Evaluating the effectiveness of an integrated community continuum of care program for individuals with serious mental illness.

OBJECTIVE: To evaluate effectiveness of an "Continuum of Care Program" (CCCP) for persons with serious mental health conditions in reducing inpatient use, and building a continuum of integrated care that enhanced employment and residential stability. The program combined components of Assertive Community Treatment with a comprehensive wrap-around program. METHODS: A cohort of 1154 individuals admitted to four outpatient CCCPs between December 2003 and May 31 2004 was identified and followed for 1 year. Outcome measures included clinical functioning level, drug/alcohol use, employment, residential arrangement and inpatient use. Regression was employed to explain changes in outcomes between baseline and follow-up as a function of services. RESULTS: Statistically significant changes were seen over a 1-year period in all outcomes. Housing, employment and mental health improved, whereas inpatient utilization and level of care need increased. Older individuals receiving higher levels of care at baseline and those with higher case management and medical service utilization reported higher inpatient use. Outcomes also varied by provider suggesting the contribution of workforce differences to outcomes. CONCLUSIONS: Although significant, changes in outcomes were small. Outcome effectiveness was mixed and generally unrelated to services. These findings imply that significant changes in outcomes may require several years to obtain.

Implementing an E-prescribing system in outpatient mental health programs.

This study describes the implementation and evaluation of an electronic prescription ordering system and feedback report in three community-based mental health outpatient agencies and the usefulness of the system in improving psychiatrists' prescribing behavior. Using the e-prescribing system as a data collection tool, feedback on evidence based prescribing practices for patients diagnosed with schizophrenia spectrum disorder or major affective disorder was provided to agency directors and prescribers via a monthly report. The results of the project were that e-prescribing tools can be installed at a reasonable cost with a short start up period. Although the feedback intervention did not show a significant reduction in questionable prescribing patterns, we should continue to investigate how to best use HIT to improve safety, reduce costs, and enhance the quality of healthcare. A better understanding of what prescribers find useful and the reasons why they are prescribing non-evidenced based medications is needed if interventions of this type are to be effective. Given the availability of administrative claims data and electronic prescribing technology, considerable potential exists to provide useful information for monitoring and clinical decision making in public mental health systems.

Effectiveness of a high end users program for persons with psychiatric disorders.

To evaluate the effectiveness of an intensive system of case management for high end users of inpatient care in reducing psychiatric inpatient utilization. A prepost study design with a contemporaneous comparison group was employed to determine the effects of a State designed intervention to reduce inpatient care for adults with a mental health disorder who had high utilization of inpatient psychiatric care between 2004 and 2007. Logit and negative binomial regression models were used to determine the likelihood, frequency and total days of inpatient utilization in the post period as a function of the intervention. Data from administrative reporting forms and Medicaid claims were used to construct inpatient utilization histories and characteristics of 176 patients. Patients in both groups had a significant reduction in mean inpatient days. However, being in the intervention program did not result in lower odds of being re-hospitalized or in fewer episodes during the study period.

Use of HIV and psychotropic medications among persons with serious mental illness and HIV/AIDS.

Persons with serious mental illness (SMI) have higher rates of co-morbid HIV disorders compared to the general population. There are concerns that the SMI/HIV population may be receiving fewer HIV and psychotropic medications due to problems of access and concerns by providers associated with following complex medication regimes. The purpose of this study was to examine any disparity in medication treatment of the SMI/HIV population by comparing medication use and continuity of prescription fills to groups that had HIV or SMI only versus those with SMI/HIV. Study participants were adult Medicaid recipients aged 19-64 with serious mental illness and HIV receiving services in Philadelphia from 2002 through 2003. Differences between the groups in case mix characteristics, medication use rates, and continuity of psychotropic and antiretroviral medication use were compared using Chi-square, t-tests of significance, and logistic regression. Co-morbid individuals were as likely to have filled prescriptions for psychotropic and antiretroviral medications as those with a single disorder and equally persistent in their continuity of antiretroviral medication refills as those with HIV only. However, persons with co-morbid condition had lower continuity of psychotropic medication use compared to those with SMI only. Our findings suggest the need to develop an integrated medical and behavioral healthcare model to improving coordination and treatment for patients with co-occurring disorders. Future research is warranted to investigate the reasons for the discrepancy in continuity of psychotropic adherence for the SMI/HIV population.

Transporting to treatment: Evaluating the effectiveness of a mobile engagement unit.

