Deciphering Rashomon: An approach to verbal autopsies of maternal deaths

Abstract

The paper discusses an approach to verbal autopsies that engages with the Rashomon phenomenon affecting ex post facto constructions of death and responds to the call for maternal safety. This method differs from other verbal autopsies in its approach to data collection and its framework of analysis. In our approach, data collection entails working with and triangulating multiple narratives, and minimising power inequalities in the investigation process. The framework of analysis focuses on the missed opportunities for death prevention as an alternative to (or deepening of) the Three Delays Model. This framework assesses the behavioural responses of health providers, as well as community and family members at each opportunity for death prevention and categorises them into four groups: non-actions, inadequate actions, inappropriate actions and unavoidably delayed actions. We demonstrate the application of this approach to show how verbal autopsies can delve beneath multiple narratives and rigorously identify health system, behavioural and cultural factors that contribute to avoidable maternal mortality.

Keywords:

• verbal autopsy
• maternal death
• multiple narratives
• health system failures
• behavioural and cultural factors

Acknowledgements

This research by the Gender and Health Equity Project at the Centre for Public Policy, Indian Institute of Management Bangalore was funded by the MacArthur Foundation and the International Development Research Centre. The authors offer heartfelt thanks to the families of the deceased women who willingly participated in the autopsies while grieving their loss. The authors are deeply indebted to colleagues for their invaluable contributions: to Deepa Dhanraj for alerting us to the complexities of death reporting; to Manasa Patna for her input during the methodology's early development, to Pakeeramma Budakunti, Y. Poornima and Vasini Varadan for their experiences and questions that helped in its refinement. The authors also thank their anonymous reviewers for comments that have helped strengthened the paper.