Addressing an epidemic: Improving guideline-concordant opioid prescribing in surgical patients.

BACKGROUND

Excess postoperative opioid prescribing increases the risk of opioid abuse, diversion, and addiction. Clinicians receive variable training for opioid prescribing, and despite the availability of guidelines, wide variations in prescribing practices persist. This quality improvement initiative aimed to assess and improve institutional adherence to published guidelines.

METHODS

This study represented a quality improvement initiative at an academic medical center implemented over a 6-month period with data captured 1 year before and after implementation. The quality improvement initiative focused on prescribing education and monthly feedback reports for clinicians. All opioid-naïve, adult patients undergoing a reviewed procedure were included. Demographics, surgical details, hospital course, and opioid prescriptions were reviewed. Opioids prescribed on discharge were evaluated for concordance with recommendations based on published guidelines. Pre- and postimplementation cohorts were compared.

RESULTS

There were 4,905 patients included: 2,343 preimplementation and 2,562 postimplementation. There were similar distributions in patient demographics between the 2 cohorts. Guideline-concordant discharge prescriptions improved from 50.3% to 72.2% after the quality improvement initiative was implemented (P < .001). Adjusted analysis controlling for sex, age, discharge clinician, length of stay, outpatient surgery, and procedure demonstrated a 190% increase in odds of receiving a guideline-concordant opioid prescription on discharge in the postimplementation cohort (adjusted odds ratio 2.90; 95% confidence interval = 2.55-3.30).

CONCLUSION

This study represented a successful quality improvement initiative improving guideline-concordant opioid discharges and decreasing overprescribing. This study suggested published guidelines are insufficient without close attention to elements of effective change management including the critical importance of locally targeting educational efforts and suggested that real-time, data-driven feedback amplifies impact on prescribing behavior.