Residential crisis services as an alternative to inpatient care

ABSTRACT: The authors report the results from a study of a residential crisis program house in Missouri. Like other programs of its kind, the program is designed to provide a home-like atmosphere for consumers experiencing acute episodes. The study contributes to a quantitative comparative evaluation of the effectiveness of such programs; it suggests that a residential crisis program is a less expensive alternative to hospitalization and appears to serve the short-term stabilization needs of its clientele. Client self-reports confirm return to preadmission functioning and high levels of satisfaction.

THE SEARCH FOR VIABLE alternatives to inpatient care dots the history of mental health treatment in America. Today, as in the past, community-based care of persons with severe mental illnesses remains shrouded in controversy. A key debate focuses on the role, if any, for state psychiatric hospitals and specialty inpatient treatment units (Carling, Miller, Daniels, & Randolph, 1987; Geller, 1991; Kiesler, 1982; Lamb, 1992; Mechanic & Rochefort, 1992; Sullivan, 1992). The focus on health care reform, the advent of managed-care models, and the privatization of mental health services represent a renewed focus on the costs of services. Clearly, institutional care is an expensive form of treatment. For example, Mechanic and Rochefort (1992) note that the cost of public hospitalization in New York exceeds $90,000 per patient per year. To assess the costs of treatment, longitudinal studies must that account for the direct and indirect costs of treatment.

Although the prohibitive costs of inpatient care are a focal point for many, others attend to the equally vexing clinical concerns and associated ethical and value dilemmas. For example, some argue that the many former inpatients among the homeless represent a direct indictment of the mental health system and perhaps proof that community-based care is ineffective (Johnson, 1990). On the other hand, many current and former consumers of services assert their desire and right to remain in the community (Okin & Pearsall, 1993). Additionally, innovations in community-based care and technologies offer hope that community integration of persons with severe and persistent mental illnesses will one day become a reality.

Regardless of these debates, clearly the majority of mental health care is now being provided in community settings, and many state psychiatric hospitals have been closed and/or downsized. This shift has increased the burden on community services and families to provide acute care, even long-term services, for persons with extensive needs. Intensive case management, crisis-response teams, and a panoply of psychosocial services have been designed to attend to consumer difficulties and needs in the least intrusive manner possible in order to maintain and support them in the community. However, given the nature of severe mental illness, some consumers still need intensive around-theclock care in a supportive environment.

During the past two decades, reports on residential-based hospital diversion programs have been presented (Bond et al., 1989; Dinitz, 1979; Kresky-Wolff, Matthews, Kalibat, & Mosher, 1985; Lamb & Lamb, 1984; Polak & Kirby, 1976; Stroul, 1988; Weisman, 1985). Stroul (1988) notes that such programs, referred to as respite programs, stabilization units, or crisis hostels, share several fundamental characteristics. Specifically, residential-crisis programs provide housing during a crisis, prevent hospitalization, provide acute treatment and support, and offer short-term services. Furthermore, although the primary focus of these programs is on crisis stabilization, attention is also directed toward helping consumers develop and/or maintain community support systems and activities.

An additional purpose of residential crisis programs is to continue the stabilization process and/or ease the community transition for those who have recently been hospitalized (Weisman, 1985). Wells (1992) noted that consumers recently discharged after a psychiatric hospitalization often experience anxiety attacks, insomnia, depression, and other threats to their community tenure. Kent and Yellowlees (1994), in a study that examined consumers who experience frequent hospitalizations, note that social factors are frequently implicated in recidivism. By providing shortterm intensive support, particularly when placed within an array of community-based services, residential stabilization programs may play a key role in reducing the need for future hospitalizations. This service alone can divert a consumer from a mental patient career (KreskyWolff et al., 1985; Weisman, 1985).

Residential crisis programs also embody a set of values and assumptions that mirror philosophical tenets in the original thinking supporting the community mental health center movement. Noteworthy among these assumptions is the hope that with prompt help in acute episodes provided in the community, the consumer can rapidly return to normal activities. Here, avoiding hospitalization, not simply a matter of financial prudence, also benefits the service consumer. As Weisman (1985) notes,

The fact that they are not hospitalized is important to their [consumers'] sense of identity-they can leave the program upon resolution of a psychotic episode and do not experience the internal and social stigmatization attendant upon a stay in a psychiatric hospital (p. 1305).

