Milestones in Surgical Complication Reporting. Clavien-Dindo Classification 20 Years & Comprehensive Complication Index (CCI®) 10 Years.

OBJECTIVE: To provide improved guidance for the consistent application of the Clavien-Dindo classification (CDC) and Comprehensive Complication Index (CCI®) in challenging clinical scenarios. BACKGROUND: Standardized outcome reporting is key for proper assessment of surgical procedures. A recent consensus conference recommended the CDC and the CCI® for assessing postoperative morbidity. Several challenging scenarios for grading complications still require evidence-based guidance, and the use of the two metrics in RCTs remains unexplored. METHODS: We assessed the use of the CDC and CCI® as an outcome measure in a systematic literature search. Additionally, we asked 163 international surgeons to critically evaluate and independently grade complications in 20 complex clinical scenarios. Finally, a core group of five experts used this information to develop consistent recommendations. RESULTS: Until July 2023, 1327 RCTs selected the CDC and/or CCI® to assess morbidity. Annual use was steadily increasing with now over 200 new RCTs per year. However, only a third (n=335) of published RCTs provided the complete range of CDC grades, including all subgrades. Eighty-nine out of 163 surgeons (response rate 55%) completed the questionnaire that served as basis for the recommendations: Repetitive interventions that are required to treat one complication, complications followed by further complications, complications occurring prior to referral, and expected and unrelated complications to the original procedure should all be counted separately and included in the CCI®. Invasive blank diagnostic interventions should not be considered a complication. CONCLUSION: The increasing use of the CDC and CCI® in RCTs highlights the importance of their standardized application. The current consensus on various difficult scenarios may offer novel guidance for the consistent use of the CDC and CCI®, aiming to improve complication reporting, and better-quality control, ultimately benefiting all healthcare stakeholders, and first and foremost, all patients.

Validation of the Clavien-Madadi Classification for Unexpected Events in Pediatric Surgery: A Collaborative ERNICA Project.

BACKGROUND: The Clavien-Madadi classification is a novel instrument for the assessment and grading of unexpected events in pediatric surgery, based on the Clavien-Dindo classification. The system has been adjusted to better fit the pediatric population in a prospective single-center study. There is a need now to validate the Clavien-Madadi classification within an international expert network. METHODS: A pediatric surgical working group created 19 case scenarios with unexpected events in a multi-staged process. Those were circulated within the European Reference Network of Inherited and Congenital Anomalies (ERNICA) and surgeons were instructed to rate the scenarios according to the Clavien-Madadi vs. Clavien-Dindo classification. RESULTS: 59 surgeons from 12 European countries completed the questionnaire. Based on ratings of the case scenarios, the Clavien-Madadi classification showed significantly superior agreement rates of the respondents (85.9% vs 76.2%; p < 0.05) and was less frequently considered inaccurate for rating the pediatric population compared to Clavien-Dindo (2.1% vs 11.1%; p = 0.05). Fleiss' kappa analysis showed slightly higher strength of agreement using the Clavien-Madadi classification (0.74 vs 0.69). Additionally, intraclass correlation coefficient was slightly higher for the Clavien-Madadi compared to the Clavien-Dindo classification (ICCjust 0.93 vs 0.89; ICCunjust 0.93 vs 0.89). More pediatric surgeons preferred the Clavien-Madadi classification for the case scenarios (43.0% vs 11.8%; p = 0.002) and advantages of the Clavien-Madadi were confirmed by 81.4% of the surgeons. CONCLUSION: The Clavien-Madadi classification is an accurate and reliable instrument for the grading of unexpected events in pediatric surgery. We therefore recommend its application in clinical and academic pediatric surgical practice. LEVEL OF EVIDENCE: III.

Experts vs. machine - comparison of machine learning to expert-informed prediction of outcome after major liver surgery.

BACKGROUND: Machine learning (ML) has been successfully implemented for classification tasks (e.g., cancer diagnosis). ML performance for more challenging predictions is largely unexplored. This study's objective was to compare machine learning vs. expert-informed predictions for surgical outcome in patients undergoing major liver surgery. METHODS: Single tertiary center data on preoperative parameters and postoperative complications for elective hepatic surgery patients were included (2008-2021). Expert-informed prediction models were established on 14 parameters identified by two expert liver surgeons to impact on postoperative outcome. ML models used all available preoperative patient variables (n = 62). Model performance was compared for predicting 3-month postoperative overall morbidity. Temporal validation and additional analysis in major liver resection patients were conducted. RESULTS: 889 patients included. Expert-informed models showed low average bias (2-5 CCI points) with high over/underprediction. ML models performed similarly: average prediction 5-10 points higher than observed CCI values with high variability (95% CI -30 to 50). No performance improvement for major liver surgery patients. CONCLUSION: No clinical relevance in the application of ML for predicting postoperative overall morbidity was found. Despite being a novel hype, ML has the potential for application in clinical practice. However, at this stage it does not replace established approaches of prediction modelling.

Surgical Outcome Reporting. Moving From a Comic to a Tragic Opera?

OBJECTIVES: To assess the current quality of surgical outcome reporting in the medical literature and to provide recommendations for improvement. BACKGROUND: In 1996, The Lancet labeled surgery as a "comic opera" mostly referring to the poor quality of outcome reporting in the literature impeding improvement in surgical quality and patient care. METHODS: We screened 3 first-tier and 2 second-tier surgical journals, as well as 3 leading medical journals for original articles reporting on results of surgical procedures published over a recent 18-month period. The quality of outcome reporting was assessed using a prespecified 12-item checklist. RESULTS: Six hundred twenty-seven articles reporting surgical outcomes were analyzed, including 125 randomized controlled trials. Only 1 (0.2%) article met all 12 criteria of the checklist, whereas 356 articles (57%) fulfilled less than half of the criteria. The poorest reporting was on cumulative morbidity burden, which was missing in 94% of articles (n=591) as well as patient-reported outcomes missing in 83% of publications (n=518). Comparing journal groups for the individual criterion, we found moderate to very strong statistical evidence for better quality of reporting in high versus lower impact journals for 7 of 12 criteria and strong statistical evidence for better reporting of patient-reported outcomes in medical versus surgical journals ( P <0·001). CONCLUSIONS: The quality of outcomes reporting in the medical literature remains poor, lacking improvement over the past 20 years on most key end points. The implementation of standardized outcome reporting is urgently needed to minimize biased interpretation of data thereby enabling improved patient care and the elaboration of meaningful guidelines.