In sub-Saharan Africa, intense investment in community-based primary health care is aimed at reducing the high level of child mortality. Programs and policies currently promote the hiring and deployment of community health workers to village health posts. Such policies lack adequate grounding in statistically rigorous analyses of the interaction of social contextual factors with the programmatic strategies that are employed. In this setting, social systemic determinants of health behavior are highly influential i determining child survival, yet the link between health systems determinants and social system influences is often poorly understood. Moreover, socio-demographic studies fail to provide direct evidence of system strengthening effects, or to indicate which investments are most productive, because survey evaluation procedures are unconnected with service strengthening inputs. The means of optimizing health systems inputs are unassessed or only crudely elucidated by evaluation research. Therefore, the goal of our proposed research is to successfully model the complex relationships between the context of health services at the community and higher levels and the health seeking behavior of individuals and communities on child survival. The model we propose to develop will identify which health system strategies are responsible for improvements in health, given social contextual effects. We will develop and estimate a complex multilevel model of child survival and life years gained that represents a theory of health systems strengthening and health seeking behavior based on extant data from northern Ghana. This model will assess the relationship between mortality among children under five and the addition of community-engaged primary health care services during the development and scale-up of Ghana's national primary health care policy, known as the Community-based Health Planning and Services Initiative (CHPS), from 1994 to 2003. We will use existing multilevel, longitudinal data on health systems strengthening in conjunction with a decade of yearly socio-demographic survey data and demographic surveillance of over 6,000 children under five years of age in 257 communities zoned for primary health care scale-up in Ghana's Kassena-Nankana District. The proposed work will demonstrate practical means of combining survey research with programmatic monitoring data to study the impact of health systems strengthening inputs on health behavior. It will also bridge the gap between individual level behavior and service system dynamics. Using demographic methods, parameters measured through this analysis will be used to calculate child life years gained through health system investments at the community level. These data will allow us to elucidate the relationship between health system strengthening that addresses geographic inaccessibility, as well as social barriers to health care, individual and community level health behavior, and their effect on child mortality. We expect our results to provide keenly needed guidance to strategies for health systems strengthening in sub-Saharan Africa.
Our proposal is to demonstrate practical means of measuring the complex and multidimensional issues involved in developing public health service capability in resource constrained settings. We will develop and estimate a novel multilevel longitudinal model of health systems strengthening over a ten year period in rural Ghana. The model will include individual and community level measures of health behavior, providing a statistically rigorous analyses of the interaction of social contextual factors with the programmatic strategies that are employed, and identifying the most effective components of health systems strengthening.
