Opioid Use Disorders Among People with Severe Mental Illness

Background: The number of Americans with opioid use disorders (OUDs: prescription painkillers or heroin) has increased dramatically. However little is known about OUD among people with severe mental illness (SMI), even though people with psychopathology receive half of all opioid prescriptions.

Methods: Using the National Survey on Drug Use and Health (2006-2014: N=502,467), logistic regression was used to: (1) identify factors associated with past-year OUD among people with SMI; and (2) examine associations between OUD and adverse outcomes (psychiatric hospitalization, criminal justice system involvement, suicide attempt). Independent variables included sex, race/ethnicity, age, marital status, education, employment status, total family income, health insurance (yes/no), self-reported health, past-year DSM-IV major depressive episode, lifetime anxiety (yes/no), use of certain drugs before age 18, and ease of obtaining certain drugs.

Results: OUD was present in 5.0% (weighted) of adults with SMI and 0.7% (weighted) of adults without SMI. Adults with SMI were 7.67 times more likely (CI=6.61-8.91, p<.001) as adults without SMI to have OUD. After controlling for a number of factors, correlates of OUD among people with SMI included male gender, younger age, marital status (never been married), use of certain drugs before age 18 (especially marijuana), and ease of obtaining certain drugs.

After inclusion of control variables, men with past-year prescription painkiller use disorder (only) were 2.34 times more likely (CI=1.35-4.07, p<.01) to be hospitalized (psychiatrically) than those without substance use disorder, and men with heroin use disorder (only) were 4.58 times more likely (CI=1.34-15.71, p<.05). Among people with SMI, men with all other (i.e., non-OUD) substance use disorders were no more likely to be hospitalized than men without substance use disorder. Women with heroin use disorder (only) were 4.83 times more likely (CI=1.47-15.89, p<.05) to be hospitalized than those without substance use disorder, and women with all other substance use disorders (i.e., apart from OUD) were 80% more likely (OR=1.80, CI=1.32-2.46, p<.001).

People with prescription painkiller use disorder (only) were 7.43 times more likely (CI=4.55-12.14, p<.001) than those without substance use disorder to have criminal justice system involvement, while those with: (1) heroin use disorder (only) were 18.78 times more likely (CI=9.22-38.24, p<.001); (2) both prescription painkiller and heroin use disorder (only) were 25.83 times more likely (CI=14.06-47.47, p<.001); and (3) all other substance use disorders were 5.15 times more likely (CI=3.95-6.72, p<.001).

People with prescription painkiller use disorder (only) were 2.40 times more likely (CI=1.72-3.35, p<.001) to attempt suicide than those without substance use disorder, and those with all other substance use disorders (i.e., apart from OUD) were 79% more likely (OR=1.79, CI=1.45-220, p<.001).

Implications: This study appears to be among the first to research correlates of OUD among people with SMI, and among the first to assess adverse consequences that are associated with OUD among people with SMI (psychiatric hospitalization, criminal justice system involvement, and suicide attempt). The findings on OUD and OUD outcomes can help identify and understand individuals with SMI who could benefit from OUD prevention and intervention efforts.