An option for measuring maternal mortality in developing countries: a survey using community informants

Siti Nurul Qomariyah,David Braunholtz,Kamaluddin Latief,Karen H Witten,Trisari Anggondowati,Wendy J Graham,Endang L Achadi,Eko Setyo Pambudi

doi:10.1186/1471-2393-10-74

Siti Nurul Qomariyah, David Braunholtz + Show 6 more

Open Access

https://doi.org/10.1186/1471-2393-10-74

Copy DOI

Abstract

BackgroundThe maternal mortality ratio (MMR) remains high in most developing countries. Local, recent estimates of MMR are needed to motivate policymakers and evaluate interventions. But, estimating MMR, in the absence of vital registration systems, is difficult. This paper describes an efficient approach using village informant networks to capture maternal death cases (Maternal Deaths from Informants/Maternal Death Follow on Review or MADE-IN/MADE-FOR) developed to address this gap, and examines its validity and efficiency.MethodsMADE-IN used two village informant networks - heads of neighbourhood units (RTs) and health volunteers (Kaders). Informants were invited to attend separate network meetings - through the village head (for the RT) and through health centre for the kaders. Attached to the letter was a form with written instructions requesting informants list deaths of women of reproductive age (WRA) in the village during the previous two years. At a 'listing meeting' the informants' understanding on the form was checked, informants could correct their forms, and then collectively agreed a consolidated list. MADE-FOR consisted of visits relatives of likely pregnancy related deaths (PRDs) identified from MADE-IN, to confirm the PRD status and gather information about the cause of death. Capture-recapture (CRC) analysis enabled estimation of coverage rates of the two networks, and of total PRDs.ResultsThe RT network identified a higher proportion of PRDs than the kaders (estimated 0.85 vs. 0.71), but the latter was easier and cheaper to access. Assigned PRD status amongst identified WRA deaths was more accurate for the kader network, and seemingly for more recent deaths, and for deaths from rural areas. Assuming information on live births from an existing source to calculate the MMR, MADE-IN/MADE-FOR cost only $0.1 (US) per women-year risk of exposure, substantially cheaper than alternatives.ConclusionsThis study shows that reliable local, recent estimates of MMR can be obtained relatively cheaply using two independent informant networks to identify cases. Neither network captured all PRDs, but capture-recapture analysis allowed self-calibration. However, it requires careful avoidance of false-positives, and matching of cases identified by both networks, which was achieved by the home visit.

Highlights

The maternal mortality ratio (MMR) remains high in most developing countries
We found that the proportion of women classified as pregnancy related deaths (PRDs) who died from maternal causes in the study area was very high (i.e. 97.5%) meaning that choice of PRD or maternal death definition is not of major importance
Probability of each informant network capturing deaths In urban and in 10% of rural villages, where we used both kader and RT networks to capture eligible PRDs, 116 eligible PRDs were identified: 13 were identified only by the Kader network, 30 identified only by the RT network, and 73 identified by both. This gives a point estimate of 0.71 for the probability of an eligible PRD being identified by MADE-IN/MADE-FOR via Kader information, and an estimate of total eligible PRDs of 121

Summary

Introduction

Recent estimates of MMR are needed to motivate policymakers and evaluate interventions. Recent estimates of MMR are needed to motivate policymakers to prioritise maternal health, and to evaluate interventions. In Serang and Pandeglang districts, Banten Province, as in all of Indonesia, the administrative system includes community level volunteers (RTs) responsible typically for 10-40 house-holds. Distinct from this are health volunteers (Kaders) responsible for the integrated health posts with coverage of approximately 100 households. Since these systems are comprehensive, functional, and extend down to a level at which it would be reasonable to expect good knowledge of deaths and maybe circumstances of death, there is an obvious possibility of using them as a basis for ‘capturing’ maternal deaths

Methods

Results

Discussion

Conclusion