Diabetes mellitus, socioeconomy and self-rated health : Population-based studies in cohorts registered in the diabetes incidence study in Sweden

Sammanfattning: The studies presented in this thesis analyzed the impact of diabetes mellitus on socioeconomy and selfrated health, focusing on how the consequences of diabetes develop during the first decade after the diagnosis. The Diabetes Incidence Study in Sweden, DISS, prospectively registers all incid ent cases of diabetes in the age group 15 to 34 years; approximately 400 new cases per year, of whom 65 percent are men and 35 percent women. The cases registered in DISS in 1983 and 1992 and control groups from the general population were selected for the study. Data on healthcare utilization, health-related quality of life, socioeconomic characteristics, and satisfaction with health services were collected via a mailed questionnaire in the younger cohort I year after diagnosis, and in the older cohort 8 years after diagnosis. Costing was based on average national costs of care collected by the Federation of Swedish County Councils and the Swedish Council on Technology Assessment in Health Care, and on sales prices of the National Corporation of Swedish Pharmacies. Hospital daycare costs were derived from a comparative analysis at five hospitals in Stockholm. Analyses of the utilization and costs of care showed that the first year after diabetes diagnosis was associated with per capita costs of care equaling 5.6 times those for the general population, and 8 years after diagnosis 3.8 times those for the population controls. One year after diabetes diagnosis, the annual excess costs of care were US$ 4743 among men and US$ 4976 among women (1997 prices). Hospital inpatient care accounted for more than 50 percent of the excess costs. Eight years after diagnosis, the excess costs were US$ 2010 among men and US$ 2734 among women. The higher costs for women were mainly related to more frequent use of hospital outpatient care, but also to intensive home monitoring. A review of gender equity in diabetes care in Sweden, including population-based studies published during 1990 through 1998, showed that in several studies diabetic women had reported more frequent outpatient contacts, less patient satisfaction, an a lower health-related quality of life than diabetic men. Gender differences had not been reported in the level of glycaemic control. Young and middle-aged men with diabetes have a high excess all-cause mortality as compared to non-diabetic men. A trend toward stronger social gradient in mortality among women than men with diabetes had been reported in one nationwide study. The impact of gender and socioeconomic factors on self-rated health in the diabetic and control groups was examined in multivariable analyses where the dependent variable was the 'general health perceptions' score from the SWED-QUAL questionnaire. A model including a wide range of sociodemographic variables explained 6 percent of the variance in self-rated health I year after diabetes diagnosis and 13 percent of the variance 8 years after diagnosis. Female gender predicted poor self-rated health 8 years after diagnosis but not I year after diagnosis. A study addressing the relationships between dissatisfaction with care, healthcare utilization, and self-rated health showed that dissatisfaction was associated with poor self-rated health both in the diabetic and control groups. Socioeconomic variables were not independent predictors of dissatisfaction after adjustment for self-rated health. A significant association between female gender and dissatisfaction with care was found 8 years after diabetes diagnosis but not I year after diagnosis. The combined impact of gender and socioeconomic status on the prognosis of diabetes and other chronic diseases should be investigated in further studies. Monitoring the impact of gender should become an integrated part of quality management in diabetes care. A question for future health systems research is how the investments in medical care during the first 10 years after diabetes diagnosis impact on the use and costs of care 10 to 20 years later, when a higher incidence of long-term complications can be expected. Economic analyses and simulation models based on randomized controlled trials do not produce evidence about the real-life investments in diabetes care nor about the factual pay-off in terms of quality and quantity of life. For these purposes, followup, studies in unselected diabetic populations living in realworld health systems are needed.