Polysubstance Use Disorders in Four State Medicaid Programs Authors: Julie M. Donohue, PhD; Susan Kennedy, MPP, MSW; Logan Sheets Introduction Methods and Approach In 2019, over 71,000 Americans died from a drug overdose and, This project leveraged AcademyHealth's MODRN, a multi-state with the onset of the COVID-19 pandemic in 2020, opioid-related research collaborative, founded by members of the State-Univer- deaths have climbed to a new high.1,2 As states grapple with the sity Partnership Learning Network and the Medicaid Medical opioid epidemic amid a pandemic, the rising prevalence of other Directors Network. Comprising 13 state Medicaid and public drug use is complicating efforts to treat opioid use disorder and university research partners, MODRN enables timely, standard- prevent overdose deaths. Recent research suggests that more than ized analyses of states' Medicaid data to address issues of na- 30% of opioid deaths also involved benzodiazepines.3 In addition, tional public health importance, such as the opioid crisis, while the concurrent use of methamphetamines and opioids doubled estimating the impact of innovative policies and interventions. from 2011 to 2017.4 A study using 2016-2017 data reported that AcademyHealth serves as MODRN's administrative coordinating approximately half of Medicaid enrollees with opioid use disorder center and works closely with the data coordinating center at the had a comorbid other substance use disorder.5 Yet, little is known University of Pittsburgh. Together, they manage the core func- about recent trends in multiple or polysubstance use disorders tions of MODRN to support critical multi-state Medicaid analyses among Medicaid enrollees or the associated consequences for on OUD treatment and outcomes. medical conditions including those common among people with injection drug use. Improving understanding of substance use The University of Pittsburgh, as MODRN's data coordinating cen- disorder comorbidity burden can help states target interventions ter, distributed standardized code to the state university partners to reduce overdose deaths. who returned aggregate results, not individual data, to the data coordinating center for statistical analyses and reporting. The With funding from the Medicaid and CHIP Payment and Access model enabled efficient, standardized analyses of multiple states' Commission (MACPAC), AcademyHealth, in collaboration with Medicaid data while ensuring the security of health information. the University of Pittsburgh, Ohio State University, West Virginia University, and the University of Maryland, Baltimore County The project analyzed Medicaid enrollment, claims and encounter conducted an analysis of four states (MD, OH, WV, PA) partici- data from MD, OH, PA, and WV for the period of January 1, 2016 pating in the Medicaid Outcomes Distributed Research Network to December 31, 2018. The study population included all full-ben- (MODRN). This brief examines the prevalence of diagnoses of efit, non-dually eligible Medicaid enrollees, age 12 to 64 years polysubstance use disorders, describes the characteristics of Med- of age. Seven broad categories of substance use disorders were icaid enrollees with polysubstance use disorders and compares the identified using ICD-10 diagnoses in physical or behavioral health mental health and medical comorbidities based on the number of claims in Medicaid including opioid use disorder; alcohol use dis- unique substance use disorders. order; cannabis use disorder; cocaine use disorder; other psycho- active substance use disorder; amphetamine-type stimulant use disorder (with and without diagnoses for methamphetamine-re- 1.Katz, J., Goodnough, A., & Sanger-katz, M. (2020, July 15). In shadow of pandemic, U.S. drug overdose DEATHS RESURGE to record. Retrieved December 17, 2020, from https://www.nytimes.com/ interactive/2020/07/15/upshot/drug-overdose-deaths.html 2.National Institute on Drug Abuse. (2021, February 25). Overdose death rates. Retrieved December 17, 2020, from https://www.drugabuse.gov/drug-topics/trends-statistics/overdose-death-rates 3. McClure, F. L., Niles, J. K., Kaufman, H. W., & Gudin, J. (2017). Concurrent use of opioids and benzodiazepines: evaluation of prescription drug monitoring by a United States Laboratory. Journal of addiction 4.Ellis, M. S., Kasper, Z. A., & Cicero, T. J. (2018). Twin epidemics: the surging rise of methamphetamine use in chronic opioid users. Drug and alcohol dependence, 193, 14-20. 5.O'Brien, P., Henke, R. M., Schaefer, M. B., Lin, J., & Creedon, T. B. (2020). Utilization of treatment by Medicaid enrollees with opioid use disorder and co-occurring substance use disorders. Drug and Alcohol Dependence, 217, 108261. This report was prepared under contract to the Medicaid and CHIP Payment and Access Commission (MACPAC). The findings, statements, and views expressed in this report are those of the authors and do not necessarily reflect those of MACPAC. Polysubstance Use Disorders in Four State lated poisonings); and an 'other substance use disorder' group that OUD was the most prevalent substance use disorder in each of included several low-prevalence conditions (e.g., sedative/hypnot- the four states, however, the next most common substance use ic/anxiolytic related disorders, hallucinogen-related disorders, any disorder varied across the states (figure 2). In two states, the next pregnancy related substance use disorders, inhalant-related dis- most common was alcohol use disorder. In one state the next orders). We constructed measures at the person-year-level of the most common was cannabis use disorder and in another it was count number of unique substance use disorders (using the seven other psychoactive substance use disorder. categories defined above) coded as one, two, or three or more. Figure 1 – In 2018, 4 in 10 Medicaid enrollees with SUD had two or more. Analyses compared the characteristics of Medicaid enrollees by substance use disorder burden including age, sex, race, eligibility Prevalence of SUD category, urban/rural residence, and several comorbidities (hu- man immunodeficiency virus, hepatitis C virus, hepatitis B virus, anxiety, depression, post-traumatic stress disorder, schizophrenia and other psychotic disorders). 