Inequality in health care utilization in Germany? Theoretical and empirical evidence for specialist consultation

Aim: In view of increasing concern about a two-class system in the German health care sector, this study investigates the relevance of health insurance schemes and other socioeconomic characteristics to the level of specialist health care provision. Subjects and Methods: Referring to Ronald M. Andersen’s model of health care utilization and more content-based approaches, we implement a negative binomial hurdle regression to estimate the number of specialist visits within the last 12 months. Our data source is the German sample of the first wave of the Survey of Health, Ageing and Retirement in Europe (SHARE) in 2004. Results: The results show that men’s number of specialist visits is markedly sensitive to predisposing and enabling factors, whereas women’s health care utilization depends less on such socioeconomic characteristics. With reference to previous findings concerning general practitioner consultation, the assumption of a bipolar health care system providing general practitioner care primarily to the statutory insured and specialist care to the privately insured is supported empirically as to men. Education, which is considered to be highly correlated with health lifestyles, has a positive effect on medical health care. Every additional year of education increases by about 10% the probability of men seeking specialist consultation. Furthermore, the results indicate an unfavorable situation for the self-employed concerning health care because of their specific employment situation and health insurance coverage. Discussion: The research results suggest the existence of relevant differences in the amount of specialist consultation according to health insurance and other socioeconomic features. Further research could concentrate on the question of whether these inequalities in utilization levels indicate overprovision or underprovision of ambulant health care. Moreover, we recommend longitudinal research that is particularly suited to detangle age and cohort effects