A pilot study to assess cognition and pillbox fill accuracy by community-dwelling older adults

OBJECTIVE: To assess pillbox fill accuracy and cognition among community-dwelling older adults. DESIGN: A descriptive, cross-sectional study. SETTING: Retail pharmacy. PARTICIPANTS: Convenience sample of English-speaking adults older than 60 years of age without dementia, taking more than four medications, and naive to Mediset use. INTERVENTIONS: In face-to-face interviews, subjects provided demographic, medical, and medication information, completed the Mini-Cog and Medi-Cog (combination of Mini-Cog and medication-transfer screen [MTS]), and filled their own medications in a pillbox. Data were analyzed using descriptive statistics and stepwise regression analysis with correctly filled pill count (PC) as the dependent variable and the cognitive screens as independent variables. MAIN OUTCOME MEASURES: Accuracy of the Mini-Cog, MTS, and Medi-Cog in predicting PC. RESULTS: Among 50 subjects (58% female, mean age 76.4 years), only one subject failed to pass the Mini-Cog and two failed to reach the criterion level of correctly filled PC. The mean (standard deviation) Mini-Cog score for the sample was 4.38 (0.81), MTS score was 4.1 (1.31), Medi-Cog score was 8.48 (1.82), and the mean PC was 97% (8%). The Mini-Cog and MTS individually accounted for about 30% of the variance (P < 0.001); the Medi-Cog accounted for 44% of the variance (P < 0.001), indicating strongest PC prediction. CONCLUSION: Nearly all study participants filled pillboxes accurately. The Medi-Cog was the strongest predictor of pillbox fill accuracy. Future studies of medication selfmanagement abilities among community-dwelling older adults should include representative samples of this population, comprehensive assessment of health status, cognitive screening, pillbox fill accuracy, and the utilization of medications in filled pillboxes.