AN ECONOMETRIC STUDY OF THE DECISION TO SEEK MEDICAL CARE IN WEST AFRICA: A CASE STUDY OF THE GHANA DANFA HEALTH PROJECT USING DISCRETE CHOICE MODELS

Abstract

A theoretical and an empirical investigation using Logit Analysis, Discriminant Analysis, (Hierarchical) Log-Linear models with factor interactions and Goodman's measures of optimal prediction and uncertainty within the framework of consumer choice theory to explain the usage of health-care facilities and the behavior of individual consumers and different population segments seeking medical care within the Danfa Community in Ghana, West Africa. Based upon the household objective of utility maximization and the set of constraints--income, wealth, time, information and health, the demand for medical care is estimated as a function of individual and system characteristics, i.e. those characteristics describing in broad terms the factors of the household's needs, perception, willingness to secure care, and ability to secure care (e.g. age, sex, education, ethnicity, type of disease, literacy, health condition, occupation and costs of medication, travel and consultation. The sources of treatment or the providers of medical care were classified into five categories: self, family, drug seller, herbalist, and clinic. Our empirical results indicate that an individual's decision to seek or purchase medical care is more likely to be based on individual characteristics such as the number of unhealthy days rather than on system characteristics such as prices or costs of medication, travel, etc. This study has also revealed that some segments of the Danfa population in Ghana are more likely to exhibit an increasing preference or avoidance for certain health care facilities than others or use health-care facilities in different ways by either purchasing more or less medical care than other consumer groups. In assessing the effects of changes in the levels of particular factors on health-care decisions, our empirical results indicate that there is a reduction in total medical outlays for some consumers if there is a rise in the number of unhealthy days or an increase in the cost indices of medication, travel, and consultation. Policy measures for improvement in the future, including the reduction of the number of unhealthy days and household medical care expenditures through preventive health care education, community-based health insurance schemes for various occupational groups, and improvement of access capabilities or income earning capabilities through the encouragement of proper organization of economic activities within the rural community have been recommended in this study.