No impact of a prescription booklet on medication consumption in nursing home residents from 2011 to 2014: a controlled before–after study

Abstract

Abstract Background Older persons are particularly exposed to adverse events from medication. Among the various strategies to reduce polypharmacy, educational approaches have shown promising results. We aimed to evaluate the impact on medication consumption, of a booklet designed to aid physicians with prescriptions for elderly nursing home residents. Methods Among 519 nursing homes using an electronic pill dispenser, we recorded the daily number of times that a drug was administered for each resident, over a period of 4 years. The intervention group comprised 113 nursing homes belonging to a for-profit geriatric care provider that implemented a booklet delivered to prescribers and pharmacists and specifically designed to aid with prescriptions for elderly nursing home residents. The remaining 406 nursing homes where no such booklet was introduced comprised the control group. Data were derived from electronic pill dispensers. The effect of the intervention on medication consumption was assessed with multilevel regression models, adjusted for nursing home status. The main outcomes were the average daily number of times that a medication was administered and the number of drugs with different presentation identifier codes per resident per month. Results 96,216 residents from 519 nursing homes were included between 1 January 2011 and 31 December 2014. The intervention group and the control group both decreased their average daily use of medication (− 0.05 and − 0.06). The booklet did not have a statistically significant effect (exponentiated difference-in-differences coefficient 1.00, 95% confidence interval 0.99–1.02, P = .45). Conclusion We observed an overall decrease in medication consumption in both the control and intervention groups. Our analysis did not provide any evidence that this reduction was related to the use of the booklet. Other factors, such as national policy or increased physician awareness, may have contributed to our findings.