Combining prescription monitoring, benchmarking, and educational interventions to reduce benzodiazepine prescriptions among internal medicine inpatients; a multicenter before and after study in a network of Swiss Public Hospitals.

OBJECTIVES

Reducing the inappropriate benzodiazepine (BZD) prescriptions represents a challenge for health care systems worldwide. The 'Choosing Wisely' campaign recommends against the use of BZD in the elderly as the first choice for insomnia, agitation, or delirium. We aimed to determine whether a transparent monitoring-benchmarking together with educational interventions, on top of the internal publication of a targeted recommendation, could be effective in curbing BZD prescriptions.

METHODS

Multicenter before and after study in a network of five southern-Switzerland teaching hospitals. An intervention based on a transparent continuous monitoring-benchmarking system, called 'Reporting Wisely', able to collect, analyze, and report data on BZD prescriptions and educational interventions focused on themed meetings, audit, and feedback, was implemented. The intervention was limited to the Internal Medicine. The impact of the intervention on new BZD prescriptions and de-prescribing at hospital discharge, was assessed using segmented regression analyses of interrupted time-series and comparing Internal Medicine to Surgery.

RESULTS

Between July 12014, and June 302017, data of 45,597 hospital admissions, from Internal Medicine and Surgery departments were analyzed. Before the intervention (July 12014 to December 312015), the mean monthly new BZD prescription rate was 7.2%; value dropping to 5.5% (24% relative reduction; p < 0.001) in the intervention phase (January 12016 to June 302017). At the end of the intervention a 15% relative increase of BZD de-prescribing was also found (p < 0.01). The use of atypical antipsychotic (AAP) and other potentially harmful sedative drugs did not increase. In the surgery department, exposed to the recommendation but not to the intervention, a constant upward trend with a slope of 0.129 new prescriptions per 100 admissions per month (95% CI 0.08-0.17; p < 0.001) was seen.

CONCLUSIONS

The implementation of a dual intervention based on transparent monitoring-benchmarking and multidisciplinary education has proved useful in curbing new BZD prescriptions and in promoting BZD de-prescribing in the hospital setting.