Quality assurance in anaesthetic practice: comparison between two methods in detecting complications.
BACKGROUND: Reliable identification and documentation of complications is an essential part of a well-functioning quality system (QS) in anaesthetic practice. The criteria for the complications have to be appropriate. The QS of Kuopio University Hospital fulfils the ISO 9002 standard. The validity of the recordings in the QS was ascertained by comparing the routine recordings with external assessment. METHODS: Three types of complications were predefined: minor, severe and those specific for regional anaesthesia. A total of 1006 anaesthetic charts, including general, regional and intravenous anaesthesia, were randomly selected and retrospectively screened by an external assessor. The retrospective assessment of complications was compared to the recordings in the data management system for operative procedures (DMS) as a part of routine quality assurance. Cohen's kappa statistics was used to indicate agreement between two raters. RESULTS: Both methods identified complications in 115 procedures (11.4%). The methods, however, did not identify complications in same procedures. There was a fairly close agreement (P < 0.001) between the methods in detecting all (Cohen's kappa 0.72), minor (0.67) and severe (0.66) complications and those specific for regional anaesthesia (0.78). Fifty-eight complications were detected either by retrospective assessment or routine reporting, i.e. the two raters disagreed in 58 complications. Thirteen severe complications recorded in the DMS could not be retrospectively identified. The agreement did not depend on ASA class, the urgency or the length of procedures or on the type of anaesthesia. CONCLUSIONS: The agreement between external assessment and routine reporting was fairly good, however, there were still some relatively large differences. The selection and definitions of complications need to be reassessed.