Availability of Medications for Opioid Use Disorder in US Psychiatric Hospitals Authors: Shawn M. Cohen, Tamara Beetham, David A. Fiellin, Srinivas B. Muvvala Published: 2024 ABSTRACT This cross-sectional study examines the availability of medications for opioid use disorder in psychiatric hospitals in the US. Introduction In the midst of an unprecedented overdose crisis, medications for opioid use disorder (MOUD), namely methadone, buprenorphine, or naltrexone, remain the most effective treatment for OUD yet remain underutilized.1 Expanding availability of MOUD in all health care settings is a critical step in addressing the overdose crisis. Given the high comorbidity of psychiatric illness and OUD, access to MOUD within psychiatric treatment settings could have an outsized potential impact.2 We aimed to describe MOUD availability at psychiatric hospitals throughout the US. Methods This is a cross-sectional analysis of facility-level responses to the 2022 National Substance Use and Mental Health Services Survey (N-SUMHSS; 88% response rate).3 This study was considered nonhuman participants research by the Yale University institutional review board and did not require informed consent. The study sample included all facilities in the US that self-reported their facility type as a psychiatric hospital and reported their MOUD availability. MOUD availability was measured as a dichotomous variable indicating whether MOUD (ie, buprenorphine, methadone, and/or naltrexone) is a service psychiatric hospitals report providing. Wilcoxon rank-sum and χ2 tests were used to compare MOUD availability by facility characteristics. To assess for potential nonresponse bias on the MOUD provision rate, a Monte Carlo simulation with 1000 imputations was conducted modeling MOUD provision among nonrespondents based on characteristics associated with both MOUD provision and nonresponse. Analyses were conducted from February to September 2024 using Stata version 18.0 MP (StataCorp) with statistical significance defined as 2-tailed with P < .05. We followed the STROBE reporting guideline. Results Of 1107 psychiatric hospitals who responded to the survey, information on MOUD provision was available for 1021 (92.23% response rate) (Figure). Facilities were identified as inpatient psychiatric units within a general hospital (558 facilities [54.65%]), freestanding psychiatric hospitals (428 facilities [41.92%]), and state hospitals (35 facilities [3.43%]) (Table). Flow Diagram N-SUMHSS indicates National Substance Use and Mental Health Services Survey; MOUD, medications for opioid use disorder. Table. Psychiatric Hospital Characteristics and MOUD ProvisionaCharacteristicHospitals, No. (%)P valuebTotal (N = 1021)Offers MOUD (n =490)Facility type Freestanding psychiatric hospital428 (41.92)200 (46.73).49 Inpatient psychiatric unit of a general hospital558 (54.65)274 (49.1).44 State hospital35 (3.43)16 (45.71).78Ownership Private nonprofit organization519 (50.83)247 (50.2).03 Private for-profit organization290 (28.4)134 (47.69).80 Publicly ownedc212 (20.76)52 (35.86).002Bed capacity, No. (median) [IQR]937 (30) [18-70]417 (36) [20-75].003Payers accepted Medicare955 (93.54)477 (49.95)<.001 Medicaid968 (94.81)470 (48.55).13 Other state-financed health insurance679 (66.50)340 (50.07).06 Private insurance979 (95.89)479 (48.93).004 Military insurance700 (68.56)344 (49.14).28 Cash940 (92.07)462 (49.15).01Funding accepted County or local government327 (32.03)148 (47.28).68 State agenciesd599 (58.67)295 (49.25).34 Federally funded grantse132 (12.93)64 (48.48).90 Veterans Affairs412 (40.35)201 (48.79).68 Indian Health Services, tribal, or urban165 (16.16)78 (47.27).84 Private or community foundation34 (3.33)21 (61.76).10Age groups accepted Children (0-12 y)214 (21.27)102 (47.66).99 Adolescents (13-17 y)341 (33.43)174 (51.03).18 Young adults (18-25 y)872 (85.49)462 (52.98)<.001 Adults (26-64 y)932 (91.37)480 (51.50)<.001 Older adults (≥65 y)913 (89.51)469 (51.37)<.001Substance use services provided Integrated MH and SU treatment approach632 (62.27)358 (56.65)<.001 Co-occurring MH and SU group or program607 (60.10)344 (56.67)<.001 Medically managed withdrawal441 (45.09)308 (69.84)<.001 Medications for alcohol use disorder406 (41.51)373 (91.87)<.001 Individual counseling609 (62.40)333 (54.68)<.001 Group counseling717 (73.39)403 (56.21)<.001 12-Step groups213 (21.82)146 (68.54)<.001 Case management503 (51.54)271 (53.88)<.001Other services provided Housing services87 (8.58)49 (56.32).10 HIV or AIDS group or program193 (19.15)108 (56.67).01 Traumatic brain injury group or program129 (12.80)70 (54.26).11Geographic characteristic Medicaid expansion state737 (72.18)363 (49.25).19 US census region Midwest250 (24.49)117 (46.80).66 Northeast192 (18.81)110 (57.29).004 South390 (38.20)178 (45.64).24 West189 (18.51)85 (44.97).36Abbreviations: MH, mental health; MOUD, medications for opioid use disorder; SU, substance use. a Data are from the 2022 National Substance Use and Mental Health Services Survey. b χ2 and Wilcoxon rank-sum tests were used to compare MOUD availability by facility characteristics. c Publicly owned facilities are those owned by state, local, county, community, tribal, or federal entities. d State agencies include mental health, welfare or child and family services, state corrections and juvenile justice, education, and/or other state government. e Federally funded grants include community service block, community mental health block, and other federal grants. Psychiatric Hospital Characteristics and MOUD Provisiona Abbreviations: MH, mental health; MOUD, medications for opioid use disorder; SU, substance use. Data are from the 2022 National Substance Use and Mental Health Services Survey. χ2 and Wilcoxon rank-sum tests were used to compare MOUD availability by facility characteristics. Publicly owned facilities are those owned by state, local, county, community, tribal, or federal entities. State agencies include mental health, welfare or child and family services, state corrections and juvenile justice, education, and/or other state government. Federally funded grants include community service block, community mental health block, and other federal grants. Less than one-half of psychiatric hospitals (490 facilities [47.99%]) reported providing MOUD. Facilities were more likely to provide MOUD if they provided medications for alcohol use disorder (373 facilities [91.87%]; χ21 = 568.32; P < .001) or medically managed withdrawal services (308 facilities [69.84%]; χ21 = 173.94; P < .001); of 441 facilities offering medically managed withdrawal, 133 (30.16%) did not provide MOUD. Facilities that provided MOUD had greater median (IQR) bed capacity than the overall sample (30 [18-70] vs 36 [20-75]; P = .03), and MOUD provision was positively associated with being located in the Northeast (110 facilities [57.29%]; χ21 = 8.19; P = .004). Publicly owned psychiatric hospitals were less likely to provide MOUD (52 facilities [35.86%]; χ21 = 9.30; P = .002). We estimate an adjusted MOUD provision rate of 47.36% (SE = 1.56%) in our robustness check, indicating our reported 47.99% MOUD provision rate was unlikely to be affected by nonresponse bias. Discussion In this cross-sectional study, we found that less than one-half of psychiatric hospital survey respondents provide MOUD. It is additionally concerning that 30% of psychiatric facilities offering medically managed withdrawal did not provide MOUD given the possible harm of this approach including increased risk of overdose.4 Consistent with prior research in community outpatient mental health facilities,5 our findings suggest that low MOUD availability is a systematic issue in psychiatric care. Stigma against MOUD likely plays a role in these findings. Lack of knowledge about MOUD, its effectiveness, and regulations surrounding its provision also likely contributes. We suspect our main finding is an overestimate of true availability. Among facilities that reported providing MOUD, the survey did not specify which form of MOUD (methadone, buprenorphine, or naltrexone) was available or the frequency of availability. Availability of more regulated forms of MOUD, namely methadone and buprenorphine, may be lower. Additionally, providing MOUD conditionally or even for a single patient would allow a facility to meet criteria for provision without reflecting true availability. Limitations include potential upward bias from facilities self-reporting MOUD provision. Social desirability bias may contribute particularly because N-SUMHSS data are publicly available.6 This study exposes the deficiency in MOUD access at inpatient psychiatric hospitals and possible avenues to improve MOUD provision particularly to people with comorbid OUD and mental illness.