In an effort to move from an institutional model, these programs often strive to create a home-like atmosphere with the attendant responsibilities and expectations of consumers serving as a member of the community (KreskyWolff et al., 1985; Stroul, 1988; Weisman, 1985). Because of the accessibility of such programs, involvement with family and friends, with consumer permission, can be sustained.

Previous studies have suggested that residential crisis services are efficacious. Bond and associates (1989), in a comparison of two residential crisis programs, found that two-thirds of consumers served in both programs had avoided hospitalization at the four-month follow-up. Kresky-Wolff and colleagues (1985) note that only a small percentage (9%) of consumers admitted to a residential program were subsequently referred directly to the hospital. Studies reviewed here also note the cost benefits of these diversion programs in comparison with an inpatient stay in a state psychiatric facility. In summary, it appears that residential crisis programs are an effective alternative to inpatient care and that they may be underutilized in most community-based service systems.

This article reports results from a preliminary study of a residential-crisis-program house operating in Missouri. This program is designed specifically to provide a home-like atmosphere to address acute episodes of consumers. The services offered by this program are critical in that the state psychiatric hospital that had previously served the region is now closed.

Robberson House

Robberson House provides stabilization services for persons experiencing early signs or acute symptoms of mental illness or who are in the process of recovery from a severe illness episode. The facility is designed to provide a home-like atmosphere and includes a kitchen, living room, and activity area. The facility can serve up to 10 adults in 5 bedrooms. At least two staff are present 24 hours a day, with a psychiatrist and nurse assigned approximately 10 hours per week to provide medication management. Care plans are individualized to each consumer with the primary focus on stabilization. Meals are provided on site, and residents have a choice in daily activities. Meal preparation and housekeeping are provided by current mental health consumers participating in a work unit and/or in supported employment slots. Robberson House is adjacent to the host community-support program but is nested in a residential neighborhood. The program also serves as an anchor for the crisis-assistance teams and case managers working within the mental health center, but referrals can come from throughout the community or be self-initiated. At the time of opening the average daily cost of the program was $175, at least one-half the average cost of a day's hospitalization in Missouri.

In discussion with key center staff, four main objectives were established:

To stabilize crisis situations efficiently and in a satisfactory manner. This objective includes the reduction of presenting symptoms.

To maintain or reduce the frequency of hospitalizations at preadmissions levels.

To return consumers to preadmission levels of functioning with regard to housing and vocational status.

To attain a high level of consumer satisfaction with services.

This report summarizes data gathered from June 1, 1993, to May 31, 1994. The first three months of the project were devoted to the identification of program goals and modes of operation. The second quarter of 1993 was devoted to evaluation design and the pretesting of instruments. During those months, data-collection protocols were refined, evaluation staff were trained, and the obtrusiveness of the evaluation on the day-to-day operation of Robberson House was monitored and minimized. All consumers admitted to Robberson House on or after June 1, 1993, made up the sample for the formal quantitative evolution, which was conducted for one year.

Data Collection

The initial data-collection design involved selective use of existing Robberson House forms. Efforts to refine this process revealed that the evaluation purpose (as opposed to therapeutic purpose) was ill served by (1) reliance on open-ended and highly interpretative items, (2) the many forms sifted through to yield the few items needed for evaluation, and (3) the need to rely on student researchers to obtain a standard set of information when charts were complex and active. These problems inspired the collaborative development of a short, standard, admission form that relied on closed-ended responses. This form was completed by Robberson House staff on the 191 admissions from June 1, 1993, to May 31, 1994. Information drawn from a follow-up telephone interview based on a standard protocol completed the database.

The follow-up form was to be administered at least one month after discharge. A total of 96 follow-up interviews were completed, 77 of which were on the sample collected after standardization of the admission forms in June 1993. This 40% response rate compares favorably with the rates obtained by others with low-intrusive, telephonebased protocols on unstable populations (Hartman, Sullivan, & Wolk, 1991; Wallace, McNeil, Gilfillian, Maclean, & Fanella, 1987).

