Barfar 2017.

Study characteristics
Methods	Study design: RCT   Duration of study: 12 months; 2010 to 2011
Participants	Country: Iran  Income classification: upper‐middle income in 2010 to 2011   Geographical scope: urban  Healthcare setting: 3 university‐affiliated hospitals and participants’ homes  Mental health condition: serious mental disorders (schizophrenia, schizoaffective disorder, bipolar I disorder)  Population (mention whether patient, carer, or dyad)  1. Age: 15 to 65 years  2. Gender: both  3. Socioeconomic background: more than 60% of participants were unemployed   4. Inclusion criteria  a. Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM‐IV), diagnosis of schizophrenia, schizoaffective disorder, or bipolar I disorder b. Age 15 to 65 c. Previous history of psychiatric hospitalisation(s) d. Residing in catchment area of the hospital   5. Exclusion criteria 1. Mental retardation 2. Severe organic condition 3. Non‐Farsi‐speaking
Interventions	Interventions  Stated purpose: to determine whether providing services for patients with severe mental disorders (schizophrenia, schizoaffective, and bipolar I disorder) through an Aftercare Service programme improves quality of life and global functioning, increases patient satisfaction, and is cost‐effective when compared with routine conventional care  INTERVENTION 1 (n = 80) Name: Aftercare Service (2 groups, analysed together)  Delivered by: professionals (PHPs and CPs) Title/name of PW and number: general practitioner (GP), social worker, case manager; number not specified  1. Selection: GP and social worker with experience in psychiatric home health care  2. Educational background: not specified   3. Training: specially trained; no details given  4. Supervision: for home visiting team, care plan was established and reviewed by faculty member psychiatrist, and weekly meetings were held with psychiatrists to discuss problems and to establish decisions.   Intervention details: home visits by general practitioner and social worker for 25 patients who were non‐compliant, were difficult to engage, and/or had high service use (home visiting care group). Telephone follow‐up by case manager for 55 patients who did not meet above requirement for home visits (telephone follow‐up group)  1. Duration/frequency: monthly home visit (home visit team), phone calls before monthly outpatient follow‐up (telephone follow‐up group), psychoeducation sessions (frequency not specified), intervention duration 1 year   2. Content of intervention: home visiting care included assessing patients, co‐ordinating care, managing symptoms/medication, prescribing and adjusting medications, ensuring compliance, educating patient and family about the illness and about medications and warning signs of a relapse, recognising early phases of a relapse, managing a relapse (which may involve raising medication dosage or referring for hospital admission), and guiding the family on how to access supportive and community resources. Telephone follow‐up involved telephone calls by a case manager to encourage patients to attend outpatient clinic at the hospital for monthly follow‐up visits by a resident of psychiatry or a general practitioner, where medication management and patient education were provided. Both groups received family psychoeducation and social skills training in individual (in‐home visiting care) or group format (telephone follow‐up, 6 sessions)    CONTROL (type and descriptions) (n = 80) Existing services provided by outpatient or inpatient services inside or outside hospitals that might be contacted by patients. Outpatient services included visits by a psychiatric resident at the hospital who prescribed medications for the patient. No active follow‐up or any form of rehabilitation services  CO‐INTERVENTIONS: nil
Outcomes	Patients  1. 1‐Year rehospitalization rate*  2. Clinical Severity Index (CGI)  3. Global Functioning (GAF)  4. General symptoms of psychosis (PANSS)  5. Positive symptoms of psychosis (PANSS)  6. Negative symptoms of psychosis (PANSS)  7. Total symptoms of psychosis (PANSS)  8. Depression (HDRS)  9. Mania (YMRS)  10. Quality of life (WHOQOL‐BREF)  Carers  Nnil  Process/health workers  CSQ‐8: Satisfaction  Economic outcomes  1. Intervention costs 2. Patient and family costs  3. Total costs  4. Tables 3, 4, and 5  (asterisk for study's primary outcomes; star: outcomes that we have not reported in this review)  Time point: 0 months
Notes	Source of funding: Mental Health Office at Ministry of Health and Tehran University of Medical Sciences  Notes on validation of instruments (screening and outcomes): validated  Additional information: Hajebi et al. A multicenter randomized controlled trial of aftercare services for severe mental illness: study protocol. BMC Psychiatry 2013;13:17. Declarations of interest ‐ none  Handling the data: nil  Prospective trial registration number: IRCT201009052557N2
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"A total of 160 post‐discharge eligible patients were randomized into two equal patient groups using stratified balanced block randomization method. A psychiatrist at the center stratified patients by sex and severity of illness in an effort to insure that equal numbers of similar patients were allocated to each group"
Allocation concealment (selection bias)	Unclear risk	Allocation of participants between arms of the study was equal (allocation ratio 1:1). Randomisation was provided by an independent statistician at the medical university. Eligible patients were assigned to intervention or control (TAU) groups by stratified balanced block randomisation method with allocation concealment. A psychiatrist at each centre was responsible for concealment procedures. It is unclear how allocation was concealed
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	The nature of such services prevents adequate blinding of participants. It was not possible to blind raters who will perform follow‐up evaluations on cases and controls because of direct contact with patients when rating them. It was not possible to conceal knowledge of allocation to groups. Lack of blinding of personnel and participants may not have influenced study results
Blinding of outcome assessment (detection bias) all outcomes	High risk	Blinding of outcome assessors was not performed
Baseline outcome measurements similar	Unclear risk	Baseline outcomes have been clearly reported in the table. Although similar between groups, it is unclear whether any differences were adjusted for during analyses
Baseline characteristics similar?	Low risk	Patients in experimental and control groups were comparable at baseline with regard to demographic and clinical data, with no significant differences (Table 1)
Incomplete outcome data (attrition bias) Efficacy data	Low risk	The proportion of missing data was similar in intervention and control groups and therefore was unlikely to bias the results
Incomplete outcome data (attrition bias) Safety data (e.g. adverse events)	Low risk	The proportion of missing data was similar in intervention and control groups and was not linked to adverse events
Protection against contamination	Unclear risk	As control group received treatment as usual with visits from psychiatrists, it is unclear whether communication between intervention and control could have occurred
Selective reporting (reporting bias)	Low risk	All outcomes reported in trial protocol and registry have been reported
Other bias	Low risk	No other biases were present