[Correction added on 20 May 2015, after online issue publication: Iwao Oshima’s affiliation has been corrected.]
Preliminary outcome study on assertive community treatment in Japan
Version of Record online: 26 JUL 2012
© 2012 The Authors. Psychiatry and Clinical Neurosciences © 2012 Japanese Society of Psychiatry and Neurology
Psychiatry and Clinical Neurosciences
Volume 66, Issue 5, pages 383–389, August 2012
How to Cite
Nishio, M., Ito, J., Oshima, I., Suzuki, Y., Horiuchi, K., Sono, T., Fukaya, H., Hisanaga, F. and Tsukada, K. (2012), Preliminary outcome study on assertive community treatment in Japan. Psychiatry and Clinical Neurosciences, 66: 383–389. doi: 10.1111/j.1440-1819.2012.02348.x
- Issue online: 26 JUL 2012
- Version of Record online: 26 JUL 2012
- Received 28 March 2008; revised 12 March 2012; accepted 24 March 2012.
- assertive community treatment;
- care management;
- mental health system in Japan;
- psychiatric rehabilitation
Aims: The beneficial effects of assertive community treatment (ACT), which has been widely acclaimed as being successful in several foreign countries, must also be objectively evaluated with respect to the transition from inpatient to community-based mental health treatment in Japan. This was the first study that examined effects of the ACT program in Japan using pre/post design data of the pilot trial of the ACT program in Japan project.
Methods: The study included 41 subjects hospitalized at Kohnodai Hospital, National Center of Neurology and Psychiatry between May 2003 and April 2004 for severe mental illness and who met inclusion criteria for entry regarding age, diagnosis, residence, utilization of mental health services, social adjustment, and ability to function in daily activities. All subjects provided informed consent for study participation and were followed for 1 year after hospital discharge.
Results: Comparison of the number of days and frequency of inpatient psychiatric hospitalization and frequency of emergency psychiatric visits between the 1-year period before hospitalization and 1-year period after hospital discharge showed a significant decrease in number of days and frequency of hospitalization. Comparison at 1 year after discharge with baseline showed no change in satisfaction with overall quality of life or Brief Psychiatric Rating Scale scores, but the Global Assessment of Functioning score significantly increased, and the antipsychotic dose (chlorpromazine equivalent) significantly decreased.
Conclusion: Despite some limitations in methodology and conclusions, this study suggests that ACT enables persons with severe mental illness to live for longer periods in the community, without worsening of symptoms, decreased social function, or deterioration in quality of life.
COMMUNITY MENTAL HEALTH care in Japan is undergoing a transition from hospital-based to community-based care. We are now in an era of evidence-based medicine emphasizing scientific principles. Assertive community treatment (ACT) programs, with deinstitutionalization of care, have been successful in other developed countries and have now been introduced in Japan with the hope of promising results.
ACT is a community-based program to help patients with severe mental illness who have required frequent or lengthy hospitalizations to live a decent life outside a hospital setting. Staff members comprise a team of experts in various disciplines that visit patients to provide most services. ACT is an intensive and comprehensive care management model to help persons with mental illness live in the community. Stein et al. have shown that an ACT program in the state of Wisconsin in the USA has significantly decreased hospital stays, stabilized living conditions, and increased satisfaction with services provided.1,2 Based on this success, several countries have introduced ACT programs into their community mental health care systems.3–12
This paper reports the beneficial effects experienced by patients in the pilot stage of the first ACT model program in Japan (ACT-J) started in May 2003 in the region covered by the Kohnodai Hospital National Center for Neurology and Psychiatry.
The primary aim of ACT-J is to reduce lengthy hospital stays of persons with severe mental illness who are high users of psychiatric services and enable them to live in the community with a stable and high quality of life. Entry into the study was based on patients meeting all of the following inclusion criteria: (i) inpatient psychiatric hospitalization between May 2003 and April 2004 at Kohnodai Hospital (excluding treatment for acute drug intoxication or complications); (ii) age 18–59 years; (iii) primary diagnosis of a mental illness, such as schizophrenia or mood disorder, at the time of admission (excluding primary diagnosis of intellectual impairment, dementia, drug or alcohol abuse, and personality disorder); (iv) residence in one of the three cities of Ichikawa, Matsudo, or Funabashi; (v) use of mental health services within the preceding 2 years (dating back from day of index hospital admission; hereinafter the same) as listed in Table 1 (I); (vi) antisocial behavior within the preceding 2 years as listed in Table 1 (II); (vii) a ‘no’ response to either of the two items in Table 1 (III) regarding daily activities within the preceding 1 year; (viii) total of ≥3 items corresponding to those for (vi) social adjustment and (vii) daily activities; (ix) consent of the patient's attending physician for ACT participation; and (x) voluntary consent given by the patient to participate after receiving an explanation on the research nature of the study.
