Document resource
Background The existent classifications of the distinct evidence-based psychological interventions (EBPs) are broad, not systematically developed or restricted to one mental disorder.Objective To identify the unique EBPs and their variations evaluated for six severe mental disorders (schizophrenia and psychosis, bipolar disorders, bulimia nervosa, anorexia nervosa, one type of substance use and borderline personality).Study selection and analysis We established a large cohort of EBPs by selecting all randomised controlled trials (RCTs) from six recent and large network meta-analyses for severe mental disorders in adults. Pairs of researchers independently selected all psychological intervention arms and assigned them to macro-families based on labels and theoretical background. Subsequently, within each macro-family, groups of identical arms, that is, referencing the same manual or protocol, were consolidated as distinct psychological interventions. Adaptations related to content, structure, delivery or culture were also described.Findings From 260 RCTs, 422 psychological intervention arms were grouped in 45 macro-families and reduced to 266 distinct interventions (63% of total arms) across disorders. Intervention variability, that is, the proportion of unique psychological interventions out of the total number of arms, was lowest for the substance use disorder selected—stimulant use disorder (47/104, 45%) and bulimia nervosa (25/45, 55.5%). Variability was high for bipolar disorders (48/59, 81%), anorexia nervosa (20/29, 69%), schizophrenia/psychosis (86/123, 70%) and borderline personality disorder (40/62, 64.5%). 79 of the 266 distinct interventions (29%) contained adaptations (12% multiple adaptations) and 75/266 (28%) had openly available manuals or protocols.Conclusions and clinical implications Mapping EBPs across macro-families and distinct interventions could inform evidence synthesis and identification of active ingredients, guide mechanisms of change exploration and support prioritising interventions for research and dissemination.