A C TA Obstetricia et Gynecologica

AOGS M A I N R E S E A R C H A R T I C L E

Anthropometric measurements as predictors of cephalopelvic disproportion

SANTOSH J. BENJAMIN1 , ANJALI B. DANIEL2 , ASHA KAMATH2 & VANI RAMKUMAR3
Department of Obstetrics and Gynaecology, Christian Medical College, Vellore, Tamilnadu, India, 2 Department of Community Medicine and 3 Department of Obstetrics and Gynaecology, Kasturba Medical College, Manipal, Karnataka, India
1

Key words Anthropometric measurements, cephalopelvic disproportion, fetal weight estimates Correspondence Dr. Santosh Benjamin, Department of Obstetrics & Gynaecology, Christian Medical College, Vellore 632004, Tamilnadu, India.E-mail: toffeetripper@gmail.com Conict of interest The authors have stated explicitly that there are no conicts of interest in connection with this article. Please cite this article as: Benjamin SJ, Daniel AB, Kamath A, Ramkumar V. Anthropometric measurements as predictors of cephalopelvic disproportion. Acta Obstet Gynecol Scand 2012; 91:122127. Received: 15 October 2010 Accepted: 28 August 2011

Abstract Objective . We assessed the efcacy of maternal anthropometric measurements and clinical estimates of fetal weight in isolation and in combination as predictors of cephalopelvic disproportion (CPD). Design. Prospective cohort study. Setting. Tertiary care teaching hospital, two afliated hospitals with facilities for conducting cesarean delivery and seven afliated primary care facilities with no operation theaters. Sample. Primigravidae over 37 weeks gestation attending these facilities during a 20-month period with a singleton pregnancy in vertex presentation. Methods . Several anthropometric measurements were taken in 249 primigravidae. Fetal weight was estimated. Differences in these measurements between the vaginal delivery and CPD groups were analyzed. The validity of these measurements in predicting CPD was analyzed by plotting receiver operating characteristic curves and by logistic regression analysis. Main outcome measure. Mode of delivery. Results . Maternal height, foot size, inter-trochanteric diameter and bis-acromial diameter showed the highest positive predictive values for CPD. Combining some maternal measurements with estimates of fetal weight increased predictive values modestly, which are likely to be greater if the estimates of fetal weight are close to the actual birth weight. Based on multivariate analysis the risk factors for CPD in our population were foot length 23cm, inter-trochanteric diameter 30cm and estimated fetal weight 3 000g. Conclusions . Maternal anthropometric measurements can predict CPD to some extent. Combining maternal measurements with clinical estimates of fetal weight only enhances the predictive value to a relatively modest degree (positive predictive value 24%).
Abbreviations: CPD, cephalopelvic disproportion; ROC, receiver operating characteristic; PPV, positive predictive value.

DOI: 10.1111/j.1600-0412.2011.01267.x

Introduction
Maternal deaths remain a major problem in some developing regions of the world such as sub-Saharan Africa and south Asia (1,2). Cephalopelvic disproportion (CPD) is in such circumstances an important cause of maternal and perinatal mortality and is also associated with considerable morbidity for both the mother and the baby (14). Perineal tears, postpartum hemorrhage, and obstetric stulae in the mother, and birth asphyxia and birth trauma in the newborn are all associated with CPD (46). Timed optimally, a cesarean delivery for CPD is best for the mother as well as her fetus; to facilitate this it is imperative that

CPD is diagnosed sufciently early. The consequences of late detection are particularly grave in the developing world where the mother may go into labor in a setting where facilities for performing cesarean section are inadequate (6). In such situations, it is vital that women at potential risk of CPD are identied prior to the onset of labor to facilitate referral to a center where a cesarean delivery can be performed. Measurement of maternal height has been used as a simple means to identify women at risk of CPD, as it is assumed that the shorter the mother, the greater the likelihood of CPD (710). However, maternal height in isolation has limited value for predicting CPD risk (11) and combining anthropometric measurements may increase the likelihood of
C 2011 The Authors 2011 Nordic Federation of Societies of Obstetrics and Gynecology 91 (2012) 122127

122

Acta Obstetricia et Gynecologica Scandinavica

S. J. Benjamin et al.

Anthropometric predictors of CPD

predicting CPD (12). Since anthropometry and stature of women of different ethnic origins vary, it is important to identify the most sensitive predictors of CPD for a particular population. We assessed the efcacy of using maternal height, foot size and other anthropometric measurements as predictors of CPD in a cohort of women from southwest India and included estimated fetal weight as an additional predictor.

