Reasons to Retire the Skilled Birth Attendant Indicator and Proposed Alternatives
Monitoring the skilled birth attendant (SBA) indicator for the Millennium Development Goals did not demonstrate the expected outcomes. It appears that SBA did catalyze significant maternal mortality reductions, but not in the low-resource districts and nations where the greatest challenges lie. As Stephen Hodgins suggested in his recent MHTF blog post about the SBA indicator, we need to reassess its global utility.
There are various reasons why SBA is a poor proxy for maternal survival, and we should not expect the correlation between the SBA at delivery indicator and prevention of maternal deaths to be especially strong. For one, only 56% of the maternal emergencies that require medical expertise occur while a woman is in the SBA’s care (during pregnancy and extending this care period through the immediate postpartum week). Therefore, the SBA indicator does not cover the full period of risk of maternal death. Then, there are problems with the validity of the indicator itself as defined and with the quality of the data used to calculate it. The definition of a skilled birth attendant is not standard; it varies greatly by site and between policymakers and implementers. Furthermore, the data comes from women who are asked during household surveys to identify the category of service provider who assisted them in their most recent delivery. Yet many, if not most, women are unaware of the skill level of the health worker(s) who attend their birth and are unable to distinguish a midwife from a nurse or from one of the many lower cadre health workers who staff most primary health care centers (PHCs) in low-resource districts. And many clients would not know which health workers had benefited from a task shifting or life-saving skills training program.
Health system strengthening strategies have established a benchmark for the number of functional Basic and Comprehensive Emergency Obstetric and Neonatal Care (B/CEmONC) facilities per 500,000 people in the population; however, these facilities rely on timely emergency referrals from the community level in order to save lives. The SBA strategy has assumed that skilled birth attendants working in PHCs would assist women having deliveries in PHCs in the catchment areas and provide such timely referrals; however, in low-resource districts, only a few PHCs actually staff SBAs. Because there is a critical shortage of front-line health workers as well as significant time and cost entailed in the training, retention and supervision of skilled birth attendants, increasing the number and distribution of SBAs providing community-based deliveries would reduce the number of SBAs currently available for more complicated deliveries at the referral facilities. Yet, the very high number of PHCs lacking SBAs and their dispersion means that too many rural women will continue to lack access to SBAs for first-line birth care or expert referral.
To address this problem, community-based emergency maternal health care referrals can be rapidly scaled up to fill this referral gap. Effective community-based lifesaving interventions have also been well documented through participatory research and action, and waiting homes have proved effective in some settings.
However, wouldn’t measuring data collected from the EmONC facilities that provide lifesaving signal functions offer a more reliable global indicator of lives saved? Perhaps we can now adapt the “Met need for emergency obstetric care” indicator, defined as the “proportion of women [and newborns] with major direct obstetric complications who are treated in such [C/BEmONC] facilities.” Or more simply and directly, we can count ‘the number of cases referred to EmONC that were treated with a signal function’.
To promote expanded availability of signal functions beyond the designated EmONC facilities, well-established health systems could designate non-EmONC facilities (PHCs) that provide specific signal functions as satellite facilities that coordinate and report their data to their nearest EmONC. Similarly, data from community-based misoprostol distribution interventions could be linked to the nearest EmONC and assigned a ‘signal function equivalency rating’ along the lines of Contraceptive Years of Protection rates. Measurement experts would need to determine the parameters for counting the signal function cases to avoid incentivizing more interventions beyond those warranted by medical indication. For example, they might include a maximum threshold for cesarean cases eligible for inclusion in the indicator data to counteract pressure for non-medically essential cesareans.
Measuring signal function use will continue the Sustainable Development Goal (SDG) focus on strengthening the maternal lifesaving capacity of health systems while ensuring greater equity in low-resource districts and nations. This revised indicator would also catalyze support for interventions to improve community-based emergency referral as an essential component of the health systems that the SDGs are seeking to strengthen.