|Home | About | Journals | Submit | Contact Us | Français|
This study assessed the effectiveness of a previously tested model, Critical Time Intervention (CTI), in producing an enduring reduction in homelessness for persons with severe mental illness who were discharged from inpatient psychiatric treatment facilities.
A total of 150 previously homeless men and women who were diagnosed with severe mental illness and were discharged from inpatient psychiatric hospitalization were randomly assigned to receive either usual care or usual care plus CTI. The nine-month CTI intervention aims to gradually pass responsibility for providing ongoing support to community sources that will remain in place after the intervention ends, thereby leading to a durable reduction in the risk of future homelessness. Participants’ housing status was assessed every six weeks for eighteen months via participant self-report collected by interviewers blind to condition.
In an intent-to-treat analysis, participants assigned to the CTI group had significantly less homelessness at the end of the follow-up period (the final three six-week intervals) than did those assigned to the control group (OR=0.22, 95% CI=.06--.88).
We have shown that a relatively brief, focused intervention for persons with severe mental illness led to a reduction in homelessness that was evident nine months after the end of the intervention. This work suggests that targeted, relatively short interventions applied at critical transition points may enhance the efficacy of long-term supports for persons with severe mental illness living in the community.
While a growing body of research has begun to assess the effectiveness of a variety of outreach, housing and treatment approaches in meeting the complex needs of persons with mental illness who are homeless or residentially unstable, there remain very few randomized trials of carefully documented models intended to reduce the incidence of homelessness among persons with severe mental illness. The period following hospital discharge is one in which persons with severe mental illness are at high risk for either first- episode or recurrent homelessness (1,2). This is also a high risk period for other adverse outcomes including suicide (3–5), psychiatric re-hospitalization (6,7) and violence against others (8). Reasons for this may include both the impact of enduring psychiatric symptoms that preceded the inpatient episode plus difficulties in establishing effective connections to sources of treatment, housing and support in the community. Furthermore, like many transitional periods, the time following institutional treatment is commonly a period of significant personal disequilibrium.
Critical Time Intervention (CTI) aims to prevent recurrent homelessness and other adverse outcomes following discharge in two ways: by strengthening the individual's long-term ties to services, family, and friends; and by providing emotional and practical support during the critical time of transition. An important aspect of CTI is that post-discharge services are delivered by a worker who has established a relationship with the client before discharge. CTI shares with long-term assertive community treatment models a focus on promoting in vivo development of independent living skills and building effective support networks in the community (9). The emphasis, however, is on maintaining continuity of care during the critical period of transition while primary responsibility gradually passes to existing community supports that will remain in place after the intervention ends. Such an approach, we believe, increases the likelihood that the impact of a time-limited intervention will persist beyond its actual endpoint, which is the primary goal of CTI.
Previous studies provide evidence of CTI’s effectiveness in producing an enduring positive effect on housing outcomes among homeless adults with severe mental illness. The first was a randomized trial among 96 homeless men with severe mental illness placed into housing from a large shelter institution in New York City (10). In this study, a nine-month CTI produced a significant reduction in post-discharge homelessness which, crucially, remained evident up to the end of an eighteen month follow-up period. Subsequently, a six month version of CTI was tested at eight Department of Veterans’ Affairs medical centers. A quasi-experimental study compared the outcomes of two cohorts of homeless persons with severe mental illness after discharge from inpatient psychiatric treatment. One cohort (n=278) received standard case management and the other (n=206) received CTI. Controlling for baseline differences, investigators found that the CTI cohort had significantly more days housed than did the comparison cohort over the one-year follow-up period and that this effect also persisted after the intervention ended (11).
In this article, we report the results of a single-blind randomized trial testing whether nine months of CTI would produce an enduring effect in preventing homelessness among formerly homeless men and women with severe mental illness following discharge from inpatient psychiatric treatment. The trial was approved by the discharging hospital’s institutional review board under a multi-site collaborative agreement overseen by the NYS Office of Mental Health.