Material and methods

This prospective cohort study was carried out at a tertiary care teaching hospital (Kasturba Hospital), two afliated hospitals with facilities for conducting cesarean delivery, and seven afliated primary care facilities with no operation theaters. Primigravidae of >37 weeks gestation who attended any of these facilities over a 20-month period with a singleton pregnancy in vertex presentation were eligible for inclusion in the study. The women were recruited on each day from the center where the principal investigator (S.J.B.) was on duty; the sample accounts for approximately 15% of the women delivering at these facilities. The height, bis-acromial, inter-cristal, inter-trochanteric diameters, foot length, vertical and transverse diameters of the Michaelis rhomboid, symphysio-fundal height and abdominal girth were measured for each woman (Figure 1). Height was recorded on a stadiometer with an accuracy of 0.5cm; a modied Harpenden anthropometer with an

accuracy of 0.1cm and a tape measure were used for other measurements. With the patient lying supine with knees exed, fundal height was marked after correcting uterine dextro-rotation. Then, with the patients knees straightened, the distance from the symphysis to the point marked (symphysio-fundal height) was noted (13). The abdominal girth was measured with a measuring tape at the level of the umbilicus. Fetal weight was estimated both as the product of the symphysio-fundal height (cm) and the abdominal girth (cm) and expressed in grams (1315) and using Johnsons formula (16) which is calculated from the symphysio-fundal height in centimeters a constant according to level of fetal head engagement 155. Mode of delivery was recorded as normal vaginal, instrumental vaginal (forceps, vacuum extraction), cesarean delivery for CPD or cesarean delivery for other indications. The former was dened as emergency cesarean delivery performed for failure of descent of the fetal head or failure of progress following a trial of labor, where there was no change in the cervical dilation for 4 hours in the active phase of labor following rupture of the membranes despite adequate uterine contractions (three contractions/ 10 minutes, each >45 seconds) in the presence of a well exed head. Fourteen patients who were short and considered to have a large fetus had elective cesarean delivery and were excluded from analysis. Intra-observer reproducibility alone was evaluated (only one investigator). Forty women had all anthropometric measurements repeated on two separate occasions without the investigator having access to the values of the rst measurement session.

Statistical analysis
The Statistical Package for Social Sciences (SPSS version 15.0 for Windows) was used. Reproducibility was estimated by computing the coefcient of variation. The sensitivity and specicity of each anthropometric measurement was computed and receiver operating characteristic (ROC) curves plotted. Stepwise logistic regression was used to identify predictors of CPD. The optimal cut-off for each variable was obtained using the ROC analysis. To obtain predicted probabilities, all the variables with the optimal cut-off that were signicant at the 0.2 level in the univariate analysis were included for the logistic regression. For analysis, normal vaginal and instrumental deliveries were grouped together as a single outcome measure for comparison with the CPD group. Analysis of variance followed by post-hoc Tukeys test was used to compare differences between the vaginal delivery and CPD groups. Spearmans correlation coefcients were computed to compare the estimates of fetal weight with actual birth weight. A p-value of <0.05 was considered signicant.

Figure 1. Various anthropometric measurements made in the study. The transverse diameter of the Michaelis sacral rhomboid was measured between the two posterior superior iliac spines. The vertical diameter of the rhomboid was measured between the L5 spine (one space below the L3L4 disc which is in line with the uppermost point of the iliac crest) and the upper limit of the natal cleft.

