P. Krielen1, L. Gawria1,2, M. Stommel1, S. Dell-Kuster2,3, R. Rosenthal4, R. Ten Broek1, H. Van Goor1 1Radboud University Medical Center,Departement Of Surgery,Nijmegen, GELDERLAND, Netherlands 2University of Basel,Basel Institute Of Clinical Epidemiology And Biostatistics,Basel, Switzerland 3University Hospital Basel,Departement Of Anaesthesiology,Basel, Switzerland 4University of Basel,Basel, Switzerland
Introduction:
Surgical outcomes depend on the quality of preoperative, intraoperative and postoperative care. Prospectively validated classification systems for evaluation of intraoperative adverse events (iAEs) are not available. Recently, a classification of intraoperative complications (CLASSIC) to grade iAEs has been developed, retrospectively validated (Rosenthal 2015), and extended to 5 grades of severity. The aim of this retrospective analysis of prospectively collected data is to assess the inter-rater agreement of CLASSIC and its predictive value for severe postoperative complications using data from a cohort study of elective abdominal surgeries (LAPAD study (ten Broek 2013), NCT01236625).
Methods:
In the LAPAD study, detailed data on iAEs and their treatment were collected by an independent researcher present in the operating room. For the current research questions, two independent teams, each consisting of a dedicated researcher and a surgeon, retrospectively graded all previously recorded iAEs according to CLASSIC. The two teams were blinded for each other's grading. Discrepancies between the teams in the grading of iAEs were resolved through discussion. Cohen’s Kappa coefficient was calculated to determine the inter-rater agreement. Uni- and multivariable logistic regression were used to estimate the predictive value of the highest CLASSIC grade (categorized as 0, I-II, and III-V) on the occurrence of the most severe post-operative complication according to Clavien-Dindo (dichotomized as grade III-V versus 0-II). Multivariable analysis was adjusted for all variables, which were significant in univariate analysis.
Results:
Seven hundred fifty-five abdominal surgeries were reviewed for the occurrence of iAEs. iAEs were observed in 333 surgeries (44.1%) by team 1, and 324 (42.9%) by team 2. The raw agreement between both teams was 86.9% for the classifications of iAEs according to CLASSIC. The Cohen’s Kappa coefficient across all 5 CLASSIC grades was 0.87 (95% CI 0.84-0.90). A severe iAE (CLASSIC III – V) was observed in 86 (11.4%) of surgeries. Any post-operative complication was observed in 278 (36.8%) surgeries, of which 129 (46.4%) severe postoperative complications (Clavien-Dindo grade 3 or higher). In 27 (31.4%) patients with a severe iAE, a severe post-operative complication was observed, compared to 52 (20.9%) of patients with a minor iAE, and 50 (11.9%) with no iAE (p<0.001). An iAE was a significant independent risk factor for severe post-operative complications in multivariable analysis, with OR 3.07 (95% CI 1.76 – 5.34) for severe iAEs and OR 1.97 (95% CI 1.28 – 3.04) for minor iAEs as compared to no iAEs.
Conclusion:
The newly proposed CLASSIC is a reliable tool for the classification of iAEs, with a good inter-rater agreement. A higher severity grade of iAEs according to CLASSIC was associated with a higher risk of severe post-operative complications.