People with severe mental illness, particularly those with a history of comorbid substance use disorders, have a highly elevated risk of HIV, hepatitis B, and hepatitis C infection (1,2,3). Multiple studies have shown an elevated prevalence of HIV infection among persons with severe mental illness—3.1 to 22.9 percent, compared with an estimated prevalence of .32 percent to .42 percent in the U.S. adult population (3,4). The U.S. Centers for Disease Control and Prevention (CDC), citing the number of individuals with undiagnosed infection in the community, recently reemphasized the importance of viral testing for at-risk persons (5).
For people who have severe mental illness, infectious hepatitis may pose an even greater health threat than HIV. Recent findings suggest that the prevalence rates of hepatitis B (23.4 percent) and hepatitis C (19.6 percent) are approximately five and 11 times as high, respectively, among persons with severe mental illness as in the general population and that rates of HIV and hepatitis C co-infection are also high. The risk of co-infection, along with findings of increased risk of morbidity and mortality associated with co-infection (6), supports the necessity of early detection and treatment among clients with severe mental illness, including development of procedures to reduce barriers to care for people with this complex array of disorders.
As a clearly high-risk population, clients with severe mental illness and those with dual diagnoses should receive basic CDC-recommended services (7) for HIV infection or AIDS and for hepatitis, including screening for risk, testing of persons at risk, immunization to prevent hepatitis A and B, risk reduction counseling, and referral and support for medical care (8,9,10,11,12). Unfortunately, these services have not been broadly implemented. Although promising interventions to provide counseling about HIV risk reduction for persons with severe mental illness have been described (13,14), dissemination of even these limited interventions has lagged.
A majority of community mental health providers do not regularly address HIV risk or illness issues with severely mentally ill clients, even in communities where the prevalence of HIV infection and AIDS is known to be high (15,16,17,18,19). None of the studies of community mental health providers indicate that these providers are engaged in education, screening, or testing for hepatitis. Moreover, most persons who have both severe mental illness and hepatitis C do not have regular sources of medical care (20), which compounds their vulnerability to adverse medical outcomes. Most clients who are in community mental health treatment in the United States do not receive basic services for these blood-borne diseases, and effective and feasible models for responding to this problem are lacking.
A number of barriers to implementation of best-practice care have been identified in the literature (21,22,23,24). Mental health providers report too few staff, lack of funds, competing demands, a belief that clients receive these services elsewhere, psychiatric providers' reluctance to offer infectious disease services, medical providers' reluctance to provide treatment to clients who have severe mental illness, and separate funding and administration for medical and psychiatric services. Client barriers include lack of information, inconsistent motivation, lack of insurance or means to pay, lack of stable connection to a medical care provider, and confusion about needed services.
We attempted to develop and test a model of infectious disease prevention and treatment support that could address these barriers and be integrated into mental health service settings. This intervention was designed to provide a best-practice level of care for clients with severe mental illness who are receiving public-sector mental health services and includes screening, testing, immunization for hepatitis A and B, risk reduction, and referral to medical care. To enhance the feasibility of broad dissemination, the intervention was designed to be brief, to involve minimal burden on providers, and to be adopted in multiple settings.
The STIRR (screen, test, immunize, reduce risk, and refer) intervention blends public health and clinical aspects (25) to bring a small group of expert providers to a clinical site, for a limited period, to provide core services to an undertreated, at-risk population. The STIRR intervention involves three basic components: implementation support, on-site services, and treatment referral and support. These components are specified in an implementation manual that is available from the first author.
The STIRR intervention is delivered by a small team of trained specialists, including an implementation support specialist, an infectious disease coordinator (a nurse or a physician assistant), an internist with expertise in diagnosis and treatment of HIV and hepatitis, a psychiatrist with expertise in dual disorders, and an administrative support person. In the two pilot applications reported here, all team members were on the faculty of Dartmouth Medical School. Most contact with clients and staff is made by the nurse or coordinator (about one hour per client) and by the administrative support person. The implementation support expert spends ten to 12 hours at each site, and the internist and the psychiatrist communicate on-site and by telephone for between five and ten hours per site.