C 2011 The Authors Acta Obstetricia et Gynecologica Scandinavica

2011 Nordic Federation of Societies of Obstetrics and Gynecology 91 (2012) 122127

123

Anthropometric predictors of CPD

S. J. Benjamin et al.

Table 1. Comparison of maternal anthropometric characteristics and fetal weight between three modes of delivery. Vaginal delivery (n=172) Maternal age (years) Height (cm) Foot length (cm) Bis-acromial diameter (cm) Inter-crestal diameter (cm) Inter-trochanteric diameter (cm) Michaelis rhomboidtransverse (cm) Michaelis rhomboidvertical (cm) Gestational age at delivery (weeks) Symphysio-fundal height (cm) Abdominal girth (cm) Estimated fetal weight (Clinical) (g) Estimated fetal weight (Johnsons formula) (g) Actual birthweight (g)

Cesarean delivery for CPD (n=27) 26.0 3.3 152.1 6.0 22.3 1.1 32.6 2.4 26.9 2.6 28.9 3.7 10.9 1.1 10.1 1.2 39.1 0.9 34.9 2.8 90.4 7.4 3172 456 3556 410 3095 462

Cesarean delivery for other indications (n=50) 25.7 4.0 154.6 5.0 23.0 1.1 33.6 1.8 27.3 1.6 30.4 1.6 10.6 1.2 10.4 0.8 39.1 0.9 34.5 2.3 90.2 5.9 3117 357 3488 363 2971 385

p-value 0.188 <0.001 0.002 0.014 0.628 0.086 0.648 0.369 0.020 0.023 0.544 0.081 0.023 0.005

24.9 3.4 157.1 6.6 23.2 1.1 33.8 1.9 27.2 1.9 30.6 1.8 10.7 1.1 10.4 0.9 38.7 1.1 33.7 2.6 89.3 6.5 3016 398 3365 399 2861 353

MeanSD. Based on ANOVA.

The study was conducted after review and approval by the departmental review panel.

Results
We included 249 primigravidae. All the anthropometric measurements were reproducible, with the mean differences between two readings ranging from 2mm for foot length to 19mm for height; these differences were not signicant. All the coefcients of variation were below 20%. The anthropometric measurements and estimated fetal weights by mode of delivery are shown in Table 1. Several measurements were lower in mothers who had cesarean delivery for CPD when compared with mothers who delivered vaginally.

Signicant differences were noted for height, foot length and bis-acromial diameter of mothers with and without CPD. Based on the ROC curves for each measurement, the cutoff threshold values with the highest sensitivity and specicity were identied and are shown in Table 2. The areas under the curves for the maternal height, foot length, intertrochanteric and bis-acromial diameters were 0.690, 0.688, 0.647 and 0.633 respectively. With the chosen cut-off threshold values of 155.5, 23, 30 and 33.4cm, respectively, for height, foot length, inter-trochanteric diameter and bis-acromial diameter, the positive predictive values ranged between 14.4 and 18.6%. The predictive value increased modestly when estimates of the birthweight were combined with maternal height and foot length (Table 3). Adding bis-acromial and inter-trochanteric

Table 2. Maternal anthropometric measurements, estimated fetal weights and actual fetal weights as predictors of CPD Cut-off value Height (cm) Foot length (cm) Bis-acromial diameter (cm) Inter-crestal diameter (cm) Inter-trochanteric diameter (cm) Michaelis rhomboidtransverse (cm) Michaelis rhomboidvertical (cm) Symphysio-fundal height (cm) Abdominal girth (cm) Estimated fetal weight (clinical) (g) Estimated fetal weight (Johnsons formula) (g) Actual birthweight (g) PPV, positive predictive value. 155.5 23.0 33.4 27.5 30.0 10.4 10.1 33.3 88.3 2 931 3 294 2 894 Sensitivity percentage 70.4 77.8 55.6 55.6 63.0 55.6 59.3 70.5 70.4 77.8 70.4 66.7 Specicity percentage 52.1 58.6 89.2 42.8 65.3 61.7 46.8 42.3 47.7 41.4 42.3 50.0 PPV percentage 15.4 18.6 14.4 10.6 18.1 15.0 11.9 12.9 14.1 13.9 12.9 14.0

