EMR quantity autopopulation removal on hospital discharge prescribing patterns: Implications for opioid stewardship

Prescription drug stewardship is critical. Autopopulation(AP) of medication quantities may influence prescriber behaviour. We investigate the impact of AP removal(APR) on opioid prescribing.

Inpatient and emergency department(ED) discharges with opioid pain medications 2 years before and after APR were identified. Milligrams of morphine equivalents(MMEs) prescribed were recorded. Group comparisons were performed using Mann-Whitney U tests. Spearman's rho was used to analyse correlations between pain level and quantity of prescribed opioids. Mann-Kendall tests assessed trends in prescription patterns. Generalized estimating equations assessed trends in total quantity of prescribed MME.

A total of 53 608 patient encounters were included for analysis. In surgical patients, there were no trends in the frequency of prescriptions below, at or above the AP quantity pre-APR. Post-APR, there was a decrease in the percentage of prescriptions written for the AP quantity(τ = -.493, P = .001) and an increase in prescriptions for <30 tablets(τ = .468,P = .001). In non-operative patients, the pre-APR period was associated with a lower percentage of prescriptions >30 tablets and a greater percentage of prescriptions for <30 tablets. Interestingly, APR reversed this trend in prescriptions for >30 tablets and resulted in an increase in larger prescriptions. Multivariate analysis of the total prescribed quantity of MME found no significant trend across months for inpatients prior to and after APR (0.997, P = .065 and 1.003, P = .142; respectively). The ED model found a monthly downward trend in amount of prescribed MME prior to and after APR (0.986, P < .001 and 0.990, P < .001; respectively). In the inpatient setting, pain was positively correlated to discharge MME (ρ = .028, P < .001); with those reporting the highest pain receiving the greatest amount of opioids both pre- and post-APR. Interestingly, in the ED, this finding was negatively correlated (ρ = -.086, P < .001); with those reporting the lowest pain receiving the greatest amount of opioids both pre- and post-APR.

AP removal may have unintended consequences, such as increased prescriptions for greater quantities. To drive down prescription amounts, lower anchor values may be of more utility than APR. The poor correlation of pain values with prescribed medications warrants further investigation.