Use of Psychotropic Medications Among HIV-Infected Patients in the United States

To our knowledge, this is the first national probability study to examine the use of different types of psychotropic medications by HIV-positive patients. The results indicate that 27.2% of HIV-positive patients who received medical care in 1996 were treated with psychotropic medications in the 6 months before the study interview. This rate is consistent with that obtained through the baseline assessment of mental health services in the HIV Cost and Services Utilization Study sample (11), which was conducted 6 months before the survey of psychotropic use here reported. Antidepressants were the most commonly used psychotropic, followed by anxiolytics and, much less frequently, antipsychotics and psychostimulants. Among antidepressants, SSRIs were more commonly used than tricyclics. Most patients taking psychotropics took more than one class of medication. Comorbidity of mental disorders was associated with greater drug use. These results are generally consistent with previous reports regarding psychotropic prescribing patterns among patients in primary care and psychiatric practice settings (21, 22).

The estimated rates of mental health disorders during the 6 months before the interview are lower than the 12-month prevalence rate previously reported for the HIV Cost and Services Utilization Study sample (3). Several factors may have accounted for these lower rates, such as the shorter time frame (6 months), use of the long form of the Composite International Diagnostic Interview, which may be more specific than the short form used by Bing et al. (2, 23), and the repeated interviewing of the same patients with consequent possible regression to the mean and diagnostic attenuation.

We found that 43.9% of the HIV-positive patients with a DSM-IV diagnosis of major depression supported by the Composite International Diagnostic Interview took antidepressant medication. This rate is somewhat higher than expected on the basis of previous reports that only one-half of the primary care patients with major depression are accurately diagnosed (5, 22) and only one-half of those diagnosed are actually treated with antidepressants (24, 25). Some of these studies, however, were conducted in the mid-1980s, and there are indications that the rate of treatment of depression markedly increased in the 1990s (26). HIV-positive patients may be also more likely to be treated for psychiatric disorders, perhaps because of the extensive medical contacts related to HIV infection. More than one-half of the patients with a diagnosis of major depression under care for HIV infection reported not having been treated with antidepressants. How much this finding may be due to failure to recognize depression, failure to treat a diagnosed disorder, concern about polypharmacy and drug interactions in this population, or a preference for nonpharmacological interventions cannot be fully disentangled based on these data. On the basis of previous reports (5, 22), failure to recognize the disorder is rather common. However, a recent report that only about one-third of patients seeking treatment for mood or anxiety disorders may receive appropriate care (13) suggests that other causes besides failure to diagnose contribute to undertreatment. There are also indications that most primary care patients with depression prefer psychotherapy over medications (27). In our study, the rate of use of psychotherapy among depressed patients not receiving antidepressant medication was 25.2% for whites, 31.9% for African Americans, and 35.5% for Hispanics. The survey used a broad definition of “psychotherapy,” encompassing any contact with a mental health provider for individual or group therapy (including family and marriage counseling) and attendance at support group meetings. Even after an accounting for use of these psychosocial interventions, it appears that at least one-third of all patients with a mood disorder did not receive treatment.

Antidepressants are effective for both depressive and anxiety disorders, and this can explain the high use of these drugs by patients with anxiety disorders (Table 2). An estimated 12.3% of the patients with no mental disorder, based on the Composite International Diagnostic Interview, received antidepressant medication (Table 1). Given that 29.1% of all of the patients had a mental disorder and only 41.7% of them received antidepressants, it can be inferred that a sizable amount (41.8%) of all antidepressant use was by patients who did not have a diagnosis of a mental disorder according to the long-form Composite International Diagnostic Interview during the last 6 months. It is possible that a mental disorder had been previously present and that the medication was used to prevent recurrence (28). However, the average duration of antidepressant treatment is less than 3 months, and most patients discontinue medication within 6 weeks of starting treatment (28–30). Antidepressants could have been prescribed to treat subsyndromal forms of depression or anxiety (31) or to manage chronic pain (32). The study interview inquired about “medications for personal or emotional problems, such as emotions, nerves, alcohol, drugs, or mental health,” but the relationship between pain and depression is complex, and the two conditions can overlap (33). Likewise, anxiolytics are often used for managing insomnia, a common symptom in HIV-positive patients (34).

