PhD defence Christelle Boyi Hounsou

Supervisors

• Prof. dr. Bruno Marchal (ITM)

• Prof. dr. Lenka Benova (ITM)

• Dr. em. Thérèse Delvaux (ITM) 

• Prof. dr. Marjolein Zweekhorst (VU Amsterdam)

• Prof. dr. Edgard-Marius Ouendo (IRSP, Benin)

• Dr. Jean-Paul Dossou (CERRHUD, Benin)  

Abstract

Introduction

Maternal mortality remains unacceptably high in sub-Saharan Africa, with Benin reporting 518 deaths per 100,000 live births in 2023. Over 80% of these deaths are considered preventable through high-quality maternal care. In response, Benin institutionalized Facility-Based Maternal Death Reviews (FbMDRs) in 2013 as a key strategy for improving care quality. FbMDRs involve systematic, in-depth reviews of the circumstances surrounding maternal deaths. Although recommended by the World Health Organization and implemented in many sub-Saharan Africa countries, their effectiveness has been inconsistent, shaped by contextual factors at national, institutional, and provider levels.

This thesis investigates the extent to which, how, why, and under what conditions FbMDRs contribute to improving the quality of maternal care in Benin. It aims to generate both theoretical insights and practical recommendations to strengthen their implementation.

Methods

A realist evaluation approach was employed, utilizing a multiple case study design. The study commenced with an analysis of how various dimensions of maternal care quality manifest in Benin and how they can be meaningfully captured. This informed the development of an initial program theory, grounded in a literature/document review, key-informant interviews, and the expertise of the research team.

Empirical research was conducted in two hospitals where FbMDRs were implemented alongside other quality improvement initiatives. Quantitative analysis of FbMDR reports from 2021 to 2023 examined review completion rates, identification of root causes, and the quality of proposed recommendations. Qualitative data were collected through semi-structured interviews and non-participant observations with providers and managers. All data were analyzed using the ICAMO framework (Intervention–Context–Actors–Mechanism–Outcome), combining within-case and cross-case synthesis with a retroductive approach to refine the initial program theory.

Results

The findings revealed a low completion rate of reviews (13%) and limited follow-up on recommendations, many of which addressed only superficial issues. Providers often experienced the review process as emotionally taxing and ineffective, leading to detachment and disengagement. Mechanisms such as fear of blame, low perceived utility, and lack of ownership inhibited meaningful participation. However, in contexts where trust, self-reflection, and a sense of efficacy were present, providers demonstrated greater engagement and commitment to learning. Systemic constraints, hierarchical cultures, and weak feedback loops created a cycle of frustration and disempowerment. The analysis highlighted that effective FbMDR implementation requires the alignment of policy goals, institutional capacities, and provider motivation. A redesigned FbMDR procedure, informed by these findings, was co-developed with stakeholders  and piloted to promote contextual relevance and system-level responsiveness.

Discussion and implications

The refined program theory demonstrates that the success of FbMDRs hinges on the coherent activation of interrelated components, each step in the causal chain enabling the next. When key contextual barriers and mechanisms remain unaddressed, this chain is disrupted, and the review process loses its potential to drive improvement.

This research offers actionable insights for policymakers and practitioners. Embedding FbMDRs within a system of continuous learning and supportive leadership offers strong potential to improve maternal health outcomes in Benin and similar context.