Before staff members meet with clients, materials and space are arranged to support the infectious disease clinic, including materials for processing blood samples as well as refrigerated storage for the Twinrix vaccine. In addition, the STIRR team establishes specific referral mechanisms with local medical providers for clients who need further assessment and treatment.
Implementation support. On the basis of Green's model (26) as elaborated by Dieterich and associates (27,28), the STIRR intervention uses proven methods of practice change. These methods include principles from educational psychology and academic detailing and are used to first negotiate service system support and then to educate and motivate mental health treatment teams to accept and support an infectious disease initiative (29,30,31). Implementation support procedures for STIRR include administrative system interventions, provider education, use of concise materials, repetition of essential messages, social influence, performance feedback, and targeting of specific opinion leaders.
Implementation begins with a meeting to educate and motivate the organization's management to support the infectious disease intervention. At this meeting, the epidemiology of infectious diseases among persons with severe mental illness is reviewed, and the STIRR model, including projected requirements of providers and benefits for participants, are presented. The STIRR team next attends scheduled treatment team meetings to introduce the same material and secure cooperation in implementing the STIRR intervention. Before the STIRR services are actually introduced, an implementation expert meets with providers to determine how best to facilitate and adapt the services at the particular site. Focus groups at each setting assess reactions and concerns, particularly concerns about implementation barriers at the client, treatment team, and organizational levels.
Once barriers are understood and entrée is negotiated, brief on-site training is offered to the treatment team to supplement its knowledge about HIV, hepatitis B, and hepatitis C and to familiarize treatment team members with the STIRR team and procedures. The treatment team is provided with informational materials to educate, persuade, and reinforce provider support for the STIRR intervention, including reminders such as posters about hepatitis C and sticky notes to flag clients' charts. Members of the treatment team are not asked to provide STIRR services except to support participation in medical care for clients who have tested positive for infection.
On-site services. The STIRR nurse, who is a member of the treatment team, offers screening, diagnostic testing, and prevention services to clients who are receiving community support services at the site of usual care. Three individual meetings are held with eligible clients for risk assessment, counseling, blood testing, and immunization.
The first individual session provides client education, personalized risk assessment, pretest counseling, blood draw, and first immunization with hepatitis A inactivated and hepatitis B (recombinant) vaccine. Screening for risk is accomplished by giving each client an educational brochure and reviewing a list of personal risk factors. Items listed include high-risk behaviors that are commonly reported by clients with severe mental illness, including needle or drug sharing, receipt of a blood transfusion before 1992, and risky sexual practices.
After assessing personal risk factors, the STIRR nurse offers clients individual pretest counseling and a blood test for HIV, hepatitis B, or hepatitis C. Blood is drawn on-site and is then either spun by centrifuge for storage or refrigerated and then shipped to the laboratory for testing.
After counseling and blood draw, each client is offered immunization with hepatitis A inactivated and hepatitis B (recombinant) vaccine. Alternatively, sites could use monovalent hepatitis A and hepatitis B vaccines. However, this approach would require a series of five rather than three injections. Because the STIRR strategy is to minimize clients' burden and barriers to participation, we used Twinrix, a sterile, bivalent vaccine used for primary immunization for adults. Injections were given at zero, one, and six months. The vaccine provides comparable protection to monovalent hepatitis A and hepatitis B vaccines. Although the optimal series includes three injections, a majority of benefit is attained after completion of the first two injections. Twinrix has proved to be well tolerated and safe and has the advantage of protecting against both hepatitis A and hepatitis B with one vaccine.
After the immunization, the STIRR nurse once again reviews the client's personal risk factors and provides basic information about how to avoid getting or spreading hepatitis and HIV. Major topics include the dangers of intravenous drug use; risks associated with sharing needles, razors, toothbrushes, or nail clippers; the need for sterilization of drug paraphernalia; and reduction of sexual risk. Finally, clients are encouraged to make a list of "what you want to do to avoid getting or spreading the virus" and to discuss that list with their physician or case manager. To reinforce these messages, clients are also given a new toothbrush, a disposable razor, condoms, and several refrigerator magnets to serve as reminders.
A second session is scheduled approximately one month after the first. In this session the STIRR nurse provides clients with test results, posttest and risk reduction counseling, medical referral and linkage (if needed), and the second Twinrix immunization.
