AbstractBackgroundInconsistent definitions of complications and unexpected events have limited accurate analysis of surgical outcomes. Perioperative outcome classifications currently used for adult patients have limitations when used for children.MethodsA multidisciplinary group of experts modified the Clavien–Dindo classification to increase its utility and accuracy in paediatric surgery cohorts. Organizational and management errors were considered in the novel Clavien–Madadi classification, which focuses on procedural invasiveness rather than anaesthetic management. Unexpected events were prospectively documented in a paediatric surgery cohort. Results of the Clavien–Dindo and Clavien–Madadi classifications were compared and correlated with procedure complexity.ResultsUnexpected events were prospectively documented in a cohort of 17 502 children undergoing surgery between 2017 and 2021. The results of both classifications were highly correlated (ρ = 0.95), although the novel Clavien–Madadi classification identified 449 additional events (organizational and management errors) over the Clavien–Dindo classification, increasing the total number of events by 38 per cent (1605 versus 1158 events). The results of the novel system correlated significantly with the complexity of procedures in children (ρ = 0.756). Furthermore, grading of events > grade III according to the Clavien–Madadi classification showed a higher correlation with procedure complexity (ρ = 0.658) than the Clavien–Dindo classification (ρ = 0.198).ConclusionThe Clavien–Madadi classification is a tool for the detection of surgical and non-medical errors in paediatric surgery populations. Further validation in paediatric surgery populations is required before widespread use.