Effects of psychosocial program for preparing long-term hospitalized patients with schizophrenia for discharge from hospital: Randomized controlled trial


Correspondence: Sayaka Sato, PhD, Department of Psychiatric Rehabilitation, National Institute of Mental Health, National Center of Neurology and Psychiatry, 4-1-1 Ogawa Higashi, Kodaira, Tokyo 187-8553, Japan. Email: sayakas@ncnp.go.jp



The aims of the present study were to revise the Community Re-entry Program–Japanese version and to review the effectiveness of the revised Program, named the Discharge Preparation Program.


This study was a randomized controlled trial. The Discharge Preparation Program (DPP) was the intervention condition (n = 26), and the usual rehabilitation program was the control condition (n = 23). Outcome indicators included factors that make patient discharge difficult (which nurses evaluated), psychiatric symptom, knowledge about the illness or medication, and the number of patients who were discharged within 6 months after the end of a program.


Significant improvements were found in the score of the ‘Issues on treatment compliance’ factor and the score of the ‘Autistic life’ factor, which are subscales of the Discharge Difficulty Scale, for the DPP group, on two-way ANOVA (‘Issues on treatment compliance’: F = 3.818, P < 0.10; ‘Autistic life’: F = 4.155, P < 0.05) These factors affected discharge outcome. Thus, the program may be capable of promoting discharge of long-term hospitalized psychiatric patients. With regard to the number of patients discharged in 6 months after the end of a program, there was no significant difference between both groups.


The present result is in agreement with past studies, and the DPP is useful in discharge support for patients with schizophrenia in Japan.

ACCORDING TO Mental Health Atlas 2005 issued by the World Health Organization, Japan has the largest number of psychiatric beds (28.4 per 10 000 people).[1] Moreover, the number of psychiatric sickbeds in Japan tended to constantly increase from 1960 to 2000, and a similar trend is seen only in South Korea among advanced countries.[2]

Psychiatric hospitals in Japan have many wards, and tend to deliver continuous psychiatric treatment by hospital staff. As a result, hospitalized psychiatric patients often end up being in hospital for many years. According to a patient survey in 2005, the average number of days in hospital of patients who have a mental or behavioral disorder, and were discharged from hospital, is 298.4 days, the longest length of stay among the disorders and diseases covered by the survey.[3]

The large number of psychiatric beds in Japan, and the associated psychiatric care centering on hospitalization and extension of hospitalization, have been targeted by overseas specialists for improvement.[4]

Under these circumstances, the Health, Labor and Welfare Ministry, in its document titled Vision of Reform for Mental Healthcare and Welfare, discusses long-term hospitalized patients, estimated to be approximately 72 000 throughout the country, who are judged to ‘be able to be discharged if the acceptance conditions are met’, and declares that the Ministry will eliminate such a situation over the next 10 years.

A number of activities have been initiated, thus far, as part of psychosocial programs for promoting the discharge of Japanese long-term hospitalized psychiatric patients, and some results have been achieved.[5-7] These various activities share a common approach, which complements traditional services such as the care by medical institutions and welfare support for housing or employment through hospitals or NPO networks, with psychological education on medication and disorders, as well as training programs for skills in daily-life and human interaction. A group led by R.P. Liberman of UCLA conducted activities similar to those of such programs in Japan, but in a more systematic manner. The group commenced its initiatives in the 1970s, to supplement the intensive case management model (ICM) or the assertive community treatment model (ACT), and based on the knowledge and experience gained in these activities, they developed the UCLA Social and Independent Living Skills Program (The SILS Program) from 1980 to 1990.[8] The SILS Program and its predecessors were composed of psychological education and skill training centering on Social Skills Training (SST), like the aforementioned programs in Japan. It is suggested that improvement of compliance and social functions can be expected to occur.[9, 10] The Community Re-entry Program[11] was translated into Japanese,[12] and the effects of the Japanese version were reported by Kumagai et al.[13]

The Community Re-entry Program[11] is based on the US system of mental health care and welfare, and on US social environments; and in that system, psychiatric patients are usually hospitalized for approximately 1 week, and the program is designed to be executed during this short period of hospitalization. The Community Re-entry Program–Japanese version[12] translates the tools and procedures used in the original version. But in light of our research, and with long-term hospitalized patients as the target of discharge promotion, it is necessary to revise the Japanese version, making it easier to use, based on other programs run in Japan.

