The aim of this study was to evaluate the potential determinants for emergency medical utilization by elderly patients in Taiwan. The data were drawn from the 'Survey of Health and Living Status of the Elderly in Taiwan', a population-based, longitudinal study of a nationally representative random sample of older adults aged 60 years and older, which was conducted from 1989 to 2007. Face-to-face interviews were conducted at the respondents' homes by trained interviewers accompanied by local health workers. The Andersen Behavioral Model helped us to evaluate the potential determinants for emergency medical utilization that included predisposing factors, enabling factors, and need factors. The measurements of determinants were repeated five times in the period of this study, and the longitudinal data were analyzed through the generalized estimating equation (GEE) by SPSS 17.0 software. The eligibility criteria were that respondents had to be more than 65 years old at baseline in 1993, and then they had to be enrolled in a 14-year follow-up period from 1993 to 2007. At the beginning of this study in 1993, there were 2961 eligible respondents in total, and in 2007, there were 1136 survivors. The loss in follow-up was mainly due to death. The results demonstrated that the significant determinants of emergency medical utilization by the elderly population were gender, age, education, self-ranked health status, chronic disease, and medical accessibility. The GEE model provides a suitable method to predict the long-term trend of emergency medical utilization by the elderly.
Copyright © 2012 Elsevier Ireland Ltd. All rights reserved.