|I Use of mental health care services within past 2 years|
|1. Hospitalization 1 or more times in 2-year period prior to current hospitalization|
|2. Use of emergency mental health-care services 3 or more times in 2-year period prior to current hospitalization|
|3. Patient discontinuation of outpatient treatment or failure to return for 3 months or longer after a scheduled outpatient visit|
|II Behavior with impaired social adjustment in previous 2 years|
|1. Violent or antisocial behavior (outside of family), such as property damage, shoplifting, robbery, or public disturbance|
|2. Drug or alcohol abuse over a period of 6 months|
|3. Disappearance, loss of residence, receipt of eviction notice, or homelessness|
|4. Suicide attempt|
|5. Abandonment, verbal abuse, or violence in family|
|III Activity in previous 1 year|
|• Do not assess based solely on period of symptom exacerbation just before hospitalization; select ‘yes’ if able to perform tasks listed below for 6 months or longer during 1-year period before hospitalization|
|• If living with family and daily assistance was provided with tasks listed below, assess based on whether could perform activity alone|
|1. Ability to function alone in social settings listed below for 6 months or longer (e.g. attend school, keep a job, use social resources like activity or daycare center, be responsible for household chores)|
|(1. Yes 2. No)|
|2. Ability to perform daily tasks alone, as an adult in the community, for 6 months or longer (e.g. personal hygiene, nutritional needs, safety control, manage personal documents, avoid dangerous situations|
|(1. Yes 2. No)|
A total of 922 persons were treated as psychiatric inpatients by Kohnodai Hospital between May 2003 and April 2004. For 868 patients (response rate: 93.9%), their attending physician submitted a checklist regarding inclusion criteria (ii)–(vii) mentioned above. There were 55 patients who met each of the inclusion criteria, but in six patients, the attending physician was reticent to allow their participation, and three others refused to participate. For the remaining 46 patients, initial contact with a case manager was made during hospitalization, and finally, 43 patients provided informed consent to participate in the ACT program. The baseline characteristics of the 43 patients meeting the inclusion criteria were ascertained by clinical staff members by review of the medical record and interview of the patient. This information was entered in a survey response format.
There were 19 men (44.2%) and 24 women (55.8%) with a mean age of 35.8 (SD = 10.5) years (range: 19–57 years). About 80% of all patients were in their 20s–40s. Diagnosis at the time of hospital admission was schizophrenia in 31 patients (72.1%), mood disorder in eight patients (18.6%) and other in four patients (9.3%). The mean duration of disease was 12.8 (SD = 8.4) years, and the mean frequency and number of days of inpatient hospitalization in the previous year was 1.7 (SD = 1.6) times and 119.4 (SD = 104.1) days, respectively. The mean frequency of emergency psychiatric visits to Kohnodai Hospital in the previous year was 3.1 (SD = 7.0) times. Marital status was: married, five patients (11.6%); never married, 32 patients (74.4%); and divorced, six patients (14.0%). Residential status was: living with family, 30 patients (69.8%); living with someone other than family, one patient (2.3%); and living alone, 12 patients (27.9%). Living expenses were covered by (some multiple responses): family working/pension, 29 cases (67.4%); own income from work, one case (2.3%); own income from pension, 15 cases (34.9%); own income from other assets, three cases (7.0%); public assistance, nine cases (20.9%); other, two cases (4.7%); and unknown, one case (2.3%).
