Development of the Diabetes Resource Consumption Index and profiling quality of diabetes care in the Veterans Health Administration

Abstract

The purpose of this study was to develop and validate a risk-adjustment index for one year healthcare resource use specific to diabetic patients, based on severity of illness measures; and to profile quality of diabetes care between outpatient clinics. The data for this study was collected from four outpatient clinics within the Southern Arizona Veterans Affairs Healthcare System, Tucson, AZ. The DRCI was developed using a sample size of 367 diabetic subjects that had complete information on diabetes-specific variables. Individual DRCI weights, based on the magnitude of one year healthcare resource use and socio-demographic characteristics, ranged from -471.5 to 3,081.2 for total healthcare costs, from -304.3 to 1,582.1 for outpatient costs, and -0.19 to 0.93 for risk of hospitalization. The DRCI was better than or equivalent to the Chronic Disease Score in predicting health care costs. Diabetics in the second cohort were predominantly elderly (mean = 66yrs ± 11.1), married (61%), white (73%), males (96%), had a high BMI (31 ± 6.3 kg/m²), and mean comorbidity score of 4.2 ± 1.8 conditions. Screening for HbA1c and microalbuminuria was frequently performed in all clinics. Overall, 61% and 36% of study patients did not have evidence of foot or eye examinations during the entire study period, respectively. Approximately, 27% (n = 408), 41% (n = 643), and 26% (n = 515) of the study patients had poor glycemic, renal function, and lipid control, respectively. Significant differences (p < .05) in HbA1c and creatinine clearance rates between the clinics were observed after adjusting for patient case-mix. However, differences between the clinics in cardiovascular outcome were not observed after adjusting for patient case-mix. This study demonstrated an association between diabetes severity with healthcare resource and costs. The DRCI, using laboratory data, is a diabetes-specific severity measure for prediction of one year healthcare resource use. Future studies are needed to validate this index in other settings. Finally, the results from this study emphasize the need to adjust for case-mix variable when comparing quality of diabetic care outcomes between outpatient clinics.