Aims To investigate whether health-related quality of life (HRQOL) predicts hospital admission in a nationally representative sample of adults with diabetes. Methods We conducted a prospective study on persons aged ? 18 years with self-reported physician-diagnosed diabetes (n = 797) who participated in the National Health Interview Survey in Taiwan, 2001. Of these potential participants, 674 provided consent for data linkage and were successfully linked to the National Health Insurance claims data. We analysed the associations between the Short Form 36 (SF-36) subscales and summaries and the occurrence of hospital admission for any cause during 2002. Results Approximately 23% of participants with diabetes had at least one hospital admission during 2002. After adjusting for demographic characteristics, co-morbidities and diabetics-related attributes, those who had been admitted to hospital had significantly poorer mean scores on each of the physical dimensions, physical components summary (PCS) and social functioning domain of the SF-36 at baseline. In logistic regression models, poorer scores on the PCS [odds ratio (OR) = 1.80; 95% confidence interval (CI) = (1.14-2.86)], duration of diabetes ? 10 years [OR = 2.10; 95% CI = (1.14-3.89)] and the presence of heart disease [OR = 1.63; 95% CI = (1.01-2.63)] were significantly associated with an increased risk of hospital admission. Conclusion In people with diabetes, poorer scores on the PCS of the SF-36 at baseline may provide additional information for assessment of hospital admission risk, independent of other measures of health outcomes.
