Summary of CMDS Domains and Predictive Accuracy
The CMDS has an AUC of 0.73 for the need of an SBA and 0.85 for mortality, which suggests that it can be used for the purpose of identifying pregnant women who require an SBA prior to delivery and those who are at risk of mortality. The CMDS identified women who required further obstetrical evaluation and skilled care by an SBA.
The CMDS is based on 7 risk factors from the literature that require limited medical knowledge to measure (Table 1). The assessment of pregnant women using the CMDS requires few supplies, all of which are readily available at low cost: a thermometer, measuring tape, blood pressure cuff, height chart, protein and glucose urinalysis strips, and a weight scale for adults.
In this study, signs of pre-eclampsia, obstetrical history, and co-existing conditions were highly associated with the need for an SBA (Table 3). Although the domains of age, parity, fundal height, and BMI were not found to be significantly associated with the need for a SBA at delivery, these factors nonetheless represent important indicators of clinical evaluation that should be considered during routine obstetrical assessment.
Signs of pre-eclampsia was the only risk factor significantly associated with mortality (Table 4). The remaining factors were not significant, which may be attributed to the low absolute number of pregnant women who died in-hospital (n = 7). The estimate of the maternal mortality ratio of 1,189 deaths per 100,000 live births is more than double the Nigerian national ratio of 512 deaths per 100,000 live births (95%CI: 447–578).2 However, the confidence intervals overlapped, which prevented us from determining if the MMR in Benue State was significantly higher. Future studies with a larger sample size would allow for narrower confidence intervals for MMR in Benue State, as well as sub-group analyses to further examine risk factors for mortality in Benue State.
The CMDS required an interaction term to account for a positive relationship between age and parity with the need for an SBA. Older primigravida and younger multigravida women were more likely to require an SBA than was predicted by these factors alone in the CMDS. A future model of the CMDS would benefit from stratified age and parity risk indicators for these women. In addition, it is possible that the importance of BMI changes in the CMDS criteria was underestimated, as we did not measure perinatal BMI and only reported excessive 3rd trimester weight gain in the pre-eclampsia domain. BMI changes have been noted to be associated with a plethora of maternal and neonatal complications,36 thus future studies on the CMDS should utilize BMI change rather than limiting evaluation to prenatal BMI. This would provide a more accurate AUC estimate in future iterations of the CMDS.
This is the first study in the literature to develop a comprehensive risk analysis tool for all pregnant women at high-risk for mortality and morbidity in low-resource settings. Maternal mortality and morbidity in low-and-middle income countries is disproportionately high, but with targeted interventions such as the one presented, high-risk pregnant women may be more easily identified and encouraged to seek out an SBA in preparation for delivery. The presence of a skilled birth attendant at delivery is critical to ensure a safe, successful delivery for both mother and child. By applying the CMDS in environments where SBA-seeking behaviours are low, SBA uptake may increase which would effectively work towards reducing the high rates of maternal mortality and morbidity in Nigeria and Benue State.
There is the potential for adaptations of the CMDS to identify women at risk and promote improved care, such as by developing the CMDS into a point of care of community screening mobile application with SMS text messages directed to pregnant women who are at high risk for complications to seek an SBA. Indeed, the use of mobile health technologies is rapidly increasing in the African continent,37–38 and the transformation of a validated scoring system into a mobile application for use by healthcare workers has previously proven successful.39 Due to the widespread use of smartphones in most African countries, mobile health technologies allow the end-user to virtually access evidence-based resources, such as the CMDS presented here, anywhere and at any time. In the future, we hope to make the CMDS available as a mobile application for use by healthcare personnel in Benue State, Nigeria, as well as pilot the use of directed SMS messaging to promote SBA uptake.
The primary limitation of this study is the limited sample size used in the assessment of the CMDS. This was due to logistical and feasibility constraints, which made it unrealistic to conduct data collection for a longer period of time to obtain a larger effective sample. An additional limitation is that the CMDS was assessed using a hospital cohort of patients, who were more likely to require an SBA. Future research is required to prospectively validate the CMDS on a cohort of pregnant women within the community to determine if care is improved and skilled birth attendant utilization is increased.
The CMDS was found to hold moderate-to-good discriminative capability within a clinical setting. The major strength of the CMDS is how it is designed for use in low-resource settings by non-specialist healthcare workers. In areas where majority of obstetrical evaluations are done by non-specialists, the CMDS can be used to supplement their assessments of pregnant women during antenatal visits, thus improving care and promoting SBA uptake, ultimately working to reduce maternal morbidity and mortality. We recommend that the CMDS be validated prospectively within a community setting after improvements based on this study have been included. We propose that the CMDS eventually be incorporated into the standard clinical assessment of pregnant women done by non-specialist health workers in low-resource settings.