At the third and final scheduled session, the nurse again assesses risk level and reinforces risk reduction. The nurse provides the final immunization and, if relevant, monitors progress on medical treatment and linkage as well as primary and secondary prevention plans.
Treatment referral and support. Referral for further medical evaluation and treatment linkage for infectious diseases does require the mental health treatment team to be involved in collaboration with the STIRR team. Although it is important to educate the mental health treatment team in basic aspects of managing blood-borne diseases, the mental health team is likely to play only a supportive or facilitative role in medical evaluation and possible treatment for persons who test positive for at least one infection.
Because mental health providers may have minimal contact with other medical specialists, the STIRR team establishes referral resources with local infectious disease providers and gastroenterologists. The team also provides consultation to clients and to mental health providers about the complex medical decisions that infected clients must make as well as consultation about linkage to psychiatric support and monitoring during medical treatments.
Implementation experience at two sites
Beginning in January 2002 we introduced a "health promotion intervention," with a focus on blood-borne infections, to two state-funded community mental health centers. The costs of the intervention, which was part of an intervention development study of the National Institute of Mental Health, were borne by the research team. The leadership groups at both sites were highly supportive of the STIRR team's providing services but also were very wary about using time of already overburdened personnel. Nurses are particularly valuable to the treatment teams and are very hard to recruit and retain in the current funding environment for mental health services.
The terms for introducing the infectious diseases-related services included absolute minimization of new demands on treatment team nurses. To achieve this objective, it was agreed that the STIRR team nurse, who was paid from external funds and had no other role as part of the regular mental health treatment team, would do all pre- and posttest counseling, all blood draws, and all STIRR immunizations. Providers and administrators readily agreed on the value of the STIRR intervention, although individual case managers varied in the degree to which they seemed to prioritize infectious diseases services over competing issues—for example, housing instability. All procedures were approved by the institutional review boards of Dartmouth and the state of New Hampshire as well as by human subjects committees at the clinical sites at which the study was conducted. Clients who participated in the risk survey or in focus groups provided written informed consent, and all clients provided verbal consent before receiving STIRR services.
The site of the first pilot study was a community mental health center located in a rural area known for high rates of poverty, substance abuse, and HIV infection. The center serves primarily clients with low incomes and severe mental illness. At this site, we targeted clients who were receiving community support services (integrated case management, psychiatric care, and rehabilitative services) from two multidisciplinary treatment teams. The total caseload for these teams was 184 clients. A total of 107 clients (58 percent) were women. The average age of the clients was 43±10.97 years, and the primary diagnoses included schizophrenia (63 clients, or 34 percent), major depression (63 clients, or 34 percent), and bipolar disorder (24 clients, or 13 percent). The average number of years since receiving mental health services was 11±7.26 (range, less than one year to 37 years).
All 184 eligible clients were approached by their providers to participate in the screening and intervention. Eleven of these clients had already been immunized or were ineligible for the study for other reasons, leaving 173 eligible for immunization. The STIRR team was available on-site for two days a week and typically met with clients as they came into the center for scheduled doctors' visits or other usual mental health services. Over a period of approximately 20 weeks, 136 clients completed blood draws for infectious disease testing, and 137 (79 percent) were immunized for hepatitis A and hepatitis B with Twinrix. Thirty-six clients (21 percent) refused or deferred immunization.
To learn about the feasibility and generalizability of STIRR dissemination, we chose as our second pilot site the largest mental health center in New Hampshire. This site serves a small city population, and clients share many of the characteristics of inner-city residents with severe mental illness—for example, poverty, poor social support, history of homelessness, and high rates of substance use disorders. To further test the limits of the intervention, we chose the highest-risk and most difficult to treat clients at the center: those being served by the dual disorders treatment team (severe mental illness plus an active substance use disorder) as opposed to all severely mentally ill persons receiving community support services. Clients with dual diagnoses frequently drop out of care, have high rates of incarceration, and often do not appear for scheduled appointments.
