Table 2 Claims-based measures and definitions for study outcomes

Supply measures

Percent of physicians certified to prescribe buprenorphine

Number of physicians on SAMHSA list of buprenorphine or naltrexone prescribers/# Medicaid-participating physicians in county

Percent of primary care practices delivering MAT

Number of physicians with any prescribing of buprenorphine or naltrexone/# Medicaid-participating physicians in county measured in claims data

Prevalence measures

OUD diagnoses

OUD ICD-9/ICD-10 code on inpatient, outpatient or professional claims

Overdose events

Inpatient stays or ED visits with an opioid overdose ICD-9/ICD-10 code

OUD diagnoses among those with evidence of misuse of prescription opioids

Rate of OUD diagnosis among those with opioid misuse using an algorithm from Sullivan et al. [46] that uses # prescribers, # pharmacies, and # days short- and long-acting opioids

Utilization measures

Use of MAT

Any prescription fill for buprenorphine or IM naltrexone among those with OUD

Use of psychosocial supports

Any visit with a service code for psychosocial support for OUD

Access to counseling for OUD

Any visit with a service code for counseling regarding psychosocial and pharmacologic treatment options among those with OUD

Duration of MAT

Number of months with proportion of days covered > 80%

Access to tele-psychiatry

Any visit with a service code for a tele-psychiatry visit

Outcome measures

Likelihood of any and SUD or mental health-related emergency department visit

SUD or mental health ICD-9/ICD-10 code as primary diagnosis for an emergency department visit

Likelihood of any and SUD or mental health-related inpatient hospitalization

SUD or mental health ICD-9/ICD-10 code as primary diagnosis for an inpatient hospitalization