The study was designed to investigate the rate of stillbirths in eight referral hospitals in four out of the six geo-political zones of Nigeria, and to identify associated obstetric causes and socio-demographic factors. The results showed a mean stillbirth rate of 39.6 per 1000 births in the 8 hospitals that is slightly less than the national average . However, two General hospitals in the north-central and south-west regions had stillbirth rates in excess of the reported national average. Thus, the results indicate a persisting high rate of stillbirths in referral hospitals without evidence of any substantial regional variation. Indeed, after adjusting for confounding variables, the rates did not differ significantly between the eight hospitals. This suggests that efforts to reduce the rate of stillbirths in the country should focus on all regions of the country, rather than on specific regions.
An assessment of the obstetric and clinical causes of death showed that more stillbirths occurred at the time of delivery (fresh stillbirths) than in the antepartum period (macerated stillbirths). An estimated 22% of the deaths were macerated compared to 47% of deaths that occurred during the intrapartum period. The large number of fresh stillbirths occurring at the time of delivery is worrisome as it indicates that the timing and method of delivery of the babies may not have followed standard obstetric practices. Indeed, the fact that a large proportion of the women were referred on emergencies during the intrapartum period after they had attempted to deliver in non-orthodox places of delivery (homes, traditional birth attendants, etc.) may have accounted for this. Efforts devoted to ensuring that women choose health facilities as their primary source of delivery rather than non-certified outlets would help to overcome this bottleneck.
Similarly, a large proportion of the macerated stillbirths were due to maternal obstetric complications – pregnancy hypertension, fetal growth retardation and maternal infections. This suggests that the high rate of pregnancy related complications among the women which are left unattended during the antenatal period predispose to these stillbirths before the time of delivery.
Among the associated factors for stillbirths, the results identified non-booking for antenatal care, referral from TBAs and faith-based clinics for emergency care, multiparity, previous experience of stillbirths, experience of obstetric complications in previous pregnancy, birth weight less than 1.50 kg, pregnancy less than 37 weeks gestation, and delivery by others means aside from normal vaginal and caesarean section as increasing the likelihood of stillbirths in the hospitals. However, after adjusting for confounders in the full logistic regression model, referral from TBAs and other non-facilities sources of care, multiparity, previous experiences of stillbirths, low birth weight, gestational age at delivery and mode of delivery remained significant predictors of the likelihood of stillbirth.
These results confirm the importance of antenatal care for identifying pregnancy complications and taking prompt action to manage the complications that increase the risk of stillbirths. As preventing stillbirths becomes a more visible goal of the maternal and child health agenda, attention to the important role of antenatal care in reducing stillbirths becomes critical [17, 18]. The results also indicate the need for steps to be taken to ensure that women are educated about the importance of delivering in health facilities rather than at home or in non-orthodox places such as in the homes of traditional birth attendants or in Churches. The current situation where only about 33% of pregnant women are attended to at birth by Skilled Birth Attendants (doctors, nurses and midwives)  is worrisome, and accounts for the high rate of preventable fresh stillbirths reported in this study. Clearly, increasing women’s access to skilled delivery care is an important measure to reduce the high rate of stillbirths in the country.
The finding that multiparity, previous experiences of stillbirths, pre-term delivery and complications in any previous pregnancy are significantly associated with increased odds of stillbirth adds more credence to the importance of antenatal care. These conditions can be managed effectively and prevented when women attend antenatal care regularly, with the effective delivery of quality antenatal care that can reduce the risk of severe complications leading to stillbirths. The present tendency for pregnant women in Nigeria not to receive antenatal care and to turn up only at delivery or at the time of emergency obstetric care is worrisome [19,20,21]. The lack of antenatal care has turned up repeatedly as a risk factor for maternal mortality [22,23,24,25,26], and now as a predictor of the high rate of stillbirths in the country. Clearly, increasing women’s access to evidence-based and quality antenatal care must be given priority policy attention in policies to improve maternal health and reduce the stillbirth rate in the country.
Of interest was the finding in the logistic regression model, which showed that women with instrumental delivery (vacuum extraction and forceps delivery) were four times more likely to experience stillbirths as compared to those delivering vaginally. The caesarean section rate in this study of 11.5% was within the recommended rate of caesarean section recently set by the World Health Organization . However, the increased stillbirth rate in women with instrumental delivery after adjusting for confounding variables suggests either poor skills in instrumental delivery or that women who ought to deliver by caesarean section were being delivered by instrumental delivery with resultant negative outcomes for fetal viability. The study therefore confirms previous reports which suggest the under-utilization of caesarean sections in many sub-Saharan African countries [28, 29] needed to address the burden of maternal ill-health. Although there is evidence to suggest that women are averse to caesarean delivery in Nigeria [30, 31], strategic public health education can help to convince women and care-givers that such mode of delivery is to save the lives of the women and their babies.
The major strength of this study is its multi-centre design and the involvement of multiple hospitals in four out of the six geo-political zones in the country. This has allowed regional comparison of the results, permitting its generalization to the wider Nigerian health systems context given that no statistically significant difference was found in rates between the hospitals when all factors were adjusted. The results are not only useful for preventing stillbirths within the participating health facilities; it also has implications for the development of policies for the improvement of maternal health care and the reduction of the high rate of stillbirths at a health systems level in the country.
By contrast, because this study is facility-based, the reported stillbirth rates and determinants exclude stillbirths that occur in births outside facilities. The major limitation of the study is its retrospective design and the fact that all cases of stillbirths may not have been captured due to poor record keeping in the hospitals. For example, the causes of up to 30% of stillbirths could not be determined due to inadequate record keeping. However, we made specific and rigorous efforts to ensure accurate data collection in the eight participating hospitals. We deliberately chose a retrospective design since this was part of a larger study whose objective was to assess the existing quality of emergency obstetric care in the referral hospitals, with the goal being to design appropriate interventions for addressing the identified bottlenecks in service delivery. A prospectively designed study would have compromised our ability to obtain accurate information on the state of delivery of maternal health care in the hospitals.
The importance of the need to ensure accurate data collection informed our concentration of data collection in the immediate preceding 6 months of the study in order to reduce the potential for data mis-handling. In particular, we used multiple approaches to identify all cases of stillbirths in the hospitals, including examination of records in the maternity wards, record keeping departments and in the delivery suites. Only when these multiple sources of record keeping were in agreement were the cases of stillbirths accepted as true. Also, the record of early neonatal deaths was separate from stillbirths in all the facilities, indicating accurate assessment of stillbirths. Thus, we believe that the results are accurate and represent the current state of available data on stillbirths in the referral hospitals.