Predicting Health Care Utilization in the Very Old

Abstract
The purpose of this paper is to predict health care utilization in the very old from a combination of individual-based factors such as physical and mental health, health attitudes and beliefs, sociodemographic characteristics, and life circumstances. This study was conducted within the context of the Berlin Aging Study (BASE). Higher use of medications was most strongly predicted by more medical diagnoses, better cognitive status, and health attitudes. Physician contact was only weakly predicted by physical health variables, hypochondriasis, and living alone. In contrast, living alone was the greatest predictor of the utilization of increased levels of caregiving services, while having children nearby served as a protective factor against the need for more formal caregiving services. These results show that utilization of health care depends on interaction between physical and mental health, attitudinal, and social factors.