Figure 1 summarizes sample screening, accrual and retention. Our original design called for participants to be recruited during their inpatient stay at a state-operated psychiatric hospital located in the NYC metropolitan area. However, before recruitment began, policy changes in the mental health system led most patients with housing difficulties to be discharged to transitional residences located on the grounds of this hospital and a second nearby hospital. We therefore altered our design to recruit participants at these two residences rather than on the inpatient wards. Persons were considered eligible if they met the following criteria: 1) currently living in one of the two designated transitional residences following hospitalization during the four-year recruitment period (2002–2006) and discharged from the residence before the end of this period; 2) a lifetime DSM-IV diagnosis of a psychotic disorder (codes 295.xx, 296.xx and 298.9); 3) homeless at the index hospitalization or an episode of homelessness within eighteen months preceding this admission; and 4) spent their first night after leaving the transitional residence in New York City in a place other than a jail or a hospital (so that all subjects were at equal risk of homelessness during the observation period and those assigned to the CTI condition would be accessible to the CTI worker). We excluded those who were unable to provide informed consent for the screening interview or did not speak sufficient English to participate. We also excluded those who did not stay more than three weeknights in the transitional residence or whose employment schedule made them unavailable to project staff during regular work hours.
Research staff approached individuals identified by clinical staff as having a plan for discharge to New York City. After obtaining written consent for the screening interview, research staff determined whether they had been homeless (defined as staying overnight in a shelter, on the streets, in a park, on a subway train, or in any other public space) during the eighteen months prior to hospitalization. Those who met this criterion were then provided with the study description and their written informed consent for participation in the trial was obtained. Eligible participants completed a baseline interview, including the Structured Clinical Interview for DSM-IV (SCID) (12); this was used to verify that the participant met our diagnostic eligibility criterion. Those not meeting this criterion were excluded at this point.
Participants were randomized independently by gender and by diagnosis of lifetime substance use disorder. To reduce variation on key factors, we randomized individuals in these four strata in permuted blocks of randomly varying size. The names of eligible participants and their respective randomization stratum were given to an administrator who did not need to be blind to treatment status. Working from a list produced by our statistician of identification numbers with associated random treatment condition assignments, she assigned each participant the next available identification number within the designated stratum. For participants assigned to the CTI condition, she then notified the CTI clinical supervisor who added the participant to the CTI caseload. 32 randomized participants were subsequently dropped from the study because they were never discharged from the transitional residence during the recruitment period or because their first post-discharge night was spent either outside of New York City, in jail, or in a hospital. This procedure was approved a priori by the study’s biostatistician.
The Personal History Form, employed extensively in our earlier research with this population, was used at baseline to measure demographic characteristics and personal history, including prior use of treatment services, residential history, and previous homeless episodes (10). Following discharge from the transitional residence, participants were interviewed every six weeks for eighteen months in order to document where they had spent each night during the respective follow-up period. These assessments were carried out by trained interviewers blind to the participant’s group assignment. In cases in which a participant had missed an interview, the interviewer documented where the participant had spent each night since the last completed assessment. We chose to collect data at six-week intervals for two reasons. First, we sought to minimize reporting inaccuracies on the part of study participants with respect to their housing experience. Furthermore, our previous experience with this population suggests that more frequent contact between research staff and participants reduces the likelihood of participants being lost to follow up. In some instances when participants could not be directly interviewed, we gathered residential data from a family member, caseworker, or another of the participant’s close associates who we had been granted permission to contact. We have reported previously on a test-retest study demonstrating that homelessness could be assessed with high reliability (kappa=.93) using this approach (10). Participants received $20 for each of the baseline, nine and 18-month interviews and $10 for each six-week interview they completed.
While living in the transitional residence, all participants received basic discharge planning services and access to psychiatric treatment. After discharge, participants in both conditions received a range of “usual” community-based services depending on the individual’s needs, preferences and living situation. These services usually included various types of case management and clinical treatment. 12 participants (8%) were assigned to mandatory outpatient treatment and/or assertive community treatment programs.
In addition to the services noted above, participants randomly assigned to the experimental condition received nine months of CTI following discharge from the transitional residence. Those assigned to the control condition received usual services only. Post-discharge housing arrangements were typically coordinated by discharge planning staff located at the transitional residence. These arrangements ranged from community residences and other structured programs to supported apartments and independent housing, either alone or with family members. Neither CTI workers nor research staff were involved in determining the initial housing arrangement for individuals in either condition. Some individuals left the transitional residence “against medical advice” and returned to shelters or the streets but were nonetheless retained in the study.
