Oral antidiabetic drug use and associated health outcomes in cancer patients.

Abstract

To assess oral antidiabetic drug use and associated health outcomes in American non-elderly adults with cancer A retrospective study was conducted by analysing the Marketscan(®) Commercial Claims and Encounters Database from 2008 to 2009. Individuals 18-64years with concomitant diagnoses of cancer (breast, prostate, colon or lung) and type 2 diabetes, and treated with oral antidiabetic medications were included. Medication adherence was assessed using the Medication Possession Ratio (MPR); logistic regression was used to analyse factors associated with non-adherence. Adherence was compared between patients with cancer and non-cancer controls matched by propensity scores. Negative binomial regression was utilized to examine the effect of antidiabetic drug adherence on all-cause hospitalizations and emergency room visits. The impact of adherence on total medical costs was then evaluated using the generalized linear model (GLM) with the log-link function and gamma error distribution. Of 1918 diabetic, cancer patients who newly initiated oral diabetic drugs, only 37·6% were adherent to oral diabetic medications; a similar proportion of adherence was found in the non-cancer control population (35·8%, P=0·24). Younger age, living in the southern region, using combination therapy vs. monotherapy, and using retail pharmacy vs. mail order pharmacy were significantly associated with non-adherence in patients with cancer. Adherence to oral antidiabetic drugs was associated with 24% fewer all-cause hospitalizations (P=0·02). We identified high prevalence of non-adherence to oral antidiabetic medications as well as negative consequences associated with non-adherence, among patients with cancer. These findings may underscore the importance of developing relevant intervention strategies for improving diabetes management and treatment outcomes among cancer patients with diabetes.