia: a meta-analysis. Trop Med Int Health 1996;
1:510–21.
10 Hughes AB, Jenkins DA, Newcombe RG, Pearson JF. Symphysis-fundus
height, maternal height, labour pattern, and mode of delivery. Am J
Obstet Gynecol 1987;156:644–8.
11 Mahmood TA, Campbell DM, Wilson AW. Maternal height, shoe size,
and outcome of labour in white primigravidas: a prospective anthro-
pometric study. BMJ 1988;297:515–17.
12 Frame S, Moore J, Peters A, Hall D. Maternal height and shoe size as
predictors of pelvic disproportion: an assessment. Br J Obstet Gynaecol
1985;92:1239–45.
13 Suonio S, Saarikoski S, Raty E, Vohlonen I. Clinical assessment of the
pelvic cavity and outlet. Arch Gynecol 1986;239:11–16.
14 Adinma JI, Agbai AO, Anolue FC. Relevance of clinical pelvimetry to
obstetric practice in developing countries. West Afr J Med 1997;16:
40–3.
15 Burgess HA. Anthropometric measures as a predictor of cephalopelvic
disproportion. Trop Doct 1997;27:135–8.
16 Hanzal E, Kainz C, Hoffmann G, Deutinger J. An analysis of the pre-
diction of cephalopelvic disproportion. Arch Gynecol Obstet 1993;253:
161–6.
17 Liselele HB, Boulvain M, Tshibangu KC, Meuris S. Maternal height
and external pelvimetry to predict cephalopelvic disproportion in
nulliparous African women: a cohort study. BJOG 2000;107:
947–52.
18 Kwawukume EY, Ghosh TS, Wilson JB. Maternal height as a
predictor of vaginal delivery. Int J Gynaecol Obstet 1993;41:
27–30.
19 Ould El Joud D, Bouvier-Colle MH. Dystocia: a study of its frequency
and risk factors in seven cities of west Africa. Int J Gynaecol Obstet
2001;74:171–8.
20 Sokal D, Sawadogo L, Adjibade A. Short stature and cephalopelvic
disproportion in Burkina Faso, West Africa. Operations Research Team.
Int J Gynaecol Obstet 1991;35:347–50.
21 van Roosmalen J, Brand R. Maternal height and the outcome of labour
in rural Tanzania. Int J Gynaecol Obstet 1992;37:169–77.
22 Dumont A, de Bernis L, Bouvier-Colle MH, Breart G. Caesarean section
rate for maternal indication in sub-Saharan Africa: a systematic review.
Lancet 2001;358:1328–33.
23 Torpin R. Roentgenpelvimetric measurements of 3,604 female
pelves, white, Negro, and Mexican, compared with direct measure-
ments of Todd anatomic collection. Am J Obstet Gynecol 1951;62:
279–93.
24 Chen HY, Chen YP, Lee LS, Huang SC. Pelvimetry of Chinese females
with special reference to pelvic type and maternal height. Int Surg
1982;67:57–62.
635