Practices Related to HIV Risk Assessment in General Hospital Psychiatric Units in New York State

Psychiatric units must now treat patients with HIV disease. A review of the 11 U.S. HIV seroprevalence studies in psychiatric populations published in peer-reviewed journals through 1996 reported rates ranging from 4 percent on a long-stay unit to 22.9 percent on a unit with substance-abusing psychiatric inpatients (1). Overall, 223 of the 2,873 psychiatric patients tested (7.8 percent) were HIV positive.

Studies of people with severe mental illness in the U.S. show high levels of HIV risk behavior. In a review of 28 peer-reviewed studies, Carey and colleagues (2) found that 4 to 35 percent of patients with severe mental illness reported a history of injection drug use, 5 to 8 percent in the past year. Half to three-quarters of these patients reported sexual activity in the past year, 20 to 40 percent with multiple partners.

Other risky sexual practices included sex trading (7 to 28 percent) and having sex with high-risk partners. The study found that 4 to 10 percent of the patients reported recent partners who were injection drug users, and an average of 13 percent of men reported having male sex partners. Twenty to 40 percent said they used alcohol or drugs before sex, and a quarter to a half reported that they never used condoms. Another study found that a significant minority of psychiatric patients who were sexually active met partners at clinics (3).

No data have been systematically collected on the extent to which psychiatric patients are infected with HIV, but reports indicate that HIV infection of many patients goes undetected during inpatient stays (4). Thus questions can be raised about the common practice of providing HIV assessment, counseling, and other services only when routine intake procedures signal an elevated risk. One study in New York City found that while only about 1 percent of patients reported using injection drugs in the past six months, 20 percent reported use when they were asked to report any injection drug use since 1978, the start of the HIV epidemic in the U.S. (5). In addition, some men who have sex with men do not consider themselves homosexual or bisexual if they have sex to gain money, drugs, or protection. Thus questions about sexual orientation may produce false negatives for risky practices.

An argument can be made that systematic assessment is required to detect risk, incorporating individualized, detailed, and longitudinal inquiries. To learn how common various practices related to HIV transmission might be in general hospital psychiatric units in New York State, we included questions on this topic in a semistructured interview conducted during site visits across the state.

Methods

The data reported here are from the Rutgers hospital and community study, described in detail elsewhere (6). In 1992 and 1993, members of our interdisciplinary research team visited a nonrandom sample of 53 units drawn from 47 of the 106 general hospital psychiatric units across New York State, approximately half in the New York City area and half elsewhere. Two to six key staff members were interviewed individually at each site. Data are reported on 136 interviews.

HIV-related practices were assessed with four interview items. What percentage of patients receive information on HIV and AIDS? What percentage of patients are assessed for high-risk behavior? What percentage of patients are counseled about their own HIV risk? For what percentage of patients have you sought permission to conduct HIV testing? Responses used a 5-point scale: 1, almost none; 2, a few; 3, some; 4, many; and 5, almost all.

For each question, unit staff responses were averaged to yield a single number representing a unit's practices. Because averages could fall between integer points, scores were recoded with an arbitrary decision rule to allow interpretation using terms from the scale: 1 to 1.5, almost none; greater than 1.5 to 2.5, a few; greater than 2.5 to 3.5, some; greater than 3.5 to 4.5, many; and greater that 4.5, almost all.

Agreement among respondents was calculated by determining the number of responses to a given question that were in agreement (plus or minus 1 point) as a percentage of the total possible agreement on the unit (determined by the number of staff interviewed on the unit).

Results

Unit-level agreement in responses was satisfactory, ranging from 64 percent to 85 percent. Statewide, the hospitals visited varied in their practices related to HIV and AIDS. Of the activities described, assessing high-risk behavior was the only one approaching standard-practice status, with 25 units (47 percent) reporting that they conducted such assessments for many or almost all patients. Three units (6 percent) responded that almost all patients received information about HIV and AIDS, and 13 (25 percent) responded that they provided information to many patients. However, 18 units (34 percent) reported that only a few patients received such information, and five (9 percent) reported that almost none did.

Twenty units (38 percent) reported that they counseled only a few patients about their own risk behavior; eight (15 percent) reported they provided such counseling to almost none of their patients. Urging patients to get tested was found to be the least common practice—nine units (17 percent) reported this practice with almost none of their patients, and 28 (53 percent) reported it with only a few patients.

Discussion and conclusions

In this sample of roughly half the general hospital psychiatric units in New York State, most units reported that they urged testing for only a few, if any, patients and that they failed to apply other HIV and AIDS risk assessment practices routinely. Practices were more commonly reported by units in areas where HIV is more prevalent in the general population. More labor-intensive and intrusive practices, such as counseling about individual risk or urging testing, were less common.

In some areas interviewers encountered staff who responded to questions about HIV and AIDs practices with the assurance that "that's not a problem for us here.” This response is troubling, given the evidence that only detailed, individualized interviews allow the clinician to assess risk. The danger is that patient stereotypes will serve as rules of thumb and that risk, and HIV infection, will remain undetected. Investigators have reported high rates of underidentification of HIV among newborns in predominantly white, nonurban areas of New York State (7), where clinicians may underestimate risk.

Although these data represent an advance over information from single units or programs, they have limitations. They are based on staff reports, not behavioral observations, and they may not represent current practices because they were collected at an earlier phase in the epidemic. Further, the items failed to cover many crucial questions about social welfare services, substance abuse treatment, and medical consultation.

The American Psychiatric Association's ad hoc committee on AIDS policy has issued guidelines (8) for inpatient psychiatric units that include a call for "appropriate education and supportive services for patients, families, and staff.” Just what is “appropriate”is far from clear, however. If each of the practices identified in our survey ought to be standard, a mere handful of units provide optimal care. Yet some authorities tie the need for testing to the presence of various clinical indications of risk (9), and the idea of mandatory HIV testing and counseling for all psychiatric inpatients is subject to debate (10).

Acknowledgments

The authors gratefully acknowledge the contributions of Michelle Daniel, Psy.M., and study team members Carol Boyer, Ph.D., Sara Kellermann, M.D., and David Mechanic, Ph.D. The Rutgers hospital and community study was made possible by a grant from the Robert Wood Johnson Foundation and support from the Rutgers NIMH Center for Research on the Organization and Financing of Care for the Severely Mentally Ill.

Dr. Walkup, Dr. Hansell, and Dr. Olfson are affiliated with the Institute for Health, Health Care Policy, and Aging Research of Rutgers University, 30 College Avenue, New Brunswick, New Jersey 09803 (e-mail, [email protected]). Dr. Satriano and Dr. Olfson are with the New York State Psychiatric Institute and the College of Physicians and Surgeons at Columbia University in New York City.