Major features of the ACT program include: (i) treatment of persons with severe mental illness; (ii) staff members comprised of experts in various disciplines; (iii) a limited caseload to enable provision of intensive services (caseload: about 100 per team of 10 staff members); (iv) shared participation by all staff members in the care of each client; (v) team responsibility to ensure direct provision of integrated services, including health, medical, welfare, and occupational care; (vi) crisis intervention services 24 h a day, 365 days a year; and (vii) regular visits to the home and workplace to provide consultation and support in the client's actual living environment.3,13
A fidelity scale has been developed in the USA to rate adherence to the ACT model. ACT-J is based on this ACT model, with establishment of the entry criteria listed above. In June 2004, after enrollment in the pilot study was completed, there were 10 full-time case managers from a variety of backgrounds, including nurses, social workers, occupational therapists, and a team leader as well. Others included two part-time staff to provide employment assistance, a team psychiatrist, and the full-time equivalent of a program assistant. Staff members were assigned to maintain specified caseload standards. They used a team approach, including twice-daily meetings on weekdays, to provide assistance in the form of visits, with coverage 24 h a day, 365 days a year.14–16
A wide range of services is provided, including consultations, delivery of medications, accompanying clients to outpatient visits, information on diagnosis and treatment, counseling, crisis intervention with home and community visits, preparation of crisis plans (symptom self-management), working with attending physicians to decide if hospitalization is required, protecting client rights, and hospital discharge planning. Other services include contact with real estate services, help moving, preparing and delivering food, taking clients shopping, arranging for services like insect control, escorting clients to public baths, advice on bodily care, employment counseling (including arranging and escorting clients to job interviews), family counseling and relationships, and improving leisure activity skills, such as going to movies, singing karaoke, and playing sports. ACT is a care management model, and staff members, through a series of care processes, establish relationships with clients, including assessment, care planning, monitoring, and evaluation. Staff members provide daily assistance based on the individual needs of each client. The goal of assistance and clinical support emphasizes ‘recovery’ and a ‘strength model’.13
Outcome results as of the end of March 2007 were evaluated in 42 of 43 patients who were discharged after the index hospitalization. One patient remained hospitalized. Of the 42 patients discharged after the index hospitalization, one patient was lost to follow up within 1 year after discharge (the patient died). Therefore, the 1-year outcome results reported in this paper are based on data from 41 patients. The mean number of days of the index inpatient hospitalization was 79.1 (range: 1–321) days, and the mean length of time between the index admission date and the ACT intervention launch date was 28.0 (range: 3–160) days. As for the 41 patients, all of them had been provided for services by the ACT team for a year from entry, without dropouts.
Actually, as for the subject of this intervention study, the mean monthly number of times for visiting homes and for visiting wards during 1 year after hospital discharge from index admission was 10.5 and 5.9, respectively. An average of 4.02 staff (range: 1–11) provided any services for each client in the month.
In addition, regarding the same subjects, the number of provided services for the 1-year period after hospital discharge from index admission were as follows (for each component, in descending order): medical support for psychiatric symptoms, 3271; social life support, 1656; assistance with daily living tasks, 1096; care management, 1001; family support, 675; support for physical health, 511; communications and coordination, 499; vocational and educational support, 435; financial support, 393.
The research staff collected data for number of days and frequency of inpatient psychiatric treatment, frequency of emergency psychiatric visits, and number and dose of drugs from the medical record. In addition, psychiatric symptoms (Brief Psychiatric Rating Scale [BPRS]), social functioning (Global Assessment of Functioning [GAF]), psychotropic/antipsychotic drug dose (chlorpromazine [CP] equivalent [Treatment Resistant Schizophrenia – Research Group version]), and quality of life (Quality of Life Inventory [QOLI] short version) were prospectively surveyed. Those data were obtained by patient interview. In this report, the QOLI data indicate the results for overall life satisfaction.
Employment status in 41 patients (excluding two patients [one not discharged and one died]) was compared between the 1-year period before baseline and the 1-year period after discharge from the index hospitalization. Data were based on the QOLI data, or in patients not yet surveyed, from information obtained by clinical staff.
The study was fully explained to each patient, and all provided written informed consent for participation. The research protocol, including data collection and storage, was reviewed and approved by the Institutional Review Board at the National Center of Neurology and Psychiatry.
Based on the index hospital admission when the patient was enrolled in ACT, the number of days and frequency of inpatient psychiatric treatment and frequency of emergency psychiatric visits to Kohnodai Hospital (nighttime and holiday) were compared between the 1-year period before hospitalization and the 1-year period after hospital discharge. As for BPRS, GAF, CP equivalent and QOLI, using 2 weeks post-hospital discharge as baseline, the results were compared after 1 year.
Statistical analysis was conducted using spss 14.0 J for Windows. Quantitative changes in the above outcomes were analyzed using the Wilcoxon signed rank test. The significance level was set at P < 0.05 (two-tailed).