124

Acta Obstetricia et Gynecologica Scandinavica

C 2011 The Authors 2011 Nordic Federation of Societies of Obstetrics and Gynecology 91 (2012) 122127

S. J. Benjamin et al.

Anthropometric predictors of CPD

Table 3. Validity of combining different maternal anthropometric measurements and fetal weight for prediction of CPD Sensitivity percentage Height+foot length Height+estimated fetal weight (Johnsons formula) Height+estimated fetal weight (clinical) Foot length+estimated fetal weight (Johnsons formula) Foot length+estimated fetal weight (clinical) Height+ foot length+estimated fetal weight (Johnsons formula) Height+ foot length+estimated fetal weight (clinical) PPV, positive predictive value. 63.4 81.5 Specicity percentage 61.7 60.4 PPV percentage 16.7 20.0

5000

Actual Birth Weight (g)

4000

3000

81.5

61.7

20.6

2000

74.1

68.5

22.2
1000 2000 3000 4000 5000

74.1

68.5

22.2

Clinical Estimate of Fetal Weight (g)

70.4

56.8

16.5

Figure 2. Scatter diagram showing the relation between actual birthweight and clinical estimate of fetal weight. Although there is a positive correlation, the agreement is poor.

70.4

73.0

24.1

diameters did not increase the predictive value appreciably (data not shown in Table 3). If the estimate of fetal weight was equal to the true birthweight and combined with maternal height and foot length, the positive predictive value rose to 34.9. Multivariate logistic regression analysis identied three signicant predictors of CPD; inter-trochanteric diameter 30cm (OR: 2.8; 95%CI: 1.2,6.9), foot length 23cm (OR: 4.0; 95%CI: 1.5,10.7) and clinically estimated fetal weight 3 000g (OR: 3.4; 95%CI: 1.3,9.3). The predicted probability for CPD if all three measurements were outside these thresholds was 32.3%. There was a positive correlation between the estimates of fetal weight and actual birthweight (Spearmans rank correlation coefcient =0.60, p<0.001) but agreement between them was poor (Figure 2).

Discussion
If anthropometric measurements are to be used to identify women at risk for CPD, the general reproducibility of these measurements must be conrmed. Intra-observer reproducibility may, as here in an experimental situation, be satisfactory but this may not be true in a clinical setting or between observers. However, these measurements can be made reliably with inexpensive equipment provided the measurements are sufciently sensitive and specic. In this study only patients who underwent cesarean deliveries were considered
C 2011 The Authors Acta Obstetricia et Gynecologica Scandinavica

as having CPD. The present study differs in this regard from that of Liselele et al. (9) who included instrumental vaginal deliveries as instances of CPD. The present study shows that in mothers who had instrumental vaginal deliveries, most of the anthropometric characteristics were close to those for the normal vaginal delivery group (Table 4). It is well established that the height of the mother is correlated to the size of the pelvis and several studies have demonstrated that mothers with CPD are shorter than those who have normal vaginal deliveries (712,1620) as conrmed here. There is, however, no consensus on the height below which CPD is likely to occur. Several studies have used a cutoff value of 150cm for height to predict CPD. However, this will not be appropriate for all ethnic populations, and the nutritional status of the mother and genetic factors would also determine fetal size (6). Another method of deciding a cut-off limit for an anthropometric measurement to predict CPD has been to identify the 10th percentile of the measurement for the study population and to use that value (9). Again, this is an arbitrary value and does not take into consideration fetal size, giving a low sensitivity for predicting CPD (9). Using ROC curves to decide the cut-off limits was recommended by Ferguson et al. (21) who emphasized the need for more robust statistical methods for determining predictive levels of anthropometric and radiological measurements. The anthropometric measurements that showed the highest positive predictive values for CPD were height, foot length, inter-trochanteric diameter and bis-acromial diameter. Since taking these measurements is not intrusive (unlike measuring dimensions of the pelvis), they should be acceptable to women even in clinic settings with little privacy or where touching is considered less appropriate.