Thus, this study had two objectives: (i) to revise the Community Re-entry Program–Japanese version so that its format and contents are best suited to cope with factors related to discharge of long-term hospitalized patients in Japan, the target of discharge promotion; and (ii) to review the effectiveness of the revised Program.


Revision of the Community Re-entry Program

Based on the current status of Japanese mental health care and welfare, the Japanese version is revised as follows, in order to make the conventional program tools more user-friendly and effective.

Major revision of content considered less familiar to Japanese participants

Because ‘church’, ‘YMCA’, and other facilities involved in the US program, as social resources for the community in which a patient lives, are not familiar institutions for Japanese patients, this session is completely deleted. Instead, we added new sessions to our program, which provide information about Japanese social resources and systems enabling patients to picture their lives in communities.

Because the US original program has multiple intensive sessions during a brief period, there is no homework for participants in each session. The Japanese program, however, was done once or twice a week over the course of 4–5 months, because it is feasible in Japanese medical settings. We therefore gave participants homework in individual sessions to prevent not only participants but also staff from having a lack of motivation throughout the program.

Development of a program tool for Japanese patients and staff

The workbook for participants used in the US program is only text, and it would be difficult for elderly participants to understand the workbook. The workbook was therefore revised to make it more user-friendly, for example using relevant pictures and figures. In addition, the US program incorporates a video containing various scenes and behaviors foreign to Japanese hospitals, such as a group-session room resembling a reception room, or characters overreacting. Therefore, although the basic framework of a given scenario is maintained in the Japanese program, such scenes were re-filmed using Japanese actors.

The staff manual for the US program describes only major points concerning how to conduct a session. This makes it difficult for Japanese staff who are inexperienced in group operation, to appropriately conduct sessions. Therefore, the manual's dialogues for sessions have been re-written using natural spoken language, so that staff inexperienced in group operation are better supported by the manual. Similarly, the title of the new Japanese program was revised to ‘Discharge Preparation Program for Mentally Disordered Persons’ (the DPP).[14] The DPP differs from conventional rehabilitation programs in Japan in that good evidence has been collected to support it, and it is structured with the use of tools.

Implementing the DPP

The DPP is implemented using psychological education or the SST method, and three primary tools: (i) video imaging; (ii) the leader's manual; and (iii) the participants' workbook; all deployed as in the US program. Essentially, a total of 24 sessions are conducted, including 17 indoor sessions and seven outdoor sessions (the Practice Program) that help participants use the knowledge and skills they learned indoors and gain a clear vision of living in the community. The standard procedure involves one or two sessions per week, with 60–90 min devoted to each session. The expected number of participants is 8–10 per group, and the leader's manual recommends a session be conducted by at least two staff members.

Research design

The effects of the DPP were examined using a randomized controlled trial (RCT) and the Waiting List technique. This research is a joint project of five hospitals. Therefore, a 2-day workshop was held by doctors familiar with SST and the Community Re-entry Program, prior to implementation of the DPP, so that no major differences in its implementation should occur among the hospitals.

There were a total of 57 subjects, which included 34 men and 23 women (average age, 53.46 ± 11.82 years) who had been diagnosed with schizophrenia and hospitalized at research-cooperating hospitals, either because of voluntary admission or admission for medical care and custody, and who agreed, in writing, to participate in the research. In the case of patients admitted for medical care and custody, their families' consent was obtained, in addition to that of the participating subjects.

If assigned to a control group, the consenting subjects were guaranteed to participate in the second program (at their request), after a waiting period of 6 months following the completion of the first program.


After having been classified by hospital of residence, age, and gender, the research subjects were assigned at random to one of two groups: (a) a group that would participate in the DPP (the participation group); and (b) a control group. Assignment was conducted by a clinical psychologist, who had never seen or met the subjects, using random numbers. As a result 29 were assigned to the participation group, and 28 to the control group.

After assignment, however, it was found that four of the subjects had been diagnosed with a disorder other than schizophrenia, and one had been hospitalized for <1 year. These subjects were ruled out from the research.