We also tested the possibility of abbreviating the intervention to reduce costs and personnel requirements. The STIRR team attempted to recruit all clients and complete the first session, including blood draw and first immunization, within eight weeks. This approach required clients to visit the mental health center at times when they were not receiving another routine mental health service, such as a medication check. In addition, we asked participants at this site to respond to questionnaires about their knowledge, motivation to reduce risk, and actual risk behavior in terms of HIV and hepatitis. The total caseload for these teams was 110 clients. A total of 32 clients (29 percent) were women. The average age of the clients was 41±8.37 years. Major diagnoses included schizophrenia (78 clients, or 71 percent) and bipolar disorder (15 clients, or 14 percent). The number of years since receiving mental health services was 11±8.68 (range, less than one year to 27 years).
Not surprisingly, participation rates at the second site were somewhat lower. Of the 110 clients of the dual diagnosis team, 99 were referred to the STIRR team. Of these, 67 (68 percent) met with the STIRR nurse for screening and pretest counseling, 65 agreed to blood testing, 64 completed the first two immunizations, and 54 completed the third. Fifty clients were assessed at baseline and at their third immunization (six months) with a modified version of the AIDS Risk Inventory, a structured interview for assessing knowledge, attitudes, and risk behaviors associated with acquiring and transmitting blood-borne infections. This scale is reliable and valid among respondents with severe mental illness (32,33,34). Results, using t tests for paired samples, showed increased knowledge (t=3.85, df=39, p≤.01) and increased motivation to reduce risk (t=5.21, df=41, p≤.01) but no decline in self-reported risk behavior.
Delivery of STIRR services required one nurse full-time equivalent per 200 clients. Direct costs, including start-up costs, averaged less than $220 per participating client, plus the cost of Twinrix vaccine (approximately $37 per dose if purchased through a state agency). This figure includes all time spent by the STIRR team at the two pilot sites; travel time and costs to the sites; community mental health center administrator and treatment staff time (approximately 16 percent of total intervention costs, or $35 per patient); costs of materials, including those given to consumers; equipment such as needles and refrigerators for vaccine storage; and testing costs. Although per-client costs clearly would be lower at larger sites, it is also possible that the costs would be higher in some provider organizations. The two pilot studies provided evidence of broad staff acceptance of STIRR, high rates of client referral and participation in testing and immunization, appropriate referral for medical follow-up, and improved knowledge about blood-borne infections and motivation for prevention.
Our experiences at two community mental health centers support the feasibility and efficacy of the STIRR intervention for clients with severe mental illness and those with dual diagnoses. The STIRR approach provides best-practice interventions for this high-risk population in a relatively efficient manner that should have benefits for clients' health and quality of life in addition to public health benefits. Both clients and providers are motivated to improve basic prevention, detection, and care for HIV and viral hepatitis. However, to optimize participation, information must be presented in a simple, clear, and concise manner, and services should be delivered at the usual site of mental health care. Mental health providers themselves were generally unwilling to master and deliver the infectious disease-related services, even though recognizing their value. Rather, when appropriate groundwork was laid through implementation support procedures, the clinicians preferred to support these services when provided on-site by an outside team of experts. We hypothesize that this approach will prove feasible for broad dissemination in a variety of treatment settings that serve clients with severe mental illness and dual diagnoses, but issues of funding and implementation in more urban and diverse populations must be further examined.
Finally, it should be emphasized that the STIRR intervention provides only basic, minimum recommended services. The most efficacious approaches to risk reduction for persons with severe mental illness should also include integrated dual diagnosis treatment (35,36,37) and more intensive interventions especially adapted for and tested in this population. These interventions use cognitive-behavioral therapy approaches to reduce sexual risk of HIV (13,14,38,39).
This research was supported by grant R21-MH-62270 from the National Institute of Mental Health to Dr. Rosenberg. The authors acknowledge the invaluable participation of Robin Boynton, Cynthia Hewitt, Ravindra Luckoor, M.D., Amy Noack, M.D., Nancy Walsh-Robard, A.R.N.P., Mark Iber, P.A., and Deborah Gottel, R.N.
The authors are affiliated with the New Hampshire-Dartmouth Psychiatric Research Center and Dartmouth Medical School, 2 Whipple Place, Suite 202, Lebanon, New Hampshire 03766 (e-mail, firstname.lastname@example.org).