2 SUDs Where the prevalence was similar across states, pooled results are pre- sented. For results with between-state variation we presented results stratified by state with state identities masked for reporting purposes. 3 or more 1 SUD SUDs Key Findings Across the four states, the pooled prevalence of any substance use disorder increased from 11.2% of adolescent and adult Medic- aid enrollees to 11.8% from 2016 to 2018 (Figure 1). In 2018, among enrollees with any substance use disorder, 59.1% had one disorder, 21.5% had two disorders, and 19.4% had three or more 1 SUD (59.1%) 2 SUDs (21.5%) 3 or more SUDs (19.4%) substance use disorders, which was similar to prior years. Figure 2 - Prevalence of specific substance use disorders varies by type and state 2018 Prevalence of SUD by Type (%) State A State B State C State D OUD Alcohol Cannabis Cocaine Other psychoactive substance Amphetamine-type stimulant (+OD) Amphetamine-type stimulant Other SUDs .0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 Prevalence among SUD Population (%) 2 Polysubstance Use Disorders in Four State Table 1 –Characteristics of Medicaid enrollees by (54.9%) or three (58.8%) compared to one (50.2%) substance use number of substance use disorders, 2018 data disorders. Enrollees with multiple substance use disorders were slightly more likely to live in rural areas. pooled across four states Table 1 compares the demographic and health status characteristics The burden of mental health comorbidities was highly correlated with of enrollees with one, two, or three or more diagnosed substance the number of substance use disorders. Among enrollees with three use disorders in 2018. Enrollees with polysubstance use disorders or more substance use disorders, 59.0% and 66.4% had anxiety and were more likely to be aged 21-34 or 35-44 compared to enrollees mood disorders, respectively, compared to only 36.2% and 39.9% of with one substance use disorder. Enrollees with multiple substance those with one substance use disorder. The prevalence of schizophre- use disorders were also more likely to be male. Enrollees with three nia and other psychotic disorders was 19.3% among enrollees with or more substance use disorders were more likely to be non-His- three or more substance use disorders compared to only 6.7% among panic White compared to those with one substance use disorder those with one. Similarly, the prevalence of post-traumatic stress (66.7% vs. 62.2%) and less likely to be in a racial/ethnic minority disorder in those two groups was 19.4% and 7.9%, respectively. group. Expansion eligibility was more likely among those with two Pooled characteristics of Medicaid enrollees with one, two, or three or more SUDs, 2018 Characteristics Number of SUDs 1 SUD 2 SUDs 3 or more SUDs Overall (row %) 59.1 21.5 19.4 12-17 (column %) 3.8 2.6 1.2 18-20 4.6 3.4 2.4 21-34 36.1 39.7 43.8 Age 35-44 23.3 24.4 26.4 45-54 18.0 17.8 17.0 55-64 14.3 12.1 9.3 Female 47.8 43.4 41.8 Gender Male 52.2 56.6 58.2 Non-Hispanic White 62.2 63.5 66.7 Non-Hispanic Black 26.6 25.6 23.3 Race/Ethnicity Hispanic 4.3 4.1 3.6 Others 6.8 6.8 6.3 Pregnant Women 6.7 5.9 5.6 Children 7.5 5.7 3.3 Eligibility Disabled Adults 17.7 17.9 18.4 Non-Disabled Adults 17.8 15.6 13.8 Expansion Adults 50.2 54.9 58.8 missing category 0.4 0.3 0.3 Living Area Urban 83.2 82.4 82.4 Rural 16.4 17.3 17.9 3 Polysubstance Use Disorders in Four State Pooled characteristics of Medicaid enrollees with one, two, or three or more SUDs, 2018 Number of SUDs 1 SUD 2 SUDs 3 or more SUDs Overall, row % 59.1 21.5 19.4 Anxiety disorder, column (%) 36.2 45.3 59.0 Mood disorder 39.9 51.3 66.4 Schizophrenia and other psychotic disorders 6.7 10.8 19.3 Post Traumatic Stress Disorder (PTSD) 7.9 11.9 19.4 Comorbidities Hepatitis C (HCV) 7.1 13.3 24.5 Human Immunodeficiency Virus (HIV) 1.3 1.9 2.9 Hepatitis B (HBV) 0.6 1.1 2.3 Abscess 0.1 0.2 0.7 Osteomyelitis 0.5 1.1 2.0 Endocarditis 0.2 0.6 1.8 Soft skin tissue infections 10.9 15.0 22.1 The prevalence of medical comorbidities common among those states' focus on opioid use disorder is understandable given that with injection drug use was also markedly higher with increased it remains the most prevalent of all diagnosed substance use number of substance use disorders. The prevalence of hepatitis C disorders. We also found that four in ten Medicaid enrollees with virus was three times higher, osteomyelitis was four times higher, substance use disorders have polysubstance use disorders and and endocarditis was nine times higher among those with three that these enrollees had increased risk of several mental health or more substance use disorders compared to those with a single and medical comorbidities. The complexity of treating Medicaid diagnosed substance use disorder. enrollees with polysubstance use disorders, mental health condi- tions, infectious diseases and medical complications common in Conclusions those with injection drug use has implications both for Medicaid Medicaid officials have made concerted efforts to improve access expenditures, and for the way states design delivery systems for to high-quality treatment for opioid use disorder in recent years. this vulnerable population. Findings from analyses of four states' Medicaid data suggest that 4