2011 Nordic Federation of Societies of Obstetrics and Gynecology 91 (2012) 122127

125

Anthropometric predictors of CPD

S. J. Benjamin et al.

Table 4. Maternal anthropometric characteristics and fetal weight in normal vaginal deliveries, instrumental vaginal deliveries and cesarean deliveries for CPD. Normal vaginal delivery, n=146 157.0 6.9 23.2 1.1 33.9 1.8 27.1 1.8 30.5 1.8 10.7 1.1 10.4 0.9 33.6 2.5 89.0 6.5 2840 344

Variables (cm) Height Foot length Bis-acromial diameter Inter-crestal diameter Inter-trochanteric diameter Michaelis rhomboidtransverse Michaelis rhomboidvertical Symphysio-fundal height Abdominal girth Birthweight (g)

Instrumental vaginal delivery, n=26 157.3 4.9 23.1 1.0 33.4 2.6 28.1 2.1 31.0 1.9 10.7 1.2 10.1 0.9 34.5 2.7 91.0 6.2 2981 379

Cesarean delivery for CPD, n=27 152.1 6.0 22.3 1.1 32.6 2.4 26.9 2.6 28.9 3.7 10.9 1.1 10.1 1.2 35.9 2.8 90.4 7.4 3095 462

p <0.001 <0.002 <0.01 0.693 0.086 0.644 0.373 0.020 0.533 <0.005

Values of the group differed from those of the cesarean for CPD group at p<0.001. Values of the group differed from those of the cesarean for CPD group at p<0.05. Based on ANOVA. Using the post -hoc Tukeys test.

Anthropometric measurements cannot estimate pelvic dimensions precisely (22) but they do help in drawing attention to the possibility of CPD. The area under the curves of the ROC curves plotted for the most predictive anthropometric data in this study was comparable to those reported by Awonuga et al. (22), who compared them to actual pelvic size. Reliable antenatal estimation of fetal weight should help in diagnosing CPD more accurately. We used two different methods of estimating the fetal weight in a primary care setting where ultrasound is not available, but both methods showed poor agreement with actual birthweight as reported by earlier investigators (23). Attempts have been made before to combine more than one anthropometric measurement in the hope that the predictive value of the combined measurements would be greater (9,11,12). In this study combining maternal height, foot length and clinical estimate of fetal weight increased the predictive value to 24.1 from 15.4 when maternal height alone was used as a predictor for CPD. Cephalopelvic disproportion can never be predicted with certainty (21) as it depends on several maternal and fetal factors other than just the dimensions of the maternal pelvis and fetal size, such as the degree of stretch in the maternal pelvis, the capacity of the fetal head to mould, the position adopted by the woman during labor and fetal position. Yet its recognition and indication of possible CPD is necessary at every birth to prevent the serious complications associated with undiagnosed disproportion. Finally, although the results of this study are of primary importance to obstetricians practising in this region, they may also be of relevance to obstetricians in several parts of the world to which women from this region have emigrated.

