Table. Changes in Opioid Prescribing After Implementation of Laws Restricting Prescribing to 7 Days or Less.

Days of any opioid prescription	Control states (n = 18)			States exposed to policy (n = 23)			Difference-in-differences estimate (95% CI)a
	2013	2018	Difference	2013	2018	Difference
Per enrollee	43.43	33.33	−10.1	44.23	32.66	−11.57	−1.75 (−2.87 to −0.62)
Per enrollee by specialty
Primary care	25.29	16.3	−8.99	24.07	14.39	−9.68	−0.26 (−1.05 to 0.53)
Surgeon or dentist	9.65	9.44	−0.21	11.79	11.2	−0.59	−0.90 (−1.37 to −0.42)
Emergency medicine	0.73	0.34	−0.39	0.51	0.22	−0.29	0.06 (−0.02 to 0.13)
Pain specialist	2.68	3.23	0.55	2.98	3.51	0.53	−0.45 (−0.73 to −0.17)
Other specialistb	4.11	3.49	−0.62	3.98	2.85	−1.13	−0.29 (−0.50 to −0.09)
Trainee/midlevel/institutionc	0.98	0.54	−0.44	0.9	0.48	−0.42	0.10 (0.02 to 0.17)

^a Estimates are from linear probability models adjusting for race, ethnicity, urbanization, income, tobacco use, alcohol use, serious mental illness, region, year, and state-level fixed effects. The steeper reduction of days of opioid prescription per enrollee in states with laws restricting opioid prescribing was primarily due to relative differences in prescribing among surgeons and dentists, pain specialists, and other specialists.

^b Other specialists included 23 specialties that did not fit into the other groups.

^c Trainee/midlevel/institution included trainees, physician assistants, nurse practitioners, and institutions.