Identification of potentially inappropriate medications characteristic of older individuals with diabetes: a study using pharmacy claims data

Older patients with diabetes often require multiple medications to manage complications, and have a high risk of potentially inappropriate medications (PIMs). Given limited information regarding diabetes-specific PIMs, we characterized PIM prescriptions in older patients with and without type 1 or 2 diabetes, and examined gender differences in diabetes-specific PIMs, using pharmacy claims data obtained from nationwide branches of a chain pharmacy company in Japan. Pharmacy claims data of patients aged 65 years and over who used one of the 905 community-based pharmacy branches of Sugi Pharmacy Co., Ltd. during a month were anonymized and analyzed. The enrolled individuals were separated to diabetic and non-diabetic groups on the basis of anti-diabetic medications. The two groups were balanced for the age, gender ratio, and status of home healthcare insurance application through propensity score matching (PSM). PIMs, i.e. medications prone to cause adverse reactions and replaceable with others, were classified to twelve categories other than anti-diabetic PIMs, and the incidence of each PIM category in diabetic and non-diabetic groups was analyzed. Of 333,869 older patients, 37,606 received anti-diabetic medications. Analysis of pharmacy claims data after PSM showed that aged individuals with diabetes had significant increases in the number of medications and the incidence of PIMs, compared to individuals without diabetes. Multivariate logistic regression analyses indicated that older individuals with diabetes had significantly increased incidences of six PIM categories, sleeping drugs, α-adrenoceptor blockers, H2-blockers, drugs for overactive bladder, diuretic drugs, and concomitant intake of multiple antithrombotic drugs. Sub-group analyses of each gender data detected a diabetes-related significant increase in the incidence of the PIM category of drugs for overactive drugs in older females, but not males [adjusted odds ratios (95% confidence interval): 1.59 (1.35–1.88) and 0.91 (0.77–1.08), respectively], while significantly increased incidences of the residual five PIM categories in the diabetic group were detected regardless gender. Older individuals with diabetes had higher risks of five PIM categories regardless of gender, and of one female-specific PIM category, than individuals without diabetes. These findings are helpful for pharmacists to make recommendations of deprescribing PIMs in pharmaceutical care of older patients with diabetes Geriatrics.