CTI is described in detail in our previous publications (13,14). In brief, it is a nine-month case management intervention delivered in three phases, each of which lasts approximately three months (see Table 1). Phase one--transition to the community--focuses on providing intensive support and assessing the resources that exist for the transition of care to community providers. Ideally, the CTI worker will have already begun to engage the client in a working relationship before he or she moves into the community. This is important because the worker will build on this relationship to effectively support the client following discharge from the institution. The CTI worker generally makes detailed arrangements in only the handful of areas seen as most critical for community survival of that individual. Phase two—try out-- is devoted to testing and adjusting the systems of support that were developed during phase one. By now, community providers will have assumed primary responsibility for delivering support and services, and the CTI worker can focus on assessing the degree to which this support system is functioning as planned. In this phase, the worker will intervene only when modification in the system is needed or when a crisis occurs. Phase three—transfer of care-- focuses on completing the transfer of responsibility to community resources that will provide long-term support. One way in which CTI differs from services typically available during transitional periods is that the transfer of care process is not abrupt; instead, it represents the culmination of work occurring over the full nine months.
CTI was delivered by three workers trained by several of the model developers. Two were bachelors level employees of the NYS Office of Mental Health re-assigned to this project from their regular duties. The third worker, who also performed some supervisory activities, was a more experienced worker who had previously delivered CTI in an earlier trial. Weekly supervision was carried out by clinically trained staff experienced in the model.
Figure 1 summarizes the flow of participants from screening through follow-up. Of the 150 participants randomized, 77 participants (51%) were assigned to the experimental condition (CTI) and 73 (49%) were assigned to the control condition (usual care). 58 participants (75%) assigned to the experimental condition completed the eighteen month follow-up period, while 59 participants (80%) assigned to the control condition completed the full follow-up period. Complete follow-up data were obtained for significantly more males than females (85% of males vs. 58% of females, (chi square 12.7, df 1, p=.001). Those with a substance dependence diagnosis were also more likely to have complete follow-up (91% of substance dependent vs. 61% of non-substance dependent, chi square 18.9, df 1, p<.001). There were no other group differences in loss to follow-up that were related to baseline characteristics, including prior homelessness.
Some participants assigned to the experimental condition did not receive all components of the intervention. In particular, a key ingredient of the CTI model is that post-discharge services are provided by a worker who has established a relationship with the client before he or she is discharged from the institution to the community. Workers were instructed to develop this relationship via multiple face-to-face contacts with the participant during the pre-discharge period. In our previous work, we have established a threshold of at least three such pre-discharge contacts as minimally sufficient for this purpose (15). In the current study, 42 participants (56%) received three or more such contacts while 35 (44%) received two or fewer contacts. The failure to deliver the desired number of pre-discharge contacts was most often the result of limited time between participant randomization and discharge from the transitional residence. As noted earlier, our original plan was to recruit participants during their inpatient stay where delivering multiple pre-discharge contacts would not have presented a problem. Instead, we recruited in the transitional residences (in response to policy changes described above) where time to make contact with participants was limited by a variety of factors including significantly greater unpredictability in participants’ discharge dates.
We defined as our primary outcome a dichotomous measure of homelessness during the last three follow-up intervals (18 weeks) of the study. Since the primary goal of CTI is to produce a long-lasting effect on homelessness risk, the key test of its efficacy is whether the risk of homelessness is reduced at the end of the observation period. We chose the final three intervals because we felt that the final six-week observation interval itself was too short to generate a stable estimate of the treatment effect. Although we also measured the actual number of nights homeless reported by participants during each follow-up interval, the highly skewed distribution of homeless nights we anticipated (and observed) led us to use the dichotomous measure as our primary endpoint.
While the proportion of the sample for whom we were unable to obtain complete follow data was low relative to other studies, it was not insubstantial. In order to verify the robustness of our findings, we carried out a multiple imputation procedure employing five imputations and repeated the analysis. Our findings remained virtually unchanged.