BACKGROUND: Substance use treatment providers have increasingly developed novel engagement and low-threshold treatment services (such as mobile treatment units) to meet the needs of people with opioid use disorder (OUD). Use of these service models has outpaced the research on their effectiveness. The current study examines the effectiveness of a mobile engagement unit in connecting individuals with OUD to a treatment program. METHODS: This retrospective cohort study included 468 Medicaid-enrolled individuals served through a managed care behavioral health system. Analyses used administrative data from 2018 to 2019 to compare the characteristics and service use of individuals transported to an intake appointment by a mobile engagement unit with individuals who arrived through typical referral routes such as walk-in, other providers, and court order. The authors employed a difference-in-differences analysis to adjust for prior service history. The outcomes of interest were any utilization of substance use treatment services. RESULTS: The groups were virtually identical in age and gender, prior to matching, except for race where there was a lower proportion of Black individuals (17% versus 44%) and lower pre-service utilization of outpatient and methadone services by the mobile group. Following intake, mobile participants used significantly more outpatient substance use treatment services (23 percentage point relative increase) and methadone maintenance (32 percentage point relative increase) than the comparison group. CONCLUSIONS: The results of this study suggest that mobile engagement units designed to identify and serve individuals with OUD in the community hold promise for reaching underserved high-risk populations and reduce barriers to treatment entry and recovery.

Cost of treating seriously mentally ill persons with HIV following highly active retroviral therapy (HAART).

BACKGROUND: Mounting evidence of high HIV prevalence rates among persons with serious mental illness underscores the importance of identifying and treating this population in order to prevent morbidity, mortality and the spread of the disease. Continual monitoring of services and costs is important for public health purposes to insure that persons with serious mental illness receive care for their HIV disorder that is at least comparable to those with HIV only and that the care is considered to be of equal quality. AIM OF STUDY: This current study examines 2003 Medicaid expenditures associated with the treatment of adults with both serious mental illness and HIV, compared to those with HIV and serious mental illness only. The degree to which the occurrences of co-morbid conditions affect overall expenditures is examined, providing the first published co-morbidity expenditure ratios showing the additional cost burden associated with having these dual disorders. Also, changes in the composition of service costs for the co-morbid population are examined before and after the advent of newer antiretroviral and atypical antipsychotic medications. METHODS: Study participants were adult Medicaid recipients age 19-64 with serious mental illness and HIV receiving services from a large urban city program in 2003. The expenditures were derived from Medicaid claims records. Differences between groups were compared using Chi-square and ANOVA tests of significance. To determine the relative cost burden of having a co-morbid versus a single disorder, a co-morbidity expenditure ratio was constructed using the total expenditure per person of those with a co-morbid disorder compared to the total expenditures of those with SMI-only and HIV-only. In order to determine the relative change in inpatient, outpatient and pharmacy service costs, the composition of service costs in 1996 is compared to the service cost composition in 2003 using the share of total costs that each service contributes. RESULTS: In 2003, 788 persons with both SMI and HIV had the highest treatment expenditures at $23,842 per person followed by 2984 persons with HIV-only at $13,183, while the SMI-only group of 19,664 individuals was $11,860 per person. The comparison group had expenditures of $4,793 per person. The co-morbidity expenditure ratio in 2003 for the co-morbid population compared to the SMI-only group was 2.0 and 1.8 for the co-morbid population to the HIV-only population. Extensive redistribution of cost occurred between service categories in the co-morbid group between 1996 and 2003. The share of inpatient cost was reduced from 64% of total costs in 1996 to 30% of total cost in 2003. Conversely, the outpatient cost share increased from 17% of total costs in 1996 to 42% of total costs in 2003 as did the pharmacy share, which rose from 19% of total costs in 1996 to 27% of total costs in 2003. DISCUSSION: Consistent with previous studies, the co-morbid group is a costly population with respect to treatment, despite the fact that inpatient care has decreased. The co-morbidity expenditure analysis indicates little cost saving associated with treating individuals with the co-morbid conditions compared to the cost of treating either conditions separately. This may suggest a lack of coordination or effective care management in the current system warranting further investigation. Also, we find no difference in the percent of the co-morbid population receiving HIV medication compared to the HIV population alone. This suggests that the co-morbid SMI population was being treated similarly to the HIV only group for their HIV disorders. Finally, though all groups had changes between 1996 and 2003 in the proportion of expenditures allocated to each of the service categories, the redistribution of cost between inpatient and outpatient care was the greatest in the co-morbid group. IMPLICATIONS FOR FUTURE RESEARCH: Although the study data suggests that individuals with both HIV and serious mental illness are receiving similar treatment for their HIV disorder as those with HIV alone, a concern that requires further investigation is the finding that HAART treatment is being used by less than 50% of the co-morbid and HIV only study population. Further investigation is required to determine the reason for the relatively low utilization of HAART medications in both HIV groups. Also, the use of a co-morbidity expenditure ratio offers a promising approach for comparing the cost burden associated with multiple disorders.

