Measurement Issues

Measurement of maternal mortality remains a difficult issue, particularly in developing countries but also in developed ones. A number of innovative techniques for measuring maternal mortality (e.g. sisterhood methods) or estimating it (e.g. models-based approaches) have been developed over the last decade or more. However, these techniques are sufficiently described in the literature and have not solved all measurement problems, so the session will not focus on them. Rather, papers on measurement submitted for this session should focus on innovative approaches that have not been widely applied, beyond the familiar approaches such as sisterhood (direct or indirect) and model-based estimates. Examples might include vital registration combined with surveillance systems or census questions on household deaths.

Measures into Policy

One shortcoming of methods for estimating maternal mortality is that the resulting measures are not very useful for setting policy, identifying interventions or evaluating impact. The main reason for this failure is that the methods tend to give a single number, say the Maternal Mortality Ratio, with no differentials or potential for further analysis. For devising and evaluating interventions, a broader range of indicators is required. For example, indicators are needed which reflect health care seeking behaviour, knowledge of the symptoms of pregnancy-related morbidity, access to obstetric care and facility-based quality of care. Reproductive Age Mortality Surveys have in some cases included detailed case investigation to determine what went wrong, but other approaches are possible. For evaluation, program indicators beyond maternal mortality are necessary. Methodological or substantive papers concerning either the identification of critical failures or intervention evaluation will be considered for inclusion in the session.

Consequences of Maternal Mortality

The death of a woman in childbirth is presumed to have profound implications for the wellbeing of surviving family members. Data from Matlab, Bangladesh have been used to show greatly elevated mortality risks to older children, but broader socio-economic implications, both short-term and long-term, have received little attention. Data from demographic surveillance projects could provide a base for analysis, as could some type of case-control study. Papers describing the consequences of a maternal death (in comparison to impacts on families experiencing a non-maternal death of a mother and on families experiencing no such death) will be considered.