MEDICATION DISCREPANCIES ASSOCIATED WITH DIABETES MELLITUS IN COMMUNITY DWELLING PRIMARY CARE OLDER ADULTS

Medication management issues in persons with diabetes (PWD) are well documented. Few studies have examined community-dwelling older PWD in primary-care provider (PCP) practice setting to determine what medications PCPs prescribe and what patients actually take. \ud This secondary analysis, guided by Donabedian’s structure-process-outcome framework, examined medication discrepancies (MD) in community-dwelling PWD (n = 142), 65 years of age and older, in the PCP setting. The aims were to (1) characterize the sample, (2) characterize the discrepancies associated with prescribed medications, and (3) identify potential correlates of medication discrepancies. This study used de-identified baseline data (n = 533) from a parent study (NIH/NIA AG023129), which examined the utility of cognitive function testing of older adults in the PCP setting. The Donabedian structure component included variables for subject characteristics such as sociodemographic variables, health information, and neuropsychological variables. The process component included data from a comprehensive medication review, which generated a complete and accurate list of the subject’s current medications and allowed a comparison of the patient-generated list with the provider-generated list present in the subject’s medical record. \ud In 95% of the subjects, MD were present. Among subjects with the same number of health problems, those with a higher number of medications were more likely to exhibit MD compared to the subjects with a lower number of medications. Among patients with the same number of medications, those who had a higher number of health problems were less likely to have a MD compared to the subjects with fewer health problems. \ud Polypharmacy and the number of health problems were the most significant correlates of a medication discrepancy. While not significant, a trend was observed for diminished cognitive function and the presence of a MD (p = 0.056). Despite a high MMSE mean score (27.9 ± 1.9) and positive correlations with neuropsychological scores, mild cognitive impairment was discovered in 44% of the sample—and four or more depressive symptoms in 69.72% of the sample. \ud The pervasiveness of medication discrepancies and health problems in a population of PWD at risk for cognitive impairment and depressive symptoms has significant health care implications that deserve further study. \u