Effectiveness of a jail-based treatment program for individuals with co-occurring disorders.

OBJECTIVE: This study examines the effectiveness of a county jail program for individuals with co-occurring disorders (COD) in reducing rates of recidivism and increasing rates of engagement in community-based treatment following discharge. METHODS: Over a period of 30 months, between 2002 and 2004, 261 individuals who screened positive for COD voluntarily entered an in-jail treatment program. The program provided integrated treatment for both mental health and substance abuse by therapists who had attended a state sponsored COD core training curriculum. The same program staff provided outpatient services once individuals were discharged. An observational study design was used to examine jail recidivism and community care as a function of intensity of treatment while in jail. All study participants had a minimum one year follow-up. Data was obtained from a baseline comprehensive screening instrument, administrative claims data and county jail records. Logistic regression models were used to determine the likelihood of re-incarceration and community engagement in treatment as a function of the number of treatment sessions provided by the jail program. RESULTS: County jail records indicated that 47.5% were re-incarcerated within 12 months of discharge. During the four and a half year period following the inception of the COD program 67% were re-incarcerated, which was similar to the national three year recidivism rate. Fifty-two percent (52%) attended a community-based treatment program post discharge. The results of the regression analysis showed that a higher number of treatment sessions in the jail COD program was significant in reducing the rate of re-incarceration but was not significant in predicting who would engage in outpatient treatment post discharge. CONCLUSIONS: Although the findings are promising, the evidence suggests that the jail treatment intervention may need to be more intense than the outpatient model used in this project given that the average length of stay in the jail program was 8 weeks due to release to the community. An alternative consideration would be to only provide the program to those inmates who are sentenced for at least 90 days.

Distance matters in choice of mental health program: policy implications for reducing racial disparities in public mental health care.

The purpose of this study is to examine the influence of race, geographic distance and quality on the choice of community mental health programs. The study population was comprised of adult Medicaid recipients who received outpatient treatment for serious mental illness in FY 2001. A discrete choice model was employed to examine the likelihood of choosing one program over another. Quality was measured based on follow-up after hospital discharge and continuity of care in outpatient services. Maps showing the relationship between race and the quality of care were prepared to visually confirm the results of the statistical analysis. African American and Hispanic clients were less likely to travel further for treatment, while no significant difference was found between the Caucasian and other race groups. Caucasian subjects were more likely to choose programs with a higher quality of care compared to Hispanic or African American clients. Higher income clients were, on average, traveling longer and receiving better quality of care after controlling for race. The results suggested that clients living in higher income White neighborhoods are more likely to travel longer distances for mental health treatment. Special attention must be paid to improve the quality of care in lower income minority neighborhoods to insure equity of treatment in publicly funded programs.

Costs of implementing a computerized prescription system in a public mental health agency.

OBJECTIVE: The objective of this case report is to inform decision makers about costs associated with adding a computerized prescription component to an existing information system in specialty mental health agencies. METHODS: A computerized prescription system was implemented in four not-for-profit mental health agencies in an urban setting as part of a larger study looking at reducing racial disparities. This brief report describes the implementation costs at one agency with ten full-time-equivalent psychiatrists for which information was available on time devoted to implementation by the management information system personnel. The financial costs of the computer network hardware and software were also documented. RESULTS: The total initial cost was $27,607: preimplementation cost, $3,720; technology and system integration cost, $10,148; and training cost, $13,739. Annual ongoing cost was expected to be $14,725. CONCLUSIONS: The technology expenditure itself is not prohibitive for initial implementation as well as for ongoing support.

Service use and cost in 2002 among clients in community settings who were discharged from a state hospital in 1989.