In addition, two subjects were dropped from the participation group and one from the control group during the implementation period of the DPP, after completion of the pre-test assessment. Thus, the final number of research subjects was 29 men and 20 women, totaling 49 (average age, 54.31 ± 11.70 years; average hospitalization, 175.48 ± 127.40 months), including 26 in the participation group (average age, 53.50 ± 11.12 years; average hospitalization, 170.44 ± 144.17 months) and 23 in the control group (average age, 55.22 ± 12.50 years; average hospitalization, 181.18 ± 108.31 months).

Assessment tools

The evaluation was based on the total score on the Discharge Difficulty Scale (DDS), the total score in the hypostatic factor, the total score on the Brief Psychiatric Rating Scale (BPRS),[15] and the total score on the knowledge test for the DPP. Of these, the DDS, newly developed by the authors and utilizing a questionnaire, is a scale with eight factors and 27 items based on free descriptions by nurses of ‘reasons their patients cannot be discharged’. The eight factors are ‘Issues on treatment compliance’, ‘Anxiety for discharge’, ‘Problem behavior’, ‘Inactive drive and decline in ADL’, ‘Autistic life’, ‘Physical complications’, ‘Possibility of suicide attempt’, and ‘Support from family’; results on the reliability and validity of this scale had been already reported.[16-18] The knowledge test for the DPP assesses the knowledge of a patient concerning matters relevant to the DPP. Although the items on this test are similar to those in the Community Re-entry Program, new sessions were added in the DPP, and eventually its knowledge test incorporated new items relating to those sessions. The new items were reviewed by a medical doctor versed in the Community Re-entry Program, its Japanese version, and the DPP.

Statistical analysis

One-way analysis of variance was conducted for the subjects, upon registration for research participation, to check the homogeneity of the facilities, using the facilities as an independent variable, and the average age and average length of hospitalization as dependent variables.

To determine the group homogeneity of the scores of variables at the time of the pre-test survey, a t-test was conducted for the subjects with five exceptions from the initial registered number (n = 52, but n = 49 for the knowledge test alone), using the group as an independent variable, and the pre-test score on each clinical evaluation scale as a dependent variable. The number of subjects for the knowledge test alone was 49 because the subjects who dropped out after registration did not agree to join the pre-test assessment.

To determine the effects of the DPP, two-way analysis of variance was conducted, using the group and time period as independent variables, and the score on each clinical evaluation scale before and after the DPP as dependent variables. For the subjects who were discharged from hospital during the implementation period of the DPP, analysis was conducted using the score on each scale surveyed at the time of discharge (n = 49, but n = 47 for the knowledge test alone). The number of subjects who took the knowledge test alone was 47 because of an inability to collect data from one subject in the control group who was transferred to a different hospital, and one subject who was discharged before the survey that followed the intervention.


Demographic data and pre-test scores

The average age and average length of hospitalization by hospital were checked for homogeneity, and the results indicated no significant difference among the hospitals (average age, F(4) = 0.363, n.s.; average length of hospitalization, F(4) = 0.127, n.s.). The homogeneity of the two groups was checked, for age, length of hospitalization, and score on each clinical evaluation scale before implementation of the DPP. Measures with missing data were excluded from analysis. A significant difference was found between the groups in the score on the knowledge test for the DPP (t = −3.183, P < 0.01). No significant difference was found in the other scores (Table 1).

Table 1. Subject baseline characteristics
(n = 52)ControlParticipantst
(n = 24)(n = 28)
  1. *P < 0.05. Control (n = 23) participants (n = 26). ADL, activities of daily living; BPRS, Brief Psychiatric Rating Scale; DDS, Discharge Difficulty Scale.
Age (years)55.2512.2353.0711.610.658
Length of stay (months)178.02107.06169.62140.260.240
DDS total score15.339.1415.118.080.095
Issues on treatment compliance3.673.423.682.96−0.013
Anxiety for discharge1.671.952.072.34−0.671
Problem behavior1.331.811.141.880.371
Inactive drive and decline in ADL2.501.982.431.730.139
Autistic Life2.002.021.931.650.140
Physical Complications1.331.460.791.131.489
Possibility of suicide attempt0.291.000.320.67−0.128
Support from family2.541.742.751.73−0.431
BPRS total score49.4211.8647.8910.550.490
Knowledge test total score3.962.316.583.41−3.183*

Effects of the DPP

Two-way analysis of variance was conducted, using the group and time period as independent variables, and the score on each clinical evaluation scale before and after implementation of the DPP as dependent variables. Measures with missing data were excluded from analysis, as was done for the demographic data and pre-test scores.

