Background: The Cochrane Schizophrenia Group (CSzG) is concerned with the evaluation of the evidence for people suffering from serious mental illness (SMI). It has developed and maintained the largest database of randomized/controlled clinical trial reports and studies of people with schizophrenia with 25,328 reports for 18,079 coded studies (1948-May 29, 2018). Getting patients out of hospital and into community living (deinstitutionalization) was a process that really took hold by the 1970s. This movement produced important changes at the level of the health system, putting emphasis on community services and appending new objectives to the partial results achieved with pharmacological treatments: (a) to prevent or reduce readmission, by providing more home and community-based treatment; (b) to improve engagement with service users; and, (c) to improve clinical, social and occupational outcomes.

Objectives: To produce a map of the available randomized evidence for mental health services interventions for SMI and identify areas where data synthesis may be beneficial.


Methods: This mapping study was carried out in accordance with the methodology of the Global Evidence Mapping Initiative (GEM). The CSzG register was searched in February 2019. We extracted the study data characteristics and the main PICO (population, intervention, comparison, outcome) question for each trial. The PICO questions were mapped into the current classification systems. A broad ordering of mental health services interventions will be developed from the available evidence. As a first approach, we focused on existing classification made by the European Service Mapping Schedule and others.



Results: The registry yielded 262 records – including six duplicated or misclassified reports (~108 studies). Most studies (91%) were conducted in the USA, Canada, Australia, the UK, and China. At this point we have extracted data and mapped PICO data into broad categories:

· Residential care (including studies about housing),

· Outpatient care (including different follow-up modalities),

· Self-care and voluntary care (including self-management information programs),

· Other categories.

Reporting sometimes hindered confident classification.

Conclusions: We found common elements across the trials but classification is impaired by unacceptably poor reporting. Most lower-middle income countries seem to be underrepresented, possibly indicating regional evidence gaps, and within the countries that have tested the effects of these service interventions, we are unclear if these centres truly serve a diverse population. The final results with the incorporation of all data will be presented at the Colloquium.

Patient or healthcare consumer involvement: No patient or healthcare consumer was involved in this project