A837was tolerability related to the number of withdrawals patients in each study, due to the presence of adverse events or treatment failure. The analyses were performed using software Addis (v.1.16.5) and RevMan (5.1). Results: A total of 979 documents were initially identified and 11 of them met the selection criteria to meta-analysis. No significant differences were observed between the number of withdrawals patients due adverse events in any meta-analysis of control versus intervention. The odds ratio ranged from 0.68 (CI 032-1.45) to placebo versus asenapine, 1.37 (CI 0.29-1.33) to placebo versus iloperidone and 0,71 (CI 0,36-1,41) to placebo versus lurasidone. However, all drugs were superior to their respective controls for the outcome of number of withdrawals by treatment failure, with odds ratio between 1.70 (CI 1.21-2.39) and 2.36 (CI 1.36-4.07). These results suggest that there is a higher effectiveness among patients for the treatment intervention that should be evaluated through clinical responses. Heterogeneity between studies (evaluated by I2 values) were low or moderate, not superior than 39,5% in any meta-analysis. ConClusions: Information and knowledge reunion and confrontation on the tolerability profile of a particular drug allows safer decisions over the therapeutic approach, focused on patient's interest which directly reflects on treatment follow-through and therapy effectiveness. In this study, we report evidence on asenapine, iloperidone and lurasidone greater tolerability profile compared to placebo in schizophrenia treatment.

Extended psychometric analysis was done using the factor analysis (Horn´s Parallel Analysis, Exploratory Factor Analysis) and Rasch analysis (RA, Rating Scale model) (Scale diagnostic, Validity, Reliability, Dimensionality and Local independence, Differential item functioning (DIF)). Results: Horn´s Parallel Analysis revealed one factor (loading factor > 0.40). RA showed a correct functioning of the rating categories of the scale. As for the item fit, only one item 'It is easy to put on my device' slightly underfitted the model (Outfit MSQ = 0.720, Infit MSQ = 0.650) and item 'durability' overfitted the model (Outfit MSQ = 1.378, Infit MSQ = 1.291). The study showed a few similar allocations of items along the logit scale, weight and fit was easy to endorse, whereas the look and comfort of the orthosis were difficult to agree with. No local dependency was detected. The targeting of item difficulty to the patient ability was good. Omega reliability value of CSD-Sk was 0.9 (polychoric Cronbach's alpha level 0.9). No DIF was detected. ConClusions: Despite some limitations in terms of fit, psychometric properties of CSD-Sk are in line with previous analyses on the English, Swedish and Italian version of the tool.