Psychometrics of the global assessment of functioning-modified scale GAF-M: data from a clinical Finnish adolescent sample

Background Assessing functioning is crucial for identifying psychosis risk and treatment needs. The Global Assessment of Functioning (GAF) scale, widely used in psychiatry, has demonstrated poor inter-rater reliability (IRR) in routine clinical practice. The modified GAF (GAF-M) has demonstrated higher IRR, but no studies have been conducted with a clinical adolescent sample. This study examines the IRR and concurrent validity of GAF-M in an adolescent sample, representing a novel contribution to the existing literature.Methods We assessed 218 adolescents aged 12–22 years using the GAF-M, with two trained clinicians rating independently. We calculated the intraclass correlation (ICC) and absolute scoring difference to evaluate IRR and performed bivariate correlations with symptom measures, demographics and GAF-M scores.Results The GAF-M exhibited a good IRR (ICC = 0.88 [95% CI 0.84–0.90]). The mean absolute difference between raters was 2.80 (SD 3.21). Higher GAF-M scores were associated with younger age (−.240, p < .001), fewer depressive (−.166, p = .016) and psychosis risk symptoms (−.168, p = .015) but not with gender or anxiety. Rater scoring differences were unrelated to demographic factors or symptom scores.Discussion The GAF-M demonstrates good IRR, and the majority of results support concurrent validity, affirming its applicability in multiprofessional and early intervention teams.