Use of multiple psychotropics was common. Although it cannot be ruled out that some medications were taken sequentially during the 6-month period, concomitant use was probably high. As expected, the concomitant use of antiretroviral and psychotropic medications was also common, with more than one-fifth of the patients receiving highly active antiretroviral therapy also taking an antidepressant. The study of the possible pharmacological interactions between antiretroviral and psychotropic medications has high clinical relevance (35). For instance, protease inhibitors can inhibit several cytochrome P-450 enzymes involved in the metabolism of psychotropics (36), resulting in a higher plasma concentration of these drugs (37).

After control for mental disorder, gender, age, insurance status, and presence of AIDS, African American patients were significantly less likely to have received psychotropic medications, and antidepressants in particular, than whites and Hispanics (Table 3 and Table 4). This finding is consistent with previous reports of less use of health care services for HIV infection by African Americans (15) and for mood and anxiety disorders in the general population (13, 38). This lower use of psychotropics by African Americans may be partly compensated by a higher reliance on psychosocial interventions. This is consistent with a recent survey showing that African Americans in primary care suffering from depression were twice as likely as white patients to prefer psychotherapeutic intervention than antidepressant medication (27). Unexpectedly, Hispanic ethnicity was not associated with lower use of psychotropics or antidepressants in this study. A national survey in the general population (i.e., not focused on HIV-positive patients) reported a greater unmet need of mental health services among Hispanics than among whites (14). Another study, however, found that, in the early and mid-1990s, the pharmacological treatment of depression dramatically increased among Hispanic patients but remained low among African Americans (39).

Contrary to our prediction, female sex was not associated with higher use of psychotropics in general or antidepressants in particular. Previous studies have reported that about two-thirds of psychotropic medication visits are for women, both in primary care and psychiatric settings. The present findings may reflect characteristics of HIV-positive men as being more likely to search for pharmacological treatment of mood and anxiety disorders than HIV-negative men. A less marked—but still significant—effect of age was also detected, with middle-age patients being more likely than older patients to have been treated with psychotropics (Table 3). This effect, however, was present only in the entire sample and disappeared when patients with a mental disorder only were analyzed (Table 4).

As expected, Medicare patients have a higher rate of psychotropic use (Table 3). Patients with Medicaid or without insurance, however, had rates of use that were comparable to those of patients with private insurance, thus indicating that for patients already in medical care these differences in insurance status do not significantly affect the use of psychotropics.

In addition to the study limitations mentioned, a few others also need to be pointed out. The severity of psychopathology was not assessed, and this limits inferences on the adequacy of treatment. For instance, undertreatment of severe depression would be of greater concern than undertreatment of milder forms. Also, chart diagnoses made by the patients’ providers were not obtained. Information about the prescribers of psychotropic medications (e.g., family doctor, psychiatrist), duration of treatment, and dose were not available.

In conclusion, this study of a national probability sample of HIV-positive patients found that psychotropic medications, and antidepressants and anxiolytics in particular, were commonly used by HIV-positive patients receiving medical care in the United States. However, more than one-half of the patients suffering from major depression were not treated with antidepressants, even though a substantial portion of those not receiving medication reported use of some type of psychosocial intervention. African Americans were less likely to have been treated with psychotropic medications, thus adding to other data indicating differences in health care among ethnic groups.

Received Jan. 18, 2002; revision received Aug. 29, 2002; accepted Sept. 17, 2002. From the Division of Services and Intervention Research, NIMH; RAND, Santa Monica, Calif.; and the Center for AIDS Research, Education and Services, Charles R. Drew University, Los Angeles. Address reprint requests to Dr. Vitiello, NIMH, Room 7147, 6001 Executive Blvd., MSC 9633, Bethesda, MD 20892-9633; [email protected] (e-mail). The HIV Cost and Services Utilization Study was conducted under cooperative agreement HS-08578 between RAND and the Agency of Healthcare Research and Quality (formerly Agency of Health Care Policy and Research) (principal investigator: Dr. M.F. Shapiro, co-principal investigator: Dr. S.A. Bozzette), with additional funding provided by the Health Services Resources Administration and NIH, Bethesda, Md. (NIMH, the National Institute of Drug Abuse, and the Office of Research on Minority Health through the National Institute for Dental Research). Dr. Bing received support from NIMH as a UCLA Faculty Scholar in Mental Health Services Research (MH-00990) and through the California Universitywide AIDS Research Program (IS99-DREW203). The authors thank Joan Tucker and Geoffrey Joyce for their comments and suggestions on the article. The opinions and assertions contained in this report are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Health and Human Services, the Agency of Healthcare Research and Quality, or NIH.