NATIONAL ESTIMATES OF UTILIZATION AND EXPENDITURES FOR PATIENTS WITH DIABETES AND CANCER

UPP 113

Objective. To investigate the prevalence of cancer comorbidity in individuals with diabetes, and its association with health care utilization and expenditures. Background. The connection between cancer and diabetes has been explored comprehensively and most found that diabetes is associated with an increased risk of cancer in several sites. Most published data, however, ignored the effect of confounders in order to come up with accurate estimates. Methods. Our study was conducted with a quasi-experimental design. Subjects included patients who had ever been diagnosed with diabetes in 2010 Medical Expenditure Panel Survey (MEPS). The propensity scores technique was utilized to reduce selection bias by controlling risk factors such as age, sex, race/ethnicity, physical activity, smoking and body mass index. Further, a series of weighted inferential statistics were used to test the effect of cancer comorbidity on the variables associated with health care use and expenditures. Results. Estimated 21 million non-institutionalized adults were reported having diabetes in the US, 2010. Among them, 3.89 million (18.5%) had cancer comorbidity. Patients with diabetes were twice likely to encounter cancer compared with non-diabetes individuals (crude odds ratio 2.1, 95% CI 1.9-2.49). After controlling for the shared risk factors with propensity score, the odd ratio remained significant (odds ratio 1.33, 95% 1.146-1.554). Variables associated with health care use and expenditures for individuals with cancer were significantly higher than those without cancer (p < 0.0001). Conclusion. Our study findings indicate that diabetes patients have higher prevalence of cancer comorbidity, thus resulting in increased health care utilization and expenditures. Grants. No grants.