Our primary analysis was an intent-to-treat (ITT) comparison testing whether there was a group-level difference between participants assigned to the two groups on risk of homelessness during the last three follow-up intervals of the study. This analysis was carried out using logistic regression adjusting for baseline homelessness (number of homeless nights during the three month period preceding the index hospitalization). As noted above, we elected to use a dichotomous measure of homelessness (ever versus never homeless in the last three intervals) as our primary endpoint because the distribution of homeless nights was highly skewed. We chose not to employ survival analysis because such models are best suited for questions in which the primary interest concerns the length of time to the occurrence of an event, and our main interest was whether the CTI and usual services groups differed at the end of observation period.
We also conducted several secondary analyses. First, we tested whether there was a group-level difference between participants assigned to the two groups on total number of homeless nights during the final three intervals. We used a Poisson regression model, adjusting for baseline homelessness.
Second, we performed an as-treated analysis in order to obtain an unbiased estimate of receipt of a version of the intervention that included three or more pre-discharge contacts with the worker. This analysis used the same outcome as in our primary ITT analysis (i.e. ever versus never homeless in last three intervals). The analysis was carried out via a two-stage instrumental variables regression in which assignment to condition was the instrument and receipt of the intervention including three or more pre-discharge contacts with the worker was the treatment indicator (16). Again we adjusted for baseline homelessness.
Third, we examined homelessness over the full 18-month follow-up period. Using logistic regression, adjusting for baseline homelessness, we first compared the two groups with respect to risk of homelessness (ever versus never homeless) over 18 months. Using Poisson regression, adjusting for baseline homelessness, we then compared the two groups with respect to number of nights homeless over 18 months.
Baseline characteristics of the sample are summarized in Table 2. There were 150 participants (107 men [71%] and 43 women [29%]), with a mean age of 37.5 ± 9.5 years. 93 participants (62%) were African-American. 92 participants (61%) had a lifetime diagnosis of schizophrenia, while 52 (35%) were diagnosed with schizoaffective disorder. Co-morbid substance use disorder was common, reflecting the fact that the sample was drawn from persons who were previously homeless and difficult to place into housing; 80 participants (53%) met criteria for lifetime substance dependence and another 40 (27%) percent met criteria for lifetime substance abuse without dependence. Most participants had extensive prior homelessness; 116 (79%) reported two or more previous homeless episodes and 50 (34%) reported five or more such episodes. 58 participants (40%) reported that the duration of their longest previous homeless episode was one year or longer. The average length of stay in the transitional residences was 162 days (SD=161). There were no marked differences at baseline between the CTI and the usual care group on any of these measures and none was statistically significant employing chi square tests for categorical variables and t-tests for continuous variables (p < .05).
Of the 117 participants with complete follow-up data, 31 (27%) experienced at least one homeless episode during the course of the study. Figure 2 shows the proportion of participants in each condition who experienced at least one homeless night during each of the follow-up intervals. Among those with complete follow-up data, 3 out of 58 (5%) of subjects assigned to the CTI group experienced homelessness during the final three follow-up intervals, while 11 out of 59 (19%) of subjects assigned to the control group experienced homelessness during this period. Using logistic regression to model the impact of assignment to CTI on homelessness during the final three follow-up intervals while controlling for baseline homelessness, the odds ratio for treatment assignment was 0.22 (95% CI=.06--.88). Thus assignment to CTI was associated with a statistically significant five-fold reduction in the odds of homelessness compared to assignment to usual care only. We then repeated these analyses adjusting for sex, ethnicity and age and the results were unchanged.
Next we compared the total number of nights homeless for each group during the final three observation intervals. Among those assigned to CTI there was a total of six homeless nights during this period, while among those assigned to the control group, there were 20 homeless nights. In the Poisson regression model adjusting for baseline homelessness, this difference was statistically significant at the p<.001 level.
