Psychometric Properties of WHO Schedules for Clinical Assessment in Neuropsychiatry: A Systematic Review
Fekadu, W.; Mihretu, A.; Alem, A.; Brugha, T.; van Ommeren, M.; Chatterji, S.; Hanlon, C.; Fekadu, A.
Show abstract
BackgroundWHOs Schedules for Clinical Assessment in Neuropsychiatry (SCAN) is often used as the gold standard for psychiatric classification. We systematically reviewed studies on the psychometric properties of the SCAN to support its adaptation to the revised international classification systems. MethodsWe searched PubMed, PsycINFO, Embase, Global Health, and Global Index Medicus up to April 17, 2025, and contacted experts. The protocol was registered in PROSPERO (CRD42024522395). ResultsTitles and abstracts of 4,241 records were screened, with 296 full-text articles evaluated. Ninety-three articles were included in the final review: 46 assessing SCANs psychometric properties and 47 validating other measures using SCAN as a gold standard. The internal consistency of the SCAN and its predecessor, the Present State Examination (PSE), ranged from good to excellent. Both demonstrated acceptable intra-rater, inter-rater, and test-retest reliability, with reliability especially high for psychotic disorders. There was also evidence supporting concurrent, construct, semantic, and content validity, although there was an absence of evidence for predictive validity. We also found acceptable psychometric properties for the different syndrome-based sections of the SCAN. ConclusionAlthough recent, high-quality studies are scarce, the SCAN is a promising tool for diagnosing a variety of psychiatric issues, particularly psychotic disorders. It demonstrates established reliability and evidence of concurrent, construct, semantic, and content validity. However, there is a need to revise the current version of SCAN to align it with contemporary diagnostic systems. Additionally, further research is required, especially regarding the assessment of non-psychotic conditions.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.