OBJECTIVE: This analysis is a follow-up of a 1992 study of service use and cost of care among patients discharged from a state hospital. This study documented utilization and cost of care in 2002. METHODS: Study participants were 150 former long-stay patients who were discharged from a state psychiatric hospital in 1989. An integrated database of all mental health and medical services reimbursed by Medicaid and Medicare as well as state- and county-funded services was used to construct descriptive statistics on service use and unit cost measures. Data on mortality, homelessness and criminal arrests were obtained from vital statistics, jail records and shelter admissions. RESULTS: During 2002, 18% of study participants had a psychiatric hospital admission, with a mean length of stay of 60 days. Almost all participants (99%) received some form of outpatient mental health care, and 66% were living in publicly funded residential housing. The total annual cost per study participant for the 2002 service package was $89,699. Residential care accounted for 54% of the total cost. Between 1989 and 2005, 3% were admitted to county jails and 9% had experienced homelessness. Of the original discharged sample of 590 patients, 37% died between 1989 and 2002; the mean age at death was 63+/-15 years. CONCLUSIONS: This analysis suggests that individuals discharged from state psychiatric institutions have been integrated into community residential settings and are receiving psychiatric outpatient treatment on a regular basis. Cost estimates, using the consumer price index for 2002, were $78,773 in 1992 compared with $85,850 for the exact same service package in 2002.

Peer support for persons with co-occurring disorders and community tenure: a survival analysis.

Individuals with co-occurring mental health and substance abuse diagnoses experience high rehospitalization rates. Consumer-delivered services are recognized as an important intervention for this population, but no studies have examined the extent to which such services are associated with enhanced community tenure and prevention of rehospitalizations. This longitudinal, comparison group study examines the effect of participation in The Friends Connection, a peer support program for individuals with co-occurring disorders, on 3-year rehospitalization patterns. Results from a survival analysis suggest that program participants have longer community tenure (i.e., periods of living in the community without rehospitalization) than a comparison group. Chi-square tests also indicate that significantly more people in the comparison group (73%) are rehospitalized in a 3-year period versus those in the Friends Connection group (62%). These results suggest that Friends Connection may facilitate community tenure and prevent rehospitalizations for a group that is at high-risk for rehospitalizations. The findings lend additional support of the potential effectiveness of peer support programs as part of a service delivery system that facilitates recovery of individuals with co-occurring disorders.

Comparison of alcohol treatment and costs after implementation of medicaid managed care.

OBJECTIVE: To examine the impact of a mandatory managed care behavioral health program on utilization and cost of alcohol treatment services for high-risk Medicaid patients. STUDY DESIGN: Pre-post nonequivalent comparison group design to compare managed care clients with fee-for-service (FFS) clients in terms of behavioral treatment costs and use. METHODS: Study subjects were adult Medicaid enrollees diagnosed with alcohol abuse or alcohol dependence. Chi-square tests and analysis of variance were used to determine significant differences between managed care and FFS programs in characteristics of the subjects, service use rates, and intensity of care. A regression model was used to examine predisposing, enabling, and need factors that might explain cost differences between programs. RESULTS: The managed care site had reduced behavioral healthcare costs compared with the FFS site. However, the regression analysis, which explained 35% of the variance in behavioral health service cost per user, showed that treatment cost was not significantly lowered by the managed care intervention once predisposing and need factors were controlled for. Nineteen percent of the variance in cost was explained by increased mental health comorbidity and 12% by drug comorbidity. CONCLUSION: Consistent with other studies, the results show lower behavioral healthcare costs after the managed care intervention because of changes in management practices, service substitution, and negotiation of lower hospital fees. However, the managed care influence was insignificant in explaining cost variation between sites due to higher morbidity in the FSS site post managed care.

Quality of Diabetes Care Among Adult Medicaid Enrollees With Mental Disorders.

OBJECTIVE: In light of the national trend toward integrating mental and general medical care, this study examined disparities in diabetes treatment among Medicaid recipients in a nonintegrated, managed care behavioral health carve-out system. METHODS: A retrospective study of Medicaid claims (July 2009-June 2010) compared quality of diabetes treatment among 21,015 patients with and without mental disorders. RESULTS: Presence of a mental disorder was associated with higher use of outpatient and primary care services for diabetes, lower rates of hospitalizations for diabetes, and higher odds of receiving three or more quality measures for diabetes care. Patients with serious mental illness had better diabetes care compared with patients with other mental disorders and patients with no mental disorders. CONCLUSIONS: Findings suggest that managed care behavioral health carve-out systems should be considered among the range of approaches for improving treatment for diabetes among persons with comorbid mental disorders, particularly serious mental disorders.

