Finding Default in the Opiate Crisis

As the U.S. opiate crisis continues, opiate prescribing patterns in all settings, including the ED, have come under close inspection. One possible strategy for decreasing ED opiate prescriptions calls for eliminating electronic order entry defaults for the number of opiate tablets prescribed. 

This retrospective observational study in a single academic, urban ED (58,000 annual visits) compared the number of tablets prescribed for the 12 months before and 10 months after the removal of a default quantity (20 tablets) for opioid prescriptions from the ED’s electronic order entry. 

Investigators found that after removing the default, the median number of opioid tablets prescribed fell from 20 to 15 tablets. While 20 tablets remained the most frequently prescribed amount in both time periods, the percent of patients receiving 20 tablets dropped from 50% to 23% post-intervention, translating to a wider distribution in the number of tablets given.

Considering these results, electronic health records (EHR) and defaults appear to strongly effect the number of tablets prescribed. Eliminating defaults forces providers to choose the number of pills given, and in this study was associated with a significant decrease in opiate prescribing. Yet this finding is not universal; another study in a different ED implementing the same intervention witnessed increased opiate prescribing. 

Alternatively, one may argue that lower defaults (10 tablets) would achieve greater reductions. However, one study implementing a default of 10 tablets for ED opiate prescriptions demonstrated an unintended drop in prescriptions for less than 10 tablets, undermining the authors’ goal. Moreover, lower defaults may inadvertently lead prescribers to undertreat certain patients’ pain. 

Ultimately, administrators may consider a compromise wherein the EHR gives no default, but alerts the physician when prescribing more than a typical amount (i.e. Alert would read “Clinical Guidelines recommend a maximum of 3-5 days’ worth of opiate medications. The current prescribed amount is over this recommendation.”) The provider may then re-think their outlier behavior, but can override if needed. Educating ED providers on opiate prescribing guidelines has been shown to sustainably alter prescriber behavior. Such a strategy would leverage technological tools and reduce outlier prescribing practices, while respecting physician autonomy.

This Health Policy Journal Club review is a collaboration between Policy Prescriptions® and the Emergency Medicine Residents’ Association. It is written by Marisa K. Dowling, MD, MPP. She is a third year emergency medicine resident at University of Maryland Medical Center.

Abstract

INTRODUCTION: Opioid prescribing patterns have come under increasing scrutiny with the recent rise in opioid prescriptions, opioid misuse and abuse, and opioid-related adverse events. To date, there have been limited studies on the effect of default tablet quantities as part of emergency department (ED) electronic order entry. Our goal was to evaluate opioid prescribing patterns before and after the removal of a default quantity of 20 tablets from ED electronic order entry.

METHODS: We performed a retrospective observational study at a single academic, urban ED with 58,000 annual visits. We identified all adult patients (18 years or older) seen in the ED and discharged home with prescriptions for tablet forms of hydrocodone and oxycodone (including mixed formulations with acetaminophen). We compared the quantity of tablets prescribed per opioid prescription 12 months before and 10 months after the electronic order-entry prescription default quantity of 20 tablets was removed and replaced with no default quantity. No specific messaging was given to providers, to avoid influencing prescribing patterns. We used two-sample Wilcoxon rank-sum test, two-sample test of proportions, and Pearson’s chi-squared tests where appropriate for statistical analysis.

RESULTS: A total of 4,104 adult patients received discharge prescriptions for opioids in the pre-intervention period (151.6 prescriptions per 1,000 discharged adult patients), and 2,464 post-intervention (106.69 prescriptions per 1,000 discharged adult patients). The median quantity of opioid tablets prescribed decreased from 20 (interquartile ration [IQR] 10-20) to 15 (IQR 10-20) (p<0.0001) after removal of the default quantity. While the most frequent quantity of tablets received in both groups was 20 tablets, the proportion of patients who received prescriptions on discharge that contained 20 tablets decreased from 0.5 (95% confidence interval [CI] [0.48-0.52]) to 0.23 (95% CI [0.21-0.24]) (p<0.001) after default quantity removal.

CONCLUSION: Although the median number of tablets differed significantly before and after the intervention, the clinical significance of this is unclear. An observed wider distribution of the quantity of tablets prescribed after removal of the default quantity of 20 may reflect more appropriate prescribing patterns (i.e., less severe indications receiving fewer tabs and more severe indications receiving more). A default value of 20 tablets for opioid prescriptions may be an example of the electronic medical record’s ability to reduce practice variability in medication orders actually counteracting optimal patient care.

PMID: 29560071

Santistevan JR, et al. West J Emerg Med. 2018; 19 (2): 392-397.