A multifactorial intervention for hospital opioid management

Abstract

Objective: To determine whether an educational intervention combined with a voluntary decision support system improves inpatient pain control.

Design: Retrospective serial cross-sectional study.

Setting: Community teaching hospital.

Patients: Patients admitted to internal medicine teaching service from October to December 2011 and 2012. The study cohorts consisted of a random sample of 75 patients each from both time periods.

Interventions: Beginning in August 2012, internal medicine residents participated in an interactive training session on the use of opioids for hospitalized patients and concurrently, a user initiated voluntary computerized decision support system (CDSS), in the form of computer order entry (COE) and pocket cards were introduced. The COE options correspond to the standardized opioid dosing regimen on the pocket card. Pain scores and opioid doses and demographic information were obtained from administrative databases. Additional covariates were abstracted via programmed electronic medical record (EMR) review.

Main outcome measures: Pre- and postintervention, maximum reported pain score in every 8-hour period from first analgesic dose, to 72 hours after the first analgesic dose, were compared by fitting a multivariable linear mixed model. Naloxone use was a surrogate measure for secondary outcome of opioid overdose.

Results: The intervention had no effect on maximum pain score (MPS) over time, p = 0.0930. The estimated mean MPS (95% confidence interval) was 4.7 (3.9, 5.5) preintervention and 5.2 (4.4, 6.0) postintervention.

Conclusions: A combination of a resident educational intervention, CDSS, and pocket cards did not improve MPSs over time for patients on an internal medicine teaching service.