Evaluation of Health Utility Values for Diabetic Complications, Treatment Regimens, Glycemic Control and Other Subjective Symptoms in Diabetic Patients Using the EQ-5D-5L

Aims: This study aimed to reveal health utility values for diabetic complications and treatment regimens with adjustment for glycemic control and other clinical manifestations in a diabetic population.

Methods: The EuroQol 5-Dimension 5-Level (EQ-5D-5L) health utility values for 4963 Japanese diabetic patients were analyzed using a multivariate regression model including major complications and treatment regiments (minimally adjusted model), and that additionally included glycemic control and other subjective symptoms (musculoskeletal, dental, respiratory, gastrointestinal, urinary, and cutaneous symptoms, and hearing impairment) (further adjusted model).

Results: The mean utility value was 0.901 ± 0.137. In the minimally adjusted model, blindness, overt nephropathy, regular dialysis, cardiac symptom, sequelae of stroke, symptomatic peripheral neuropathy, decreased sensation, claudication, foot ulcer/gangrene, major amputation, and complex treatment regimens were significantly associated with lower utility values, whereas proliferative retinopathy without blindness, coronary artery disease without cardiac symptom, sequela-free cerebrovascular disease, asymptomatic peripheral artery disease, and minor amputation were not. Major complications and treatment regimens that showed significant association in the minimally adjusted model still presented significant impact on the utility decrement in the further adjusted model. However, most of their regression coefficients were lower in absolute value compared to those in the minimally adjusted model.

Conclusions: The utility decrement related to each diabetic complication varied with its severity and accompanying symptoms. Complex treatment regimens were independently associated with lower utility values. The utility decrement associated with diabetic complication and complex treatment regimens would be overestimated in the analysis without adjustment for glycemic control or other subjective symptoms.