Contextualising maternal morbidity through maternal health audits

Introduction: Recent efforts towards data collection on maternal health have focused on women who survived a life-threatening maternal morbidity (near miss). This has been based on the assumption that the causes of near misses are similar to those that lead to mortality. The work on near misses has shed significant light on challenges within health facilities. However, this focus overlooks critical insights that could be gained from an analysis of health systems more broadly defined to include the critical role played by major actors in communities. Methods: This research used a multi-method and grounded theory approach to obtain local explanations of the concept of life-threatening maternal complications and tested that understanding in a population survey in the Kassena-Nankana Districts of northern Ghana. A qualitative study with traditional healers and traditional birth attendants provided a comprehensive explanation of what communities might view as a life-threatening maternal complication. These explanations were used to screen pregnant women within the community to determine how well the explanations fit with women experiences of life-threatening maternal complications. Women who qualified as having survived a life-threatening maternal complication were administered a semi-structured audit tool to assess the underlying factors to maternal complications. Using focus groups and in-depth interviews, the study explored the relevance of sharing the audit results with community leaders in order to enlist their support in addressing challenges that women face during pregnancy. Results: The study findings revealed that communities have an understanding of what constitutes a life-threatening maternal condition and traditional healers play a key role in the management of these complications. They defined a "life-threatening maternal complication” as any health condition related to pregnancy that increased a woman's risk of dying. Using this definition in a population survey produced a prevalence of life-threatening maternal complications of 19.8%. An audit with women who had complications showed that not all women with life-threatening maternal complications access care at health facilities. Delays in recognising danger signs during pregnancy, delays in making the decision to seek care, delays in arriving at an appropriate place of care and delays in receiving treatment were reported by respondents. Community leaders expressed interest in pooling resources for their collective benefits but feared such an initiative would be limited by poverty, lack of trust, corruption and the reciprocity of benefits. Conclusions: The results show the feasibility of conducting a community audit on life-threatening maternal complications and the need to pursue an agenda to integrate traditional healers within the health system. A maternal morbidity audit model that integrates community engagement in the audit process is likely to get community leaders to think about interventions that need not directly address a specific cause but may nonetheless mitigate a pathway of causes.