A cross sectional study evaluating screening using maternal anthropometric measurements for outcomes of childbirth in Ugandan mothers at term

Background: Birth related newborn and maternal mortality/morbidity remains high in most of sub-Saharan Africa compared to the rest of the world. In this low income region there is a need for valid, low cost, easy to use mass screening tests. This study looked at the screening value of maternal: height, weight and pelvis height, for assessing the outcomes of parturition in Ugandan mothers at term. Methods: This was a multi site cross-sectional study on mothers with singleton pregnancies in labour at various hospitals in different parts of Uganda. A summary of the details of the pregnancy, maternal height, weight and the delivery record were captured and analysed to generate descriptive and inferential (multilevel logistic regression analysis) and diagnostic (Receiver Operator Curve analysis) statistics. Results: We recruited 1146 mothers from all the study sites during the study period of whom 987 (86.13%) had normal deliveries and healthy babies. Mothers with adverse outcomes included 107 mothers that had caesarean section and 52 mothers who had vaginal deliveries with foetal Apgar score of ≤7 at 5 min of whom 11 had fresh still births. Maternal height (Adj OR 0.97, 95% CI 0.94-1.00) and maternal pelvis height (Adj OR 0.73, 95% CI 0.61-0.86) were significantly associated with adverse pregnancy outcomes. The combination of maternal: height (<150 cm), weight (>55.7 kg) and pelvis height (>8.95 cm) had the best diagnostic value with a combined area under the curve of 0.60. Conclusions: It was observed that an increase in either maternal pelvis height or maternal height was associated with a significant reduction in adverse pregnancy outcomes. The cut off values of all three evaluated maternal anthropometric measurements were of low test accuracy as screening tests even when used together. Further research is needed to develop low cost screening tools for use in low income settings. This was a multi site cross-sectional study carried out at purposively selected health facilities from different geographical regions and levels of Ugandas health care system. We made this selection of hospitals to cater for geographical distribution of known major ethinic groupings within Uganda [15, 16]. First we included Mulago hospital the National Tertiary Care and Referal Teaching hospital in cental Uganda. Second Arua regional referal hospital in North western Uganda and Anaka (North cental Uganda). Third we included private not for profit level IV health center hospitals, Kumi (Eastern Uganda), St. Josephs Kitgum (extreme North central Uganda), Kilembe (Western Uganda), and Nyakibale (Southwestern Uganda). Each of these hospitals was capable of handling cesarean delivery with a minimum of 150 mothers delivered per month. We included Mothers in active labour, having pregnancies of ≥37 weeks estimated using the symphysio-fundal height, came to the above listed sites during the 14 months of the study starting January 2013, gave an informed consent, having a live singleton baby with no obvious congenital abnormalities after childbirth, and baby with cephalic presentation before childbirth. We excluded mothers with history of previous cesarean section and those with medical conditions like hypertension or diabetes or carrying a multiple pregnancy. The mothers were recruited consecutively on a daily basis from each of the centers by a team of previously trained midwifes. The target sample size of 933 mothers was calculated to give a minimum AUC of 0.555, obtained from our pilot data, compared to the NULL (0.500) for equivalent sized sub-groups, a β = 0.9 and α = 0.05 using the calculator for sample size calculation based on Area under the curve available online at (https://www.statstodo.com/SSiz2ROCs_Pgm.php). This was inflated by a design effect of 1.2 for the 7 sites to give a final total sample size of 1,120 participants [17]. For each mother we collected data on age in years, height in centimeters and weight in kilograms measured using the available hospital equipment [18], gravidity (current pregnancy + all previous pregnancies), fetal presentation, head descent in terms of number of 5ths engaged and symphysio-fundal height in centimeters, a proxy for gestational age on clinical examination to the nearest 0.1 cm. After delivery we also collected data on the mode of delivery, sex of the baby, APGAR scores, presence of caput, meconium staining, whether augmentation or stimulation of labour was done, birth weight in kilograms and whether the mother had a cesarean section at the facility. For each mother the pelvis height in centimeters was measured twice, at the time of admission by the attendant midwife, using the anterior superior illiac spine (ASIS) and the Symphysis pubis bony body landmarks using a pair of transparent rigid rulers placed at right angles to each other, as demonstrated in Figure 1 (see lines AB and BC). The average of these two measurements was used for analysis. We trained the midwifes at each site on how to measure pelvis height and complete the study questionnaire at the start of the study. There were additional refresher training sessions during the site visits by IGM. The training, which in addition placed emphasis on proper fetal monitoring, partogram use and obtaining informed consent, was accompanied with supervised hands on practice for each midwife until they demonstrated competency to take the measurements. At each site the incharge was recruited as the site supervisor for the duration of the study to ensure that all forms were completed, measurements done properly and to provide training support for midwifes that were on duty after the 1 day training sessions. Data was entered into Epidata version 3.2 (Epidata association, Denmark) and exported to STATA 12 (StataCorp LP, Texas, USA) for eventual analysis. Paired correlation coefficients were determined for each of the study variables highlighting significant correlations. Logistic regression was used to identify predictors of unfavourable outcomes of labour that included: delivery by cesarean section, vaginal delivery of a baby with an APGAR score of ≤7 at 5 min [19] or mother being referred. To obtain predicted probabilities, all the variables with the optimal cut-off that were significant at the 0.20 level in the univariate analysis were included in the subsequent modeling [20]. To cater for the study design and variability in recruitment numbers from the different sites multilevel logistic regression using the gllamm function in STATA was used to calculate the odds ratios with iterations nipped at 60 for final modeling [21]. The sensitivity and specificity of each anthropometric measurement was computed, curves plotted, and optimal cut-off values with out cost considerations identified for each variable using Receiver Operator Characteristic (ROC) using MedCalc Statistical Software version 13.0.6 (MedCalc Software bvba, Ostend, Belgium; http://www.medcalc.org; 2014). All simulations for the ROC analysis were done up to 1000 times using the bootstrapping procedures. During analysis, cut off values were generated as part of the ROC analysis with added comparisons with values from literature for maternal weight, maternal pelvis height [22] and maternal height [20, 23]. No distinction was made between normal vaginal and instrumental vaginal deliveries that were grouped together, during analysis, as a single outcome measure, vaginal delivery. The group with favourable outcome was defined as a vaginal delivery with a healthy live infant. Any deviation from having a vaginal delivery resulting into a healthy live infant was defined as an unfavourable outcome. Mothers’ with multiple outcomes were counted only once and categorised as having had an unfavourable outcome in the event that this was observed. The T test was used to obtain the level of significance for differences in the means of the study variables for the normal vaginal delivery with normal baby group and unfavourable outcomes group. Any observation found with a missing value was dropped from analysis. A P < 0.05 was considered significant for all tests. Ethical approval was obtained from Makerere University School of Biomedical Sciences IRB and Uganda National of Science and Technology. At each site we verbally obtained the consent of the participating nursing staff to be part of the study and offered an equivalent of one United States dollars compensation for each birth record filled to completion. The participating mothers, each gave informed consent after explanation of the study and signed the informed consent form to participate in the study on admission. Informed consent was obtained by the attending midwife from mothers. In Uganda, the Uganda National Council of Science and Technology (UNCST) allows emancipated minors to consent to participate in research as long as they have been informed about the risks involved [24]. Authors considered minors as emancipated adults and all mothers were free to consult their spouses or next of kins since the study required one to provide contact information as part of the consent process. With the exception of measuring maternal pelvis height there were no other procedures or modifications made to the current birthing practice at any of the participating sites. Refusal to participate in the study did not result in a mother being denied access to health care or required services at the participating facility. No identifier marks of personal information was used in the analysis and subsequent reporting of the study results.