# Evaluation of a hospital-based opioid stewardship program on high-risk opioid prescribing in a Canadian setting: an interrupted time series analysis

## TL;DR

This study evaluated an opioid stewardship program in a Canadian hospital and found it reduced high-dose opioid prescriptions but had mixed effects on other prescribing indicators.

## Contribution

The study provides evidence on the effectiveness of an audit-and-feedback opioid stewardship program in a Canadian hospital setting.

## Key findings

- The program significantly reduced high daily dose opioid prescriptions but increased long-duration prescriptions.
- A benzodiazepine intervention reversed the rise in concurrent opioid-sedative prescriptions.
- The implementation of a new EHR system influenced some outcomes, highlighting the interplay between stewardship and technology.

## Abstract

High-risk opioid prescribing (e.g., high daily dose opioids, concurrent opioid-sedatives) is prevalent in hospitals and linked to adverse outcomes. Opioid stewardship programs (OSP) have the potential to reduce high-risk opioid prescribing through audit-and-feedback recommendations.

We evaluated an audit-and-feedback based OSP implemented in January 2020 at a Vancouver, Canada tertiary care hospital using interrupted time series analysis. An electronic health record (EHR) system with computerized provider order entry (CPOE) was simultaneously operationalized. The main outcome was: any high-risk opioid prescribing (based on 10 evidence-based indicators), including high daily dose of morphine milligram equivalent (MME) prescribing (> 90MME), long opioid prescription duration (> 5 days post-admission), and concurrent opioid-sedative prescribing.

Between January 2018 and March 2022, 5,477 active opioid patient encounters were included. While no significant change occurred in overall high-risk opioid prescribing post-OSP (p > 0.05), a significant reduction was seen in the level of high daily dose of MME prescriptions (estimate: -0.044; 95% confidence interval [CI]: -0.082, -0.006). Conversely, the trend in long opioid duration increased (estimate: 0.006; 95%CI: 0.000, 0.011), likely due to the removal of automatic stop dates with the implementation of the EHR with CPOE. Post-OSP intervention, we initially saw an acute increase in concurrent opioid-sedative prescriptions (estimate: 0.013; 95%CI: 0.005, 0.020). A benzodiazepine ordering intervention implemented in May 2021 reversed this trend, reducing both the level (estimate: 0.874; 95%CI: 0.374, 1.375) and slope (estimate: -0.022, 95%CI: -0.034, -0.011) of concurrent prescriptions.

The implementation of a new EHR concordant with that of the OSP may have impacted our study’s results. While our research suggests the OSP reduced high-dose opioid prescribing, other indicators impacted by the EHR system did not benefit as highly from the OSP. Nevertheless, the OSP proved able to rapidly respond to unintended consequences by introducing interventions to reduce concurrent opioid and sedative prescribing.

The online version contains supplementary material available at 10.1186/s13722-025-00574-x.