Total of BPRS scores

The analysis of variance indicated no significant difference between the main effect, and the interaction of the time period and the group.

Discharge difficulty scale

Analysis of variance indicated a significant difference and significant trend in the main effect of the time period and the interaction between the group and the time period in the ‘Issues on treatment compliance’ factor scores (time period, F = −6.012, P < 0.05; interaction, F = 3.818, P < 0.10). The simple main effect of the time period in the participation group was significant (t = 12.360, P < 0.01). It was also found that a significant difference existed in the ‘Autistic life’ factor score between the interaction of the group and that of the time period (F = 4.155, P < 0.05). The simple main effect of the time period in the participation group was significant (t = 4.490, P < 0.05). In the case of the other factor scores, no significant difference was found in the main effect and interaction of the time period and group.

Knowledge test for the DPP

Analysis of variance indicated a significant difference in the main effect and the interaction of the group and the time period (time period, F = 21.680, P < 0.001; group, F = 20.990, P < 0.001; and interaction: F = 8.484, P < 0.01). The simple main effect of the time period in the participation group and the control group was significant (participation group, t = 25.081, P < 0.001; control group, t = 21.680, P < 0.001). It also indicated that the simple main effect of the group in the post-test assessment was significant (t = −4.861, P < 0.001).

Number of discharged patients in 6 months after the DPP

Regarding the number of patients discharged in 6 months after the completion of the DPP, according to the intention-to-treat principle that all subjects originally assigned (including the dropouts) should be analyzed, six out of 28 in the participation group, and one out of 24 in the control group, were discharged from hospital (P = 0.107, Fisher's exact test; Tables 2, 3).

Table 2. BPRS and knowledge test scores
(n = 49) 


(n = 23)


(n = 26)

  1. ***P < 0.001, **P < 0.01. Control (n = 21) participants (n = 26). BPRS, Brief Psychiatric Rating Scale.
BPRS total scorePre49.7811.9948.5010.480.5180.3560.285
Knowledge test total scorePre3.762.306.583.4121.680***20.990***8.484**
Table 3. DDS scores
(n = 49) ControlParticipantsANOVA F
(n = 23)(n = 26)
  1. *P < 0.05, P < 0.10. ADL, activities of daily living; DDS, Discharge Difficulty Scale.
Issues on treatment compliancePre3.783.453.922.926.012*0.4433.818
Anxiety for dischargePre1.611.971.962.292.3060.2050.278
Problem behaviorPre1.351.851.151.936.208*0.4350.308
Inactive drive and decline in ADLPre2.611.952.581.700.0120.2860.602
Autistic LifePre2.092.021.881.700.5532.4754.155*
Physical ComplicationsPre1.391.470.851.161.6981.7130.344
Possibility of suicide attemptPre0.301.020.270.601.3190.1690.102
Support from familyPre2.651.702.691.740.8560.2030.254
Total scorePre15.789.0715.318.216.743*0.7262.084


The present study had two objectives: (i) to revise the Community Re-entry Program–Japanese version (Liberman, 1995 and translated by Inoue, 1998) so that its format and contents are best suited to cope with the problems unique to patients in Japan; and (ii) to review the effectiveness of the revised Program in terms of the factors hindering the discharge of patients.

Particular attention was paid to two concerns: (i) to adapt the Japanese program to the system and environment of mental health care and welfare in Japan; and (ii) to make the Japanese program user-friendly to Japanese participants and staff. When the revised Program was implemented, participants, who saw the video in the session, showed sympathy with the comments and concerns of the patients played by Japanese actors. Participants also said they found it easier to understand the social resources and institutional systems explained in the workbook, which are available to them in the community. As the sessions continued, participants were also seen working on the workbook at the bedside after the session, or diligently engaged in their homework, often actively discussing it with their attending nurses.

