This clinical trial aims to evaluate the effects of an art therapy intervention carried out in museums to facilitate the process of recovery in mental health. In total 100 users of mental health services will be recruited for this study. Participants in the trial will be assessed at baseline and at 3 months.

Art therapy

QUALITY ASSURANCE PLAN. The researchers will review and report the process during the trial covering participant enrolment, consent and eligibility to protect participants, including reporting of harm and completeness, accuracy and timeliness of data collection.

STANDARD OPERATING PROCEDURES. Participants will be recruited over a period of three months by mental health professionals (mental health nurses, psychologists, psychiatrists and occupational therapists). Participants will not receive financial reimbursement for taking part in this trial. Participants will be identified from outpatient services who meet all the study eligibility criteria (listed in section Eligibility). ICD-10 (International Classification of Diseases) and GAF (Global Assessment of Functioning) will be used to screen participants at recruitment.

DATA DICTIONARY. All variables collected in this study will be listed and described in a case report form, with associated guidelines, to ensure consistency in all gathered data. The following data will be collected in this trial:

Sociodemographical data Clinical history Global Assessment of Functioning Scale (GAF) Questionnaire about the Process of Recovery (QPR-15) Visual Analogue Recovery Scale Duke-UNC Functional Social Support Questionnaire

SAMPLE SIZE ASSESSMENT. The estimated sample size is n = 102, for a t test with two independent samples, with one tail, an effect size of 0.5, an α of 95% and a power of 80%.

For the validation of the QPR-15, taking into account the subject-item ratio of at least 5: 1 (15 items), and considering the minimum number of recommended subjects to perform an Exploratory Factor Analysis, The minimum number of participants must be 100.

Thus, estimating losses of 20%, the total number of participants estimated for the study is n = 127.

STATISTICS ANALYSIS PLAN. The change of means in the main outcome variable will be analyzed with the Student's t-test and if the criteria for its application with the nonparametric Mann-Whitney U test are not met. For the categorical secondary variables, the Chi square test will be used.For all comparisons, a confidence level of 95% will be used. The program R version 3.5.1 will be used. for all statistical analyzes.

PLAN FOR MISSING DATA. Each researcher is responsible for ensuring that any missing data will be reported as missing in the study database. Procedures can sometimes be considered when using statistical methods that fail in the presence of any missing values, or when in the case of multiple-predictor statistical models all the data for an individual would be omitted because of a missing value in one of the predictors. For analyzes involving multiple regression analysis, a multiple imputation approach will be considered and used if statistically sound, depending on the proportion and pattern of missing values.

METHODS TO ENSURE VALIDITY AND QUALITY OF DATA. Accurate and reliable data collection will be assured by verification and cross-check of the case report form (CRF) against the researcher´s records (source document verification). Source document verification will be conducted for 5% of data in subjects. Discrepancies and queries will be generated accordingly in the CRF for online resolution by the researcher at the site. In addition, the CRF data will be reviewed on an ongoing basis for medical and scientific plausibility.

Inclusion Criteria:

• Participants are 18 years or older.
• Participants have a diagnosis of ICD-10 F20-29 or F31.7 with a psychotic history.
• Participants agree to be part of the study by giving signed written consent.

Exclusion Criteria:

• Score in GAF <50 points.