Empiric investigation on direct costs-of-illness and healthcare utilization of Medicaid patients with diabetes mellitus.
OBJECTIVE: To determine total direct costs-of-illness and to study the influence of different factors affecting these costs. In addition, we examined each type of service (e.g., hospitalization, outpatient care, prescription drugs, physician encounters, and laboratory tests) for diabetic Medicaid patients to provide evidence about the relationship between diabetic patients' healthcare utilization and their related predictors. PATIENTS AND METHODS: A total of 7931 patients with diabetes who were 65 years or younger in the Alabama Medicaid program from 1992 to 1995 were studied. Using a relational database created from Medicaid claims, multiple regression and canonical correlation methods were used to analyze the patients' direct costs-of-illness, including the costs associated with each healthcare service used by each patient. RESULTS: The costs of hospitalization, outpatient care, prescription drugs, and physician encounters were the four largest components of the direct costs-of-illness for diabetic Medicaid patients, comprising 29.9%, 21.3%, 28.2%, and 14.3%, respectively. After controlling for other factors in an empiric model, the direct costs-of-illness for a patient with insulin-dependent diabetes mellitus was $5160 higher than for a patient with noninsulin-dependent diabetes mellitus during the 3-year study. The cost for a patient with renal dysfunction was $59,920 higher than for other diabetic patients. Each increase in the number of different prescribing physicians per patient was associated with a cost increase of $450. Each additional comorbidity increased the cost by $735 per patient. The cost for a male patient was $2140 higher than that for a female patient, and the cost for a white patient was $1330 higher than that for a non-white patient. For a patient who relied on diet to control diabetes, there were $2750 less in costs compared with other patients during the study period. More than 20% of the variability in patients' healthcare utilization costs was explained by the set of predictive factors. CONCLUSIONS: The direct costs-of-illness and healthcare utilization for Medicaid diabetic patients were significantly accounted for by the number of comorbidities, the number of different physicians visited, insulin-dependent diabetes mellitus, and complications (especially renal dysfunction). Patients who relied on dietary therapy and exercise to control their diabetes had lower healthcare costs and utilization than other patients. A significant amount of healthcare costs and utilization might be controlled or reduced if diabetes disease management can successfully be aimed at preventing diabetic complications, controlling comorbidities, and minimizing the number of different physicians visited.