Many of the weekly homework assignments require participants to work together with their attending nurses, which will eventually increase the workload of the nurses. Given this, it was feared that nurses not directly involved in the conduct of sessions would complain about the homework. But once the program began, positive opinions were heard. For instance, attending nurses of participants discovered clear objectives in their nursing planning because of the homework assignments given in each session. The program offered new and clear incentives to the nurses, whose care of long-term hospitalized patients tends to become routine. Nurses and psychiatric social workers engaged in the conduct of sessions noted, positively, that reading the leader's manual of the revised program helped them lead the group, although they are inexperienced in such work. With the exception of hospitals in which sick wards used for research activity have disappeared due to reorganization, almost all the hospitals that participated in the present study still conduct the DPP as part of their clinical activities, even after the completion of the research. In light of these results, the revised Japanese version of the Community Re-entry Program is believed to be a user-friendly and practical tool for long-term hospitalized psychiatric patients in Japan, and for program operating staff.

Regarding the effects of the program, no major change occurred in the mental symptoms in either of the two groups as a result of participation in the DPP, but significant improvement was confirmed only in participants' scores in the ‘Issues on treatment compliance’ factor, and the ‘Autistic Life’ factor, on the DDS. We reported that these factors have affected the discharge outcome of long-term hospitalized psychiatric patients,[18] and the present results suggest that the DPP may be capable of promoting discharge of long-term hospitalized psychiatric patients.

Comparing our program with the module-format program by Liberman et al. and previous studies, Eckman et al. and Kopelowicz et al. reported the effects of the Medication Management Module and Community Re-entry Program in terms of disease consciousness and medication compliance.[9, 10] Regarding the social skills, including human interaction, Kumagai et al. reported that significant improvement occurred in the participants in the DDP, in the knowledge of disorder self-control and in the factors related to social functions measured by the Rehabilitation Evaluation Hall and Baker (REHAB).[13, 19] Xiang et al. concluded that the program was clinically significant because the participants in the program showed significant improvement in their mental symptoms and social functions, and this improved level was maintained for 6 months thereafter.[20] Furthermore, Smith et al. and Kopelowicz et al. reported, with respect to the knowledge test of the US version of the Community Re-entry Program, that significant improvement occurred in the program participation group, compared to the control group.[10, 21] These previous studies used different indices, and it is difficult to summarize their results in a few words. But they all come to the same conclusion, which is that participants improved in social functions and activity, and in knowledge of their disorder, results also confirmed by the present study.

Finally, we will describe the limits of this study, and our future possibilities. Regarding the effect study, there was no difference in the mental symptoms between the participation group and the control group. Thus, there is still need for study concerning the effects of the DPP in this respect. Further studies, with an increased number of subjects, are necessary. The result of Fisher's exact test conducted for the number of patients discharged in 6 months following the completion of the program was not significant. We consider that the following two factors influenced this result. First, in a previous study that showed good outcomes, the mean subject age was in the 30s,[9, 10] but in the present study this was in the 50s. In addition, average duration of hospitalization was >10 years in this study. This may explain the low level of knowledge or skill gain and the insufficient improvement in anxiety about discharge among participants in the DPP. Second, although 6 months follow-up was conducted using the waiting list method, the length of follow-up could be fairly short, particularly those whose average length of hospitalization exceeds 10 years. Further studies are needed to review the program for younger people, and to conduct an RCT with long-term follow-up.

Last, because the present study was intended to review the effects of a program to be provided to patients themselves, its evaluation focused only on factors presented by the patients. But in order to realize the successful discharge of long-term hospitalized psychiatric patients, improvement of their prospective living environment, such as a support system to help them lead stable lives in the community, is of significant importance, because this will help mentally challenged people live comfortably in a local setting. It is hoped that a system providing more comprehensive support for patient discharge and community re-entry will be established, by organically connecting such environmental support to patients via a program such as the DPP.


We are deeply grateful to clinical staff of the following hospitals: NHO Higashi-owari Hospital, NHO Kamo Psychiatry Medical Center, Kikuti National Hospital, NHO Shimofusa Psychiatric Center, NHO Tottori Medical Center, which contributed to research through the implementation of the DPP and data collection. This research was funded through a grant from Intramural Research Grant for Neurological and Psychiatric Disorders of National Center of Neurology and Psychiatry, Japan. The authors report no conflicts of interest.