Racial disparities in antipsychotic prescription patterns for patients with schizophrenia.

OBJECTIVE: This study aimed to examine the extent and type of variation in antipsychotic prescription patterns between African American and Caucasian patients with schizophrenia. METHOD: Subjects were 2,515 adult Medicaid recipients treated for schizophrenia in 1995 with one of four types of antipsychotic medication (traditional antipsychotics, clozapine, risperidone, or depot antipsychotics). Prescription and mental health service use data were collected from Medicaid claims files for the 12 months following the first filled antipsychotic prescription. Patterns of antipsychotic prescription were compared for African American (N=1,538, 61%) and Caucasian (N=977, 39%) subjects. RESULTS: African American subjects were significantly younger and more likely to receive Supplemental Security Income than were the Caucasian subjects, who received mental health services more continuously. African American subjects were less likely than Caucasian subjects to receive clozapine (8% versus 15%, respectively) and risperidone (25% versus 31%) and more likely to receive depot antipsychotics (26% versus 14%). The likelihood of receiving clozapine or risperidone remained significantly different after demographic and service use characteristics were controlled. CONCLUSIONS: This study found ethnic disparities in antipsychotic prescription patterns among a large number of publicly insured clients treated for schizophrenia. Given the rapidly changing pharmacological treatment environment, these findings have significant implications for differential quality of care for African American patients. Future studies employing client and provider characteristics are urgently needed to test alternative explanations for ethnic disparities.

Trends in the rate and type of antipsychotic medications prescribed to persons with schizophrenia.

This article examines trends in antipsychotic medication use in a treated population of publicly funded patients with schizophrenia between 1991 and 1996. Findings from administrative claims data show that antipsychotic prescription rates increased from 79 percent to 83 percent between 1991 and 1996. Atypical antipsychotics were used by 39 percent of the population and comprised 41 percent of all antipsychotic agents prescribed compared to 59 percent for typical agents. Duration on a typical agent was 8 months versus 7.4 months for newer atypicals, with duration 11 months for those on clozapine. The highest switching behavior is found in users of atypicals (58% versus 25% for those on typicals) as is the percent of those who received an antidepressant concurrently with an antipsychotic, which was 44 percent for newer atypical users versus 31 percent for typical users. The lowest antidepressant use was among clozapine users (28%). Atypical users were more likely to be younger Caucasian men with higher use of inpatient and ambulatory mental health services compared to those on typical medications. The newer antipsychotic medications appear to be displacing traditional medications; however, contrary to what the literature suggests, duration is shorter and switching behavior and concurrent use of antidepressants is higher than in typical users.

Cost of care for Medicaid recipients with serious mental illness and HIV infection or AIDS.

OBJECTIVE: To assist in developing public policy about the feasibility of HIV prevention in community mental health settings, the cost of care was estimated for four groups of adults who were eligible to receive Medicaid: persons with serious mental illness and HIV infection or AIDS, persons with serious mental illness only, persons with HIV infection or AIDS only, and a control group without serious mental illness, HIV infection, or AIDS. METHODS: Claims records for adult participants in Medicaid fee-for-service systems in Philadelphia during 1996 (N=60,503) were used to identify diagnostic groups and to construct estimates of reimbursement costs by type of service for the year. The estimates included all outpatient and inpatient treatment costs per year per person and excluded pharmacy costs and the cost of nursing home care. Persons with severe mental illness, HIV infection, or AIDS had received those diagnoses between 1985 and 1996. RESULTS: Persons with comorbid serious mental illness and HIV infection or AIDS had the highest annual medical and behavioral health treatment expenditures (about $13,800 per person), followed by persons with HIV infection or AIDS only (annual expenditures of about $7,400 per person). Annual expenditures for persons with serious mental illness only were about $5,800 per person. The control group without serious mental illness, HIV infection, or AIDS had annual expenditures of about $1,800 per person. CONCLUSIONS: Given the high cost of treating persons with comorbid serious mental illness and HIV infection or AIDS, the integration of HIV prevention into ongoing case management for persons with serious mental illness who are at risk of infection may prove to be a cost-effective intervention strategy.

Outcomes of shelter use among homeless persons with serious mental illness.

