Background: Why do some countries rapidly adopt policies suggested by scientific consensus while others are slow to do so? HIV treatment is a particularly salient case in point. Scientists, physicians, and the World Health Organization (WHO) spend significant effort identifying the optimal standards of medical care?yet the guideline policies that govern public health and medical practice often lag far behind evidence. Efforts to address differences in adoption of “evidence-based” policy focuses on variation in the capacities for adoption and interpretation of scientific evidence. We challenge this view, hypothesizing that socio-political institutions are often decisive in how quickly countries will adapt science into policy.
Methods: We used a mixed methods strategy. First, we constructed a database of national HIV treatment guidelines, collecting 290 published national ART guidelines for adults and adolescents from 122 countries (98% of global HIV burden). Using this database, we built a model to reflect the epidemiologic, economic, and political context and used a Cox-Proportional Hazards Model to test our hypothesis. Second we conducted interviews with 24 key informants with direct knowledge of guidelines processes to establish causality in our study.
Results: Neither HIV prevalence nor national wealth is a significant driver of policy change in our quantitative or qualitative results. Qualitative analysis shows neither interpretation of medical evidence nor formal cost-benefit analysis explain differences. Instead, the formal structures of government and the degree of ethnic cleavages predict the speed with which new medical science is translated into policy. More veto points in government is strongly associated with faster policy adoption?with a full swing in the data associated with a 275% increase in adoption speed. HIV policy change is slower in contexts with complex racial/ethnic divisions?as much as 60%, ceteris paribus.
Conclusions: Our findings challenge expectations in scholarship and practice that policy divergence and inequities are primarily addressed through greater evidence and dissemination channels. That political factors are systematic, rather than random, suggests a new approach is needed by agencies such as the WHO and UNAIDS. Building diffusion strategies, messages, and policy networks that are tailored to national political context is possible when systematic socio-political factors are identified.