标题：Trends in Hospitalization and Sociodemographic Factors in Diabetic and Nondiabetic Populations in Germany: National Health Survey, 1990–1992 and 1998
摘要：Objectives. We examined time trends of hospitalization, a main outcome measure in health care, in the diabetic and nondiabetic populations in Germany and their associations with sociodemographic variables. Methods. Using data from 2 national health surveys, we estimated hospital days per person-year in the diabetic and nondiabetic populations in 1998 (n=5422) and 1990–1992 (n=7363) in Germany. We used Poisson regression to estimate relative risks and interaction of secular time with age, gender, and educational level, considering the cluster sample design of the study. Results. Hospital days per person-year decreased between 1990–1992 and 1998—from 3.59 (95% confidence interval [CI]=2.59, 4.97) to 3.14 (95% CI=2.16, 4.56) for the diabetic population and from 1.38 (95% CI=1.23, 1.55) to 1.33 (95% CI=1.17, 1.51) for the nondiabetic population—but the decrease was not statistically significant. In the diabetic population, the decrease tended to be more pronounced (interaction year × time not significant; P =.756). Also, there was a notable decrease in men and in the group aged 25 to 39 years, and a decrease in both high- and low-educational-level subjects. Conclusions. There seems to have been a larger decrease in hospitalization in the diabetic population than in the nondiabetic population in Germany. An increase in social disparity in this health outcome measure in the diabetic population could not be confirmed. Reduction of hospitalization in chronically ill people is a main goal in health care, since hospitalization is associated with a high individual burden and social costs. Interventions to avoid hospital admissions in chronic disease patients include structured treatment, education programs, and specialized ambulatory care offers. 1 – 5 Diabetes mellitus is the prototype of such chronic diseases, and several disease management approaches have been developed. Inpatient as well as outpatient structured education programs may improve disease outcomes and reduce hospitalization. 4 – 8 In studies in Germany, metabolic control improved and hospitalization decreased from 7–11 to 4–5 days per person-year as a result of structured education programs conducted during these studies. 4 , 5 Structured education programs are now routine care in Germany. In addition to the ambition to improve health care for chronically ill people, there was a general effort to reduce hospitalization in several countries during the 1990s, mainly because of rising health care costs. 1 , 2 , 9 Between 1980 and 1992, health care costs in the hospital sector in western Germany increased by more than 150%, to about 90 million deutsche marks, in 1992 (this amount is equal to US $48 million in 2006 dollars). This sum accounted for 45% of total health care costs. Several political interventions have been introduced, including outpatient surgery, predischarge assessment and domiciliary aftercare, and specialized outpatient services. Between 1991 and 1998, the mean length of hospital stay declined from 14 to 10 days. 10 – 12 Social inequalities in health have been described often 13 – 17 and are likely to widen. 18 , 19 In the diabetic population, metabolic control and the occurrence of late complications as well as health care seeking have been shown to be associated with socioeconomic status. 3 , 20 – 27 To the best of our knowledge, no studies on hospitalization trends in the diabetic versus the nondiabetic population have been published. Furthermore, associations between hospitalization trends and sociodemographic variables such as age, gender, and socioeconomic status have not been systematically analyzed at the population level. Thus, our aims were to evaluate trends in hospitalization in the diabetic and nondiabetic populations between 1990–1992 and 1998 in Germany and to evaluate associations of hospitalization trends with age, gender, and educational level (as an indicator of social status). We used data from the 1990–1992 and 1998 National Health Survey. 28 , 29
