Effects of state opioid prescribing cap laws on opioid prescribing after surgery

Abstract

Objective: To evaluate the effects of state opioid prescribing cap laws on opioid prescribing after surgery.

Data sources: OptumLabs Data Warehouse administrative claims data covering all 50 states from July 2012 through June 2019.

Study design: We included individuals from 20 states that had implemented prescribing cap laws without exemptions for postsurgical pain by June 2019 and individuals from 16 control states plus the District of Columbia. We used a difference-in-differences approach accounting for differential timing in law implementation across states to estimate the effects of state prescribing cap laws on postsurgical prescribing of opioids. Outcome measures included filling an opioid prescription within 30 days after surgery; filling opioid prescriptions of specific doses or durations; and the number, days' supply, daily dose, and pill quantity of opioid prescriptions. To assess the validity of the parallel counterfactual trends assumption, we examined differences in outcome trends between law-implementing and control states in the years preceding law implementation using an equivalence testing framework.

Data collection/extraction methods: We included the first surgery in the study period for opioid-naïve individuals undergoing one of eight common surgical procedures.

Principal findings: State prescribing cap laws were associated with 0.109 lower days' supply of postsurgical opioids on the log scale (95% Confidence Interval [CI]: -0.139, -0.080) but were not associated with the number (Average treatment effect on the treated [ATT]: -0.011; 95% CI: -0.043, 0.021) or daily dose of postsurgical opioid prescriptions (ATT: -0.013; 95% CI: -0.030, 0.005). The negative association observed between prescribing cap laws and the probability of filling a postsurgical opioid prescription (ATT: -0.041; 95% CI: -0.054, -0.028) was likely spurious, given differences between law-implementing and control states in the pre-law period.

Conclusions: Prescribing cap laws appear to have minimal effects on postsurgical opioid prescribing.

Keywords: causal inference; difference-in-differences; law; opioid; postsurgical pain.

FIGURE 1

Trends in the proportion of patients filling any postsurgical opioid prescription for each cohort from 4 years before to 1 year after law implementation. Data points represent the average value over the preceding 6 months.

FIGURE 2

Trends among patients filling any postsurgical opioid prescription in the proportion of patients filling a prescription with greater than 7 days' supply (Row 1) or a prescription with greater than 50 morphine milligram equivalents (MME) per day (Row 2) for each cohort from 4 years before to 1 year after law implementation. Data points represent the average value over the preceding 6 months.