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.

P-value based analysis for shared controls design in genome-wide association studies.

An appealing genome-wide association study design compares one large control group against several disease samples. A pioneering study by the Wellcome Trust Case Control Consortium that employed such a design has identified multiple susceptibility regions, many of which have been independently replicated. While reusing a control sample provides effective utilization of data, it also creates correlation between association statistics across diseases. An observation of a large association statistic for one of the diseases may greatly increase chances of observing a spuriously large association for a different disease. Accounting for the correlation is also particularly important when screening for SNPs that might be involved in a set of diseases with overlapping etiology. We describe methods that correct association statistics for dependency due to shared controls, and we describe ways to obtain a measure of overall evidence and to combine association signals across multiple diseases. The methods we describe require no access to individual subject data, instead, they efficiently utilize information contained in P-values for