Gaps in information regarding how to locate the individual were the single greatest obstacle to follow-up response. Approximately 30% of the population had no telephone number or other lead included on the information provided at intake. In addition, the residential instability often experienced by this population makes follow-up difficult in the best of situations. Overall, approximately 54% of follow-up interviews were completed within three months of discharge and 87% within six months.

Findings

The majority of admissions for the period under study were women (64%) in their early thirties who lived in typical community housing (57%). Consistent with our knowledge of the vocational activity of mental health consumers, 66% of respondents reported no vocational activity, with approximately 22% engaging in fullor part-time employment. Approximately 38% of consumers reported that they lived alone, 20% with a significant other, 9% with a significant other and children, 4% with children only, 7% with other relatives, and 11% with friends.

Many clients (n = 28) did not provide information on the question of prior hospitalizations. Of the 163 consumers who did respond, more than 45% reported no hospitalizations in the two years prior to admission, 23% reported one prior admission, and 31% experienced two or more prior admissions. For those consumers who had been hospitalized (n = 80), the total number of days spent in inpatient care ranged from 0 to 578, with a median of 14. More than 56% of respondents reported no stabilization-unit admissions in the past two years, 22% reported one admission, and 22% reported two or more. The median number of stabilization days was eight, six days less than reported hospital admission stays.

Although a variety of diagnoses were assigned at admission, the global categories of schizophrenia (22%), major depression (20%), bipolar disorder (17%), and adjustment-reaction disorder (13%) were the predominant categories. Substance abuse was a primary or secondary diagnosis in 21% of cases.

Most clients had multiple conditions at admission. Suicidal ideation (33%) was the most common primary condition, followed by depression (15%), decompensation (14%), anxiety (13%), medication changes (11%), and posthospital stabilization (8%). Sixty-three percent of all clients reported having planned suicide at some point. Almost 62% of the total and 72% of those responding reported actually trying to complete a suicide. A significant proportion of consumers (nearly 50%) also reported drug- and/or alcohol-abuse problems, with 53% of this group having received treatment for this problem. This incidence of alcohol and/or drug use is significantly greater than the number of consumers in the study who were officially diagnosed with a substance-abuse problem, perhaps reflecting anew the lack of attention to these concerns in the mental health community.

Other stressors were also noted upon admission. Whereas 38% of respondents reported no problems with family in the past six months, 26% reported some difficulties and 36% reported major problems. Other stressors included problems with friends (27%), financial difficulties (43%), and medication issues (33%).

Because the form for follow-up data did not change after March 1, 1993, the 96 completed interviews collected since that time are reported in the following paragraphs. Seventy-seven of these remain in the data set linked to the final intake form.

In assessing changes in major life domains, 30 clients reported a change in their living situation, with 18 stating a change in their vocational activity. However, a consistent trend to or from positive outcomes was not found, thus suggesting some measure of stability. These resuits are important because a major objective of Robberson House was to return clients to preadmission levels of functioning in the short term.

In the area of hospitalization usage, only 20 (20.8%) respondents reported a hospital admission since their discharge, and 14 of these reported one admission. The median number of days spent in the hospital was eight. Significantly, 61% of respondents believed they would have gone to the hospital if Robberson House had not been available. The frequency and perhaps the duration of hospitalizations appears to have been reduced as a result of the stabilization program. Seven consumers reported a readmission to Robberson House, with a median stay of five days. The few clients reported on here and the brief follow-up period suggest caution in interpreting results.

Satisfaction with the services offered at Robberson House was high. Approximately 55% were highly satisfied with services, 24% reported being satisfied, and only 7% were either dissatisfied or highly dissatisfied. Consumers' overall satisfaction was consistent with a more specific evaluation of services. Nearly 85% felt that they were allowed to stay in Robberson House as long as they needed to stay, and 87% felt that this stay was helpful to them. Furthermore, more than 83% would feel comfortable returning to Robberson House. Additionally, 55% of respondents indicated that they were satisfied or very satisfied with their life currently, in contrast with the 20% who reported that they were unsatisfied (only 6% in the latter category).