As shown in Table 2, all the values decreased. For number of days and frequency of inpatient hospitalization, the Wilcoxon signed rank test showed significant changes.
|Outcome measure||1 year before hospitalization||1 year after hospital discharge||Z value||P|
|No. of days of inpatient hospitalization||115.95||102.76||56.66||98.42||−3.25||0.001|
|Frequency of inpatient hospitalization||1.68||1.64||1.27||2.04||−2.12||0.034|
|Frequency of emergency psychiatric visits||3.20||7.16||1.85||3.71||−1.45||0.147|
Table 3 compares other outcome results between baseline and 1 year after hospital discharge. The Wilcoxon signed rank test showed significant changes in GAF, BPRS positive and negative symptoms, and CP equivalent. For other BPRS and drug-therapy-related parameters, there were no significant changes. The mean scores on the QOLI (scale of 1–7) for life satisfaction were similar at baseline and after 1 year, with no significant changes.
|Outcome measure||2 weeks after hospital discharge||1 year after hospital discharge||Z value||P|
|GAF (n = 38)||46.66||8.15||50.11||6.62||−3.06||0.002|
|BPRS score (n = 37)||16.05||6.39||16.35||7.05||−0.17||0.87|
|BPRS positive symptoms (n = 37)||5.41||3.63||6.73||3.60||−2.51||0.012|
|BPRS negative symptoms (n = 37)||3.03||2.90||2.03||2.66||−2.91||0.004|
|BPRS manic (n = 37)||0.22||0.63||0.41||0.69||−1.54||0.124|
|BPRS mood (n = 37)||5.11||2.40||4.84||2.50||−0.74||0.459|
|BPRS hypochondriasis (n = 37)||2.38||1.46||2.43||1.43||−0.03||0.975|
|QOLI overall satisfaction (n = 37)||4.05||1.61||4.07||1.68||−0.27||0.789|
|No. of psychotropic drugs||5.15||2.04||5.00||2.31||−0.63||0.53|
|No. of antipsychotic drugs||1.90||0.89||1.83||0.95||−0.55||0.58|
Comparison of employment status showed that during the 1-year period before baseline, three persons were regularly employed (defined as at least 5 h per week, at least minimum wage, in a public workplace), and two were engaged in social work activities. At 1 year after hospital discharge, four persons were regularly employed (defined by the conditions above), three were employed (but hours and salary did not meet above conditions), one was attending school, and one was engaged in social work activities.
Previous ACT studies, primarily conducted in the USA, have demonstrated beneficial results, including a reduction in the number of days of inpatient hospitalization, an increase in client satisfaction with services, and a more stable living situation.11,12
This ACT-J pilot study also showed a significant reduction in number of days and frequency of inpatient psychiatric hospitalization at 1 year after index hospital discharge. It may be premature to speculate about the relation between decrease in number of days of hospitalization and effects on quality of life. However, the outcome results in this study suggest stabilization of psychiatric symptoms, a similar level of quality of life, and a small but significant improvement in social functioning. In addition, the number of patients engaged in regular employment tended to increase, and the dose of medication required (CP equivalent) significantly decreased. Therefore, our findings suggest that ACT can reduce the utilization of inpatient psychiatric care without degrading quality of life of patients living in the community.
It seems difficult to completely deny the existence of a relation between the exacerbation of positive symptoms in BPRS and the significant CP equivalent decrease at 1 year after index hospital discharge. However, the correlation coefficient between the difference in the BPRS score baseline and after 1 year and the difference in the CP equivalent dose baseline and after 1 year was low (Spearman: −0.187) and not significant (P = 0.268). Furthermore, the exacerbation of positive symptoms was not unexpected due to the definition of baseline (2 weeks post-hospital-discharge), the relatively low BPRS positive symptoms score baseline (5.41) and the patient characteristics (frequent users of psychiatric care).
In contrast, the activities of the ACT team, which included frequent medical assessment outreach services to support users living independently, and direct emotional support for relatives, might have led to the reduction in the major tranquilizer dose, which was administered to sedate psychomotor excitation due to ‘high expressed emotion’ in the families.
In addition, the BPRS indicated significant exacerbation of positive symptoms and improvement of negative symptoms. This suggests that ACT services may have some ‘activation effects’ on patients. It is possible that this is due to some sort of stimulus, such as assertive outreach and social support.
During the 1-year period after hospital discharge, there were few patients who participated in social work activities, such as day care treatment and sheltered workshops. One reason for this is because our entry criteria focused on patients who were originally reluctant to use traditional rehabilitation facilities. Another reason is because the clinical philosophy in ACT, such as ‘recovery’ and ‘strength model’, emphasizes the importance of competitive employment rather than pre-employment training.
The all-around cost for ACT is not significantly cheaper than that of inpatient care, because of the intensive support for each patient.17 Therefore, this model focuses on the care of patients with severe mental illness, and specific entry criteria have been established. These criteria should also reflect the target goals of each ACT team, such as support for dual diagnosis patients. In conducting the ACT program on a research basis, we realized that Kohnodai Hospital is a general public hospital that serves the community in providing emergency psychiatric care and treatment of complications in a minimum number of inpatient days, and that by focusing on patients requiring long-term hospitalization, it might not be possible to recruit a sufficient sample size. Therefore, patients with severe mental illness who required frequent readmissions due to severe symptoms were considered for the study. To target patients with severe mental illness who were frequent users of psychiatric care, we selected patients hospitalized at Kohnodai Hospital at the time of program entry, and we also established inclusion criteria that had to be met regarding frequency of use of psychiatric services, social adjustment, and ability to function in daily activities.