OBJECTIVE: The purpose of this study was to examine the extent to which the use of case management services predicted public shelter use among homeless persons with serious mental illness after the termination of Access to Community Care and Effective Services and Supports (ACCESS), a five-year outreach and case management program. METHOD: The sample consisted of 475 Philadelphia ACCESS program participants. Client-level interview data and case manager service delivery records that were collected during the ACCESS intervention period were linked with administrative data on public shelter use for the 12-month period after the ACCESS program was terminated. By using Cox's proportional hazards model, multivariate analyses were conducted to test how the characteristics of the participants and the intensity of case management service use affected the rate of the first entry into a public shelter. RESULTS: Homeless individuals with serious mental illness who were younger, were African American, had fewer years of schooling, and had longer shelter stays during the ACCESS intervention period were more likely to enter shelters in the 12 months after the ACCESS program ended. Although use of vocational and supportive services was associated with a lower probability of shelter entry, use of housing assistance was associated with a higher probability of shelter entry. CONCLUSIONS: The study found that the total number of case management service contacts was not significantly associated with residential outcomes. Rather, the use of specific types of services was important in reducing the use of homeless shelters. These findings suggest that case management efforts should focus on developing vocational and psychosocial rehabilitation services to reduce the risk of recurrent homelessness among persons with serious mental illness.

Psychiatric service utilization and cost for persons with schizophrenia in a Medicaid managed care program.

A pre-post study design was used to look at changes in behavioral health care services and costs for Medicaid-eligible individuals with schizophrenia in a managed care (MC) carve-out compared to a fee-for-service (FFS) program in Pennsylvania between 1995 and 1998. Statistically significant reductions of 59% were found in hospital expenditures in the MC program compared to 18.3% in the FFS program. The decline in hospital costs was due to dramatic fee reductions in the MC site. No significant differences in overall ambulatory utilization were found in either program; however, ambulatory expenditures rose 57% in the MC program versus a decline of 11% in fee for service. The ambulatory cost increase resulted from a cost shift between county block grant funds, and Medicaid funds, with no additional revenues provided to outpatient providers. Study implications are that cost reductions from MC are mainly due to reducing utilization and payments to hospitals, similar to the findings for private sector programs.

Long-term effectiveness of the ACCESS program in linking community mental health services to homeless persons with serious mental illness.

This study examined the long-term effectiveness of the ACCESS (Access to Community Care and Effective Services and Supports) project on service utilization and continuity of care among homeless persons with serious mental illness. A 3-year longitudinal analysis, using Medicaid claims data, tracked behavioral health service utilization among 146 Medicaid-eligible participants in the Pennsylvania ACCESS program. Utilization patterns of inpatient, outpatient, and emergency department services for psychiatric and substance abuse treatment were examined during the year prior to, during, and one year after the implementation of the ACCESS project. Use of psychiatric ambulatory care significantly increased among intervention participants and remained greater following ACCESS intervention. Better continuity of care following hospitalization was achieved during and after the intervention. The number of days spent hospitalized significantly decreased during the intervention. These results suggest that the ACCESS intervention was effective in linking hard-to-reach homeless persons with serious mental illness to the community mental health service system, and that this effect was maintained after termination of the intervention.

Length of inpatient stay of persons with serious mental illness: effects of hospital and regional characteristics.

OBJECTIVE: This study examined the extent to which hospital and regional characteristics are associated with length of hospitalization among patients with serious mental illness. METHODS: Data from the Pennsylvania Health Care Cost Containment Council and 2006 American Hospital Association data were obtained. The sample consisted of 106 hospitals from which 45,497 adults with serious mental illness were discharged in 2006. Guided by the extended version of Andersen's health care utilization model, hierarchical linear modeling, including patient case mix, hospital, and regional characteristics, was used to explain variations in hospitalization length. RESULTS: The average length of stay was 10.0 ± 3.0 days. Stays were longer at psychiatric hospitals than at general acute care facilities and at hospitals with a greater percentage of Medicare patients and patients with serious mental illness and a higher rate of readmission. In terms of regional characteristics, stays were also longer at hospitals in counties where the county mental health program received a larger percentage of the state's mental health budget and a smaller share of the budget was used for residential care. CONCLUSIONS: Hospital type and case mix, along with the presence of housing resources funded by county mental health programs, were found to be associated with variations in length of hospitalization. Further research of a longitudinal or prospective nature is required to determine whether the availability of housing programs for persons with mental disorders leads to shorter hospital stays for those in crisis and to determine whether longer stays are the result of differences in hospital practices.