Background: U.S. states have enacted various laws to curb opioid overprescribing, including laws requiring opioid prescribers to enroll in or query the prescription drug monitoring program (PDMP), “pill mill” laws regulating pain management clinics, and prescribing cap laws limiting the dose/duration of opioid prescriptions. Prior studies show mixed results, in part due to challenges disentangling these laws from other state laws designed to address the opioid crisis enacted around the same time.

Methods: We studied 16 U.S. states that implemented one of the laws of interest (4 states with each type of law: PDMP enrollment, PDMP query, pill mill, prescribing cap) and no confounding laws in a four-year period, two years pre/post law. We used difference-in-differences and augmented synthetic control analyses of Marketscan commercial claims data, which includes over 40 million U.S. adults, to conduct single-state analyses: each intervention state had its own study period and comparison group. Outcomes included opioid prescribing patterns, opioid overdose, and guideline-concordant non-opioid chronic pain treatment. We triangulated quantitative results with in-depth qualitative case studies (N=114 key informant interviews) of the 16 intervention states.

Results: Mandatory PDMP enrollment and query laws had no effects on outcomes. Pill mill laws were associated with 1-2 percentage-point reductions in high-dose (>90 MME) prescribing (p<0.05). Opioid prescribing cap laws were associated with small but statistically significant 2-3 percentage-point decreases in the proportion of people with chronic pain receiving opioid prescriptions and comparably sized increases in non-opioid therapies. None of the laws were associated with changes in opioid overdose. Qualitative results suggested that states lack systems to enforce opioid prescribing laws.

Conclusions: This study used a rigorous design-based strategy to isolate the independent effects of laws on outcomes. Findings suggest that state opioid prescribing laws, as implemented, have had minimal effects on opioid prescribing, overdose, and non-opioid chronic pain treatment.