Funding
No specic funding. References
1. Abou-Zahr C, Wardlaw T, Stanton C, Hill K. Maternal mortality. World Health Stat Q. 1996;42:7787. 2. Ronsmans C, Graham WJ. Maternal Survival 1: Maternal mortality: who, when, where, and why. Lancet. 2006;368:1189200. 3. Nkata M. Maternal mortality due to obstructed labor. Int J Gynecol Obstet. 1997;57:656. 4. Neilson JP, Lavender T, Quenby S, Wray S. Obstructed labour. Br Med Bull. 2003;67:191204. 5. Lawn JE, Cousens S, Zupan J, for the Lancet Neonatal Survival Steering Team. Neonatal survival 1: 4 million neonatal deaths: When? Where? Why? Lancet. 2005;365:891900. 6. Lee ACC, Darmstadt GL, Kharty SK, LeClerq SC, Shresta SR, Christian P. Maternal-fetal disproportion and birth asphyxia in rural Sarlahi, Nepal. Arch Pediatr Adolesc Med. 2009;163:61623. 7. van Roosmalen J, Brand R. Maternal height and outcome of labor in rural Tanzania. Int J Gynecol Obstet. 1992;37:16977. 8. Burgess HA. Anthropometric measures as a predictor of cephalopelvic disproportion. Trop Doct. 1997;27:1358. 9. Liselele HB, Boulvain M, Tshibangu KC, Meuris S. Maternal height and external pelvimetry to predict cephalopelvic disproportion in nulliparous African women: a cohort study. Br J Obstet Gynaecol. 2000;107:94752. 10. Hughes AB, Jenkins DA, Newcombe RG, Pearson JF. Symphysis-fundus height, maternal height, labor pattern, and mode of delivery. Am J Obstet Gynecol. 1987;156:6448.
C 2011 The Authors 2011 Nordic Federation of Societies of Obstetrics and Gynecology 91 (2012) 122127

126

Acta Obstetricia et Gynecologica Scandinavica

S. J. Benjamin et al.

Anthropometric predictors of CPD

11. 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;92:123945. 12. Rozenholc AT, Ako SN, Leke RJ, Boulvain M. The diagnostic accuracy of external pelvimetry and maternal height to predict dystocia in nulliparous women: a study in Cameroon. Br J Obstet Gynaecol. 2007;114:6305. 13. Wallin A, Gyllensward A, Westin B. Symphysis-fundus measurement in prediction of fetal growth disturbances. Acta Obstet Gynecol Scand. 1981;60:31723. 14. Dare FO, Ademowore AS, Ifaturoti OO, Nganwuchu A. The value of symphysiofundal height/abdominal girth measurement in predicting fetal weight. Int J Gynecol Obstet. 1990;31:2438. 15. Bhandary AA, Pinto PJ, Shetty AP. Comparative study of various methods of fetal weight estimation at term pregnancy. J Obstet Gynecol Ind. 2004;54:3369. 16. Shittu AS, Kuti O, Orji EO, Makinde NO, Ogunniyi SO, Ayoola OO, et al. Clinical versus sonographic estimation of foetal weight in southwest Nigeria. J Health Popul Nutr. 2007;25:1423. 17. Merchant KM, Villar J, Kestler E. Maternal height and

18.

19.

20.

21. 22.

23.

newborn size relative to risk of intrapartum caesarean delivery and perinatal distress. Br J Obstet Gynaecol. 2001;108:68996. Mahmood TA, Campbell DM, Wilson AW. Maternal height, shoe size, and outcome of labour in white primigravidas: a prospective anthropometric study. Br Med J. 1988;297:5157. Kwawukume EY, Ghosh TS, Wilson JB. Maternal height as a predictor of vaginal delivery. Int J Gynecol Obstet. 1993;41:2730. Sheiner E, Levy A, Katz M, Mazor M. Short stature an independent risk factor for cesarean delivery. Eur J Obstet Gynecol Reprod Biol. 2005;120:1758. Ferguson JE 2nd, Sistrom CL. Can fetal-pelvic disproportion be predicted? Clin Obstet Gynecol. 2000;43:24764. Awonuga AO, Merhi Z, Awonuga MT, Samuels TA, Waller J, Pring D. Anthropometric measurement in the diagnosis of pelvic size: an analysis of maternal height and shoe size and computed tomography pelvimetric data. Arch Gynaecol Obstet. 2007;276:5238. Bhuiyn SN, Begum R. Clinical versus ultrasound estimation of foetal wieight at term. Int J Gynecol Obstet. 2000;70 (Suppl 4):D149.

C 2011 The Authors Acta Obstetricia et Gynecologica Scandinavica

2011 Nordic Federation of Societies of Obstetrics and Gynecology 91 (2012) 122127

127

Copyright of Acta Obstetricia et Gynecologica Scandinavica is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.