Influence of morbidity and health care structure on hospitalization among adult diabetic patients.

Abstract

The aim of the study was to investigate the influence of outpatient care, overall hospital bed provision and the prevalence of diabetes on the hospital use by adult diabetic patients. Population based data were collected in Finland from a 3-year period. Hospital admissions, inpatient days, and mean length of stay due to diabetes were derived from the Hospital Discharge Register. Estimates of diabetes prevalence were derived from the Central Drug Register. Primary outpatient site, annual number of physician visits, and age at onset among diabetic patients were assessed by questionnaire. The unit of analysis was hospital district (n = 21) and all analyses refer to the 15-64 years age group. Linear regression models were used to explain discharge rates and inpatient days in the district. Large variations between hospital districts were observed in diabetes prevalence, discharge rates due to diabetes, and mean length of stay among diabetic patients. Prevalence of diabetes and overall supply of hospital beds in the district were not related to hospital use. Univariate analyses showed that the larger the proportion of diabetic patients primarily using a private practitioner, the lower the hospital use. In districts where early onset of diabetes was more common, the average number of inpatient days was significantly lower than in other districts. In stepwise regression analyses, early onset of diabetes was the only variable that entered the models. In the full model, 24.6% of the variance in inpatient days and 16.1% of the variance in discharge rates were explained. In conclusion, factors related to health care structure, provision of hospital beds or morbidity explain little of the regional variation in hospital use. The observed variations in hospital use probably depend mainly on local treatment policies and differences in clinical decision-making.