We then compared the outcomes of participants assigned to CTI who received three or more pre-discharge contacts by the worker with experimental participants who received fewer than three such contacts. In order to obtain an unbiased estimate of the impact of receiving this version of the intervention while controlling for the impact of all possible confounders, we employed an instrumental variables approach (16). We used a two-stage logistic model with homelessness during the last three intervals as the outcome, group assignment as the instrument, and receipt of three or more pre-discharge contacts as the treatment, while again adjusting for baseline homelessness. In this analysis, we found that receipt of CTI with three or more pre-discharge contacts was associated with a statistically significant ten-fold reduction in the odds of homelessness compared to assignment to usual care only (OR = 0.10, 95% CI=.03--.35). Finally, we repeated these analyses separately for men and women and found similar results.
Finally, we examined the impact of the intervention on homelessness over the full 18-month follow-up period. We first used logistic regression to model the impact of assignment to CTI on a dichotomous measure of homelessness over the 18 months, again controlling for baseline homelessness. The odds ratio for the CTI group was 0.28 (95% CI=0.78--1.02), indicating that assignment to CTI was associated with a substantial reduction in the odds of homelessness in comparison with assignment to usual care, although the result was at the borderline of statistical significance. We then compared the total number of homeless nights accrued by participants in both groups. Among those assigned to CTI there were 1812 total homeless nights, while among those assigned to the control group, there were 2403 homeless nights. Using Poisson regression to control for baseline homelessness, this difference was statistically significant at the p<.001 level.
This is one of very few randomized trials of a carefully specified intervention designed to produce a lasting reduction in the risk of recurrent homelessness among persons with severe mental illness following discharge from inpatient psychiatric treatment. Our results support the findings of prior studies suggesting that CTI is an effective strategy for assisting formerly homeless individuals during the period of transition from institutional to community living. The magnitude of the protective effect was substantial, ranging from a five-fold reduction in homelessness risk in the intent-to-treat analysis to a tenfold reduction in the as-treated analysis. As in our previous trial (10), the results are consistent with our view that a targeted, time-limited intervention applied during a high-risk period can have a salutary impact that endures beyond the point at which the intervention ends.
Despite its significant strengths, the study has several limitations. First, as noted above, we were not able to deliver the full intervention to all participants assigned to the experimental condition. This may account for the different pattern of results obtained relative to our original trial. In this study (as illustrated in figure 2), it took some time for different outcomes between experimental and control groups to become apparent. We attribute this to the fact that a significant number of experimental participants did not establish an adequate connection with the CTI worker before leaving the transitional residence and this delayed the effect of the intervention. This is consistent with the stronger observed effect among participants who had more pre-discharge contact with the CTI worker.
Second, while the length of follow-up is relatively long by the standards of much clinical research, the degree to which the impact of CTI endures over a still longer time period is unknown. Finally, homelessness was defined as the primary outcome and therefore is the sole domain we examine in this first report. We have, however, collected data relevant to a number of other important outcomes that will be investigated in subsequent analyses. We must also acknowledge the death of two study participants (one from cancer, one from pneumonia) who were assigned to the usual services group. We found no evidence that these adverse outcomes, both of which were reported in a timely way to the IRB, were directly related to or bear on the results of the study.
We have now shown through randomized trials in two different contexts that CTI had a substantial, lasting impact on reducing the risk of recurrent homelessness in persons with severe mental illness following re-entry to community living, results consistent with others’ findings of a similar enduring effect in a non-randomized trial (11). While the need for comprehensive, long-term support for persons with severe mental illness is undeniable, this work suggests that targeted, relatively brief interventions applied at critical transition points may enhance the effectiveness of long-term supports for persons with severe mental illness living in the community.
In the current study, we have applied such an intervention at one such time point, however, we believe that there other “critical times” at which such interventions may also be effectively deployed; for example in persons with severe mental illness being discharged from jail or prison (17). CTI might also be also be further adapted for transitions that, while not involving institutional discharge, nevertheless pose significant challenges to establishing effective and enduring supports (18). For instance, approaches based on similar principles may be relevant for persons transitioning from more intensive service models, such as assertive community treatment, to less intensive ongoing sources of support (19) or for new users first entering the service system. Future research will be needed to determine if and how the CTI model can be successfully employed in these and other circumstances.
Research supported by NIMH grant R01MH59716