Patient characteristics in this study included a relatively young mean age (in their 30s) and the fact that about 70% lived with their families. As compared to long-term patients of middle age with parents who have already died, this indicated more patients needing family support. Reports of ACT programs overseas have also emphasized family support.18 This suggests that when targeting high users of psychiatric care, family support of patients must be taken into consideration in establishing ACT programs in Japan.
The baseline survey indicated relatively good results for the BPRS, GAF, and QOLI. The mean BPRS score was low at 16 and the mean GAF score was relatively high at 46. In selecting patients with frequent hospitalizations, the inclusion criteria may have reflected medication or treatment discontinuation, thus resulting in the registration of many patients whose symptoms tended to improve even with relatively short hospitalization.
In addition, more so than the effects of inpatient treatment, this study was designed to evaluate the effects of ACT when the patient is living in the community. The baseline for resumption in community life was defined as 2 weeks after hospital discharge. Thus, when the patients were evaluated, they already had some improvement in psychiatric symptoms and ability to function in society. This factor probably contributed to the relatively good results on baseline evaluation.
The high overall life satisfaction indicated by the QOLI probably reflects the added relief felt by patients in being able to leave the hospital. The fact that ACT case managers began a relationship with patients while still in the hospital to assist them in living independently after hospital discharge probably also had an effect, although that is just an inference. Detailed analysis of the effects of ACT on patients while still hospitalized is an issue for further investigation in a controlled study.
The results of this study require careful interpretation in light of a before and after comparison without a control group, a relatively small number of patients analyzed for treatment effects, and the special clinical conditions under which the study was conducted by a project team.19
However, even with a sample size of about 40 patients, there was a reduction in number of inpatient days, which was statistically significant at a level of 1%. For patients enrolled in this study, during the 1-year period prior to ACT entry, the mean number of days of hospitalization was about 120, and the mean number of hospitalizations was slightly less than 2. This indicates a group of patients with episodes of symptoms requiring inpatient treatment about once every 6 months. The good outcomes 1 year later, including improved social functioning and decreased CP equivalent, are probably not merely due to the spontaneous course of disease after hospital discharge, but also due to the beneficial effects of ACT intervention.
This paper has reported outcomes in patients at 1 year after hospital discharge in an ACT-J pilot study. The methodology of this study as a before and after comparison has some limitations, but our findings suggest that ACT can reduce the utilization of inpatient psychiatric care without degrading the quality of life of patients living in the community. Further investigations should be conducted to confirm the beneficial effects of the ACT program model.
This study was conducted under a scientific research grant from the Ministry of Health, Labor, and Welfare in Japan (Mental Health Research Project) ‘Development of a Comprehensive Community Support Service System and Visits to Patients with Severe Mental Illness’ (2002–2004). None of the authors has any conflicts of interest associated with this study.
- 31998., . The PACT model of community-based treatment for persons with severe and persistent mental illness: A model for PACT start-up. NAMI,
- 4Designing the locality system. In : The Sainsbury Centre for Mental Health, North Birmingham MHT (eds). Locality Services in Mental Health; Developing Home Treatment & Assertive Outreach. The Sainsbury Centre for Mental Health, London, 1998; booklet no. 2..
- 6The development of a local service in Birmingham. In : Dean C, Freeman F (eds). Community Mental Health Care: International Perspectives on Making It Happen. Gaskell, London, 1993; 76–89..
- 9The Sydney experience. In : Dean C, Freeman H (eds). Community Mental Health Care: International Perspectives on Making It Happen. Gaskell, London, 1993; 35–48..
- 112003., . In The Cochrane Library, Issue 3. Assertive community treatment for people with severe mental disorders (Cochrane Review). Update Software, Oxford,
- 13Substance Abuse and Mental Health Services Administration (SAMHSA), Center for Mental Health Services (CMHS), The Robert Wood Johnson Foundation (RWJF). Assertive community treatment implementation resource kit draft version. 2002.
- 19Synopsis of the Daily Living Programme for the seriously mentally ill: A controlled comparison of home and hospital based care. In : Tyrer P, Creed F (eds). Community Psychiatry in Action: Analysis and Prospects. Cambridge University Press, Cambridge, 1995; 29–44..