Three final measures of evaluation complete the self-report utility assessment. Not only were clients pleased with the services, 97% believed they were ready to be discharged when they left, and 96% felt that they were better able to cope with their problems when they left. In terms of short-term stabilization, these self-reports are encouraging.

Associations between intake and outcome data (n = 77) were examined. The outcome measures displaying sufficient variation to allow meaningful analysis included hospitalizations since discharge, satisfaction with services, and life satisfaction. Predictor variables included gender, prior hospitalizations, family problems, substance abuse, planned suicide, and history of assaultive behavior. Not enough cases were available to allow statistical tests of primary condition at admissions with outcomes, but no discernible patterns were evident.

The 40% who responded were not statistically distinguishable from nonrespondents on most variables of interest. This group was less likely to have been hospitalized in the prior two years (47% vs. 55%) and more likely to have medication stress (43% vs. 33%).

Gender, medication problems at admission, financial problems at admission, history of substance abuse, planning suicide, and history of assaultive behavior were not associated with hospitalization after discharge from the program. As expected, those with at least one prior hospitalization were almost twice as likely to report hospitalization during the follow-up period (30% versus 17%). The small sample prevents this effect from reaching statistical significance as measured by chi-square analysis (p = .2). Nonetheless, this finding is consistent with research indicating that hospitalizations can be predicted by the number of previous hospitalizations, while present difficulties are held constant. More data may show a similar pattern for stabilization-unit recidivists. Family problems at admission have the same level of association. Those with such problems are more than twice as likely to report a postdischarge hospitalization (27% versus 13%). Again the effect misses statistical significance, but the trend is consistent with Kent and Yellowlees's (1994) research on the impact of social factors on rehospitalization rates.

Those with prior hospitalizations are marginally (p = .14) less likely to be very satisfied with their experience at Robberson House. The rate is still high, however (50% vs. 68% for those with no prior hospitalizations). Those who had a planned suicide and those who had a history of assaultive behavior were marginally less likely to be very satisfied (p = .12 and p = .09, respectively). Planned suicide is the only variable associated with lower rates of overall life satisfaction. The association is in the expected direction but is not significant (p = . ).

To summarize, outcome measures were quite positive overall and did not differ dramatically for subgroups defined by these predictors. Effects were often in the expected direction, however, with clients characterized by histories of hospitalization, suicidal and assaultive tendencies, and, to a lesser extent, problems with families having less positive outcomes.

Conclusions and Implications

This evaluation suggests that Robberson House is performing the work for which it was designed. It is a less expensive alternative to hospitalization that appears to serve the shortterm stabilization needs of its clientele very well. The short follow-up period prevents us from making firm judgments regarding hospitalization rates postdischarge although the initial data are encouraging. Client self-reports confirm return to preadmission functioning and high levels of satisfaction.

In an age of health care reform and managed care systems, institutional care will continue to be increasingly scrutinized. Consumer preference and ideological movements within the mental health community also support the search for viable alternatives to inpatient care. Nonetheless, acute crisis situations and the disorganization that sometimes accompanies major neurobiological disorders demand that quality care be available to consumers on a 24-hour basis. Not surprisingly, suicidal ideation was the most common presenting condition at admission to Robberson House. Treating the consumer who presents with suicidal ideation is an important clinical concern that carries a significant financial burden for providers (Rissmiller, Steer, Ranieri, Rissmiller, & Hogate, 1994). Rissmiller and associates (1994) argue that a higher threshold can be employed before inpatient care is initiated. They note:

Hospitalization must evolve from a first-line treatment to a transient treatment necessary only when the risk of suicide is severe. When hospitalization is needed, lengths of stay can be shortened by reorganizing provider networks to allow patients to transfer quickly between inpatient care and other less costly sectors (p. 786).

Residential crisis programs are one possible response to personal and family crisis as well as an intermediate step to ease the transition from inpatient care. Interestingly, reports of the potential of such programs stretch back at least 25 years and could arguably be recognized as the original response to these conditions. Despite this fact, the literature on these alternatives is relatively sparse, which suggests that the use of such programs is low. One certainly cannot draw firm conclusions from the preliminary report offered here. However, the consistency of results across programs, even across time, suggests the viability of residential crisis programs and argues for increased adoption of this or conceptually similar models of care.

REFERENCES

Bond, G., Witheridge, T., Wasmer, D., Dincin, J., McRae, S., Mayes, J., & Ward, R. (1989). A comparison of two crisis housing alternatives to psychiatric hospitalization. Hospital and Community Psychiatry, 40, 177-183.

Carling, P., Miller, S., Daniels, L., & Randolph, F. (1987). A state with no state hospital: The Vermont feasibility study. Hospital and Community Psychiatry, 38, 617-623.

Dinitz, S. (1979). Home care treatment as a substitute for hospitalization: The Louisville experiment. New Directions for Mental Health Services, 1, 1-13.

Geller, J. (1991). "Anyplace but the state hospital": Examining assumptions about the benefits of admission diversion. Hospital and Community Psychiatry, 42, 145-152.

Hartmann, D., Sullivan, W., & Wolk, J. (1991). A state-wide assessment: Marital stability and client outcomes. Drug and Alcohol Dependence, 29, 27-38.

Johnson, A. (1990). Out of bedlam. New York: Basic Books.

Kent, S., & Yellowlees, P (1994). Psychiatric and social reasons for frequent rehospitalization. Hospital and Community Psychiatry, 45, 347-350.

Kiesler, C. (1982). Mental hospitals and alternative care: Noninstitutionalization as potential public policy for mental patients. American Psychologist, 37, 349-360.

Kresky-Wolff, M., Matthews, S., Kalibat, F, & Mosher, L. (1985). Crossing place: A residential model for crisis intervention. Hospital and Community Psychiatry, 35, 72-74.

Lamb, R. (1992). Is it time for a moratorium on deinstitutionalization? Hospital and Community Psychiatry,

43, 669

Lamb, R. & Lamb D. (1984). A non hospital alternative to acute hospitalization. Hospital and Community Psychiatry, 35, 728-730.

Mechanic, D. & Rochefort, D. (1992, Spring). A policy of inclusion for the mentally ill. Health Affairs, 11, 128-150.

Okin, R., & Pearsall, D. (1993). Patients' perceptions of their quality of life 11 years after discharge from a state hospital. Hospital and Community Psychiatry, 44, 236-240.

Polak, P., & Kirby, M. (1976). A model to replace psychiatric hospitals. Journal of Nervous and Mental Disease, 163, 13-22.

Rissmiller, D., Steer, R., Ranieri, W., Rissmiller, F., & Hogate, P. (1994). Factors complicating cost containment in the treatment of suicidal patients. Hospital and Community Psychiatry, 45, 782-788.

Stroul, B. (1988). Residential crisis service: A review. Hospital and Community Psychiatry. 39, 1095-1096.

Sullivan, W. P. (1992). Reclaiming the community. The strengths perspective and deinstitutionalization. Social Work, 37, 204-209.

Wallace, J., McNeil, D. Gilfillian, D., Maclean, K., & Fanella, F. (1987). Six-month treatment outcomes in socially stable alcoholics: Abstinence rates. Journal of Sustance Abuse Treatment, 5, 247-252.

Weisman, G. (1985). Crisis-oriented residential treatment as an alternative to hospitalization. Hospital and Community Psychiatry, 36, 1302-1305.

Wells, D. (1992). Management of early postdischarge adjustment reactions following psychiatric hospitalization. Hospital and Community Psychiatry, 43, 1000-1004.

David J. Hartmann is Associate Professor, Department of Sociology, Western Michigan University, Kalamazoo, Michigan. W. Patrick Sullivan is Director of the Indiana Division of Mental Health and Associate Professor, School of Social Work, Indiana University, Indianapolis, Indiana.

Copyright Family Service America Oct 1996
Provided by ProQuest Information and Learning Company. All rights Reserved