MEDICAL MESOCOSMS AND COHORT DIFFERENCES IN VICTIM DECISIONS IN SPOUSAL VIOLENCE IN SUB-SAHARAN AFRICA

Patricia Elung'ata^*

• McGill University, Montreal, Canada

Background: Contradictory evidence exists on whether medicine explains social disparities in health perceptions. This study evaluates healthcare systems as mesocosms to understand social differences in spousal violence perceptions in sub-Saharan Africa, concretely cohort differences in victim decisions in spousal violence (VDSV). Conceptual Framework: Medical Dominance theory addresses power-asymmetry in medicine, while socio-ecological theory examines social disparities in partner violence in low-and-middle income countries. This study analyzes how power asymmetry influences spousal violence victim decisions in sub-Saharan Africa, investigating socio-behavioural patterns in medicine as parallels to spousal violence. Methods: A cross-sectional study analyzed data from the latest Demographic and Health Survey (2001-2024) across 31 countries, focusing on 193,232 women aged 15 to 49 and their VDSV patterns: none, risk-exacerbating (discordant), and other decision types(concordant). Unadjusted Odds Ratios(OR) examined associations between VDSV patterns and Birth Cohorts(BC), while adjusting for confounders like household wealth and age. Interaction terms assessed the impact of neighbourhood ethnic diversity, relationship power differences, and healthcare access. A two-level hierarchical multinomial logistic regression model analyzed VDSV variation, considering individual, cluster, and household level confounders, with random country effects. Spatial interpolation addressed geographical clustering, and analyses were performed using the Mclogit package in R(Version 4.4.0). Results: A greater percentage change in predicted marginal probabilities for Other VDSV compared to Discordant VDSV was observed across BC. Women with recent healthcare access had smaller percentage changes in these probabilities; while those without a final say in household decisions showed the opposite trend. The Relative Index of Inequality (RII) estimates further indicate that healthcare access affects socioeconomic health inequalities. Discussion: VDSV patterns amongst younger birth cohorts show lower SV-accepting attitudes in SSA, potentially confounded by higher incidence rates. Research indicates that variations in VDSV may be linked to exposure to anti-violence interventions at healthcare facilities; findings highlight relationship power dynamics play moderating roles. Furthermore, differences in VDSV related to healthcare access highlight its importance in reducing social health inequalities. Conclusion: The study results advocate for macro-societal policies in healthcare that address social issues, particularly through theory-based approaches. It also explores the implications of these findings amid healthcare funding cuts in developing countries.