Risk of Inpatient Stay for Mental Illness Among Individuals With Substance Use Disorders

The NSDUH is a cross-sectional survey that measures prevalence and correlates of drug use among the general population of noninstitutionalized persons in the United States. In the 2007 NSDUH, the full sample comprised 67,870 persons, a number that was reduced to 55,435 persons in the public use file. After exclusion of adolescents ages 12 to 17 years and 54 adults without data on hospitalization, the final sample comprised 37,654 persons. The survey is nationally representative, and data are weighted in order to permit national estimates of substance abuse and associated factors.

The survey asked respondents if they had stayed overnight or longer in a hospital or other facility for treatment or counseling for any problem with emotions, nerves, or mental health in the past 12 months. They were asked not to include overnight stays for alcohol or drug use.

Cross-tabulations and chi square tests of association were used to explore bivariate relationships between past-year hospitalization and each covariate listed in Table 1. Logistic regression adjusting for all of the covariates was used to predict hospitalization associated with specific types of substance abuse. By using complex survey design measures, all analyses applied sampling weights to the data to adjust for nonresponse and for varying probabilities of selection, including those resulting from oversampling.

The sample is described in Table 1. Just under 1% (.9%) of American adults had been hospitalized for mental illness in the past year. By comparison, 3.3% of people with substance use disorders and .7% of people without substance use disorders had an inpatient stay for mental illness. After adjustment for all covariates, the analysis found that people with substance use disorders were 1.71 times more likely to be hospitalized (95% confidence interval [CI]=1.10–2.65, p<.05) than those without substance use disorders. [A complete description of the sample for all covariates is available online as a data supplement to this report.].

After adjustment for all covariates, the study found that people with cocaine abuse or dependence only were 2.65 times as likely, and people with opioid analgesic abuse or dependence only were 6.85 times as likely, to be hospitalized than people without substance use disorders. About 8% of people with cocaine abuse or dependence only and over 15% of people with opioid analgesic abuse or dependence only were hospitalized in the past year, compared with about 2% of people with alcohol or abuse or dependence only, for example, and about 2% of people with marijuana abuse or dependence only.

Hospitalization rates of people with two or more co-occurring substance use disorders differed. After inclusion of the control variables, the analysis found that people with alcohol and marijuana abuse or dependence only were 3.1 times more likely to be hospitalized than individuals without substance use disorders. In addition, persons with alcohol and cocaine abuse or dependence only were 2.58 times more likely to be hospitalized than individuals without substance use disorders. About 6% of people with alcohol and marijuana abuse or dependence only and 11% of people with alcohol and cocaine abuse or dependence only were hospitalized in the past year.

Even after adjustment was made for a wide variety of sociodemographic and clinical variables, adults with substance use disorders were 71% more likely than other adults to be hospitalized in the past year. This finding underscores a well-known association between substance abuse and psychiatric hospitalization (1–3). Other results were more striking. Persons who reported abuse or dependence involving opioid analgesics had a 700% increase in adjusted risk of hospitalization compared with persons without substance use disorders. Persons who reported abuse or dependence involving cocaine only, alcohol and cocaine only, or alcohol and marijuana only had about a 300% increase in adjusted risk of hospitalization compared with persons without substance use disorders.

Despite these findings, it may be premature, at least in the absence of further research, to conclude that mental health and substance abuse professionals should focus especially on abuse or dependence of opioid analgesics, cocaine, alcohol and cocaine, and alcohol and marijuana to help prevent hospitalization. However, conclusions related to research methodology and policy implications may be reached with greater certainty. More specifically, this study showed that with adequately sized samples and detailed information on substance abuse, it is possible for research to isolate the specific types of substance use disorders that most increase the risk for one of the highest thresholds of mental illness severity (psychiatric hospitalization). This information can be used to more efficiently direct scarce resources to people with substance use disorders who are most likely to need hospitalization. Mental health treatment programs, substance abuse treatment programs, programs that address both mental illness and substance abuse, and prevention programs for the general population could develop greater flexibility in detecting and treating the drug-specific disorders that are most highly associated with inpatient stay.

Particularly if other studies substantiate these findings, the four types of high-risk substance abuse must be understood and, perhaps, addressed separately. More research is needed on the reasons why hospitalization appears to be related to the combination of alcohol and marijuana abuse or dependence. People who abuse both substances tend to drink more and to have more alcohol-related problems in areas other than inpatient stay, such as driving impairment or criminal justice system involvement, than individuals who abuse only alcohol (5–8). Similarly, the significant association between hospitalization and co-occurring alcohol and cocaine abuse or dependence may be related to the body's production of cocaethylene when metabolizing both substances. Cocaethylene is more toxic than either drug alone (9–12).

The link between hospitalization and opioid analgesic abuse or dependence contributes to a growing literature (13) on the negative outcomes of a progressively prevalent problem. In much the same way, the findings show that the negative consequences of cocaine addiction, already understood to include depression and mental health problems more generally (14), also include hospitalization.

It was not possible by using NSDUH data to compare hospitalization predictions by type of mental illness, although high levels of psychological distress on the Kessler-6 Scale for Psychological Distress (15) and past-year major depressive episode were used to adjust for presence of mental illness. In addition, for some respondents past-year hospitalization may have preceded development of past-year substance abuse; however, substance use disorders that meet DSM-IV criteria typically are long-standing.

Even after inclusion of control variables, adults with four types of substance use disorders were more likely than adults without substance use disorders to be hospitalized. High-risk disorders included abuse of or dependence on opioid analgesics, cocaine, alcohol and cocaine, and alcohol and marijuana. Efforts to prevent inpatient stays may wish to target individuals with high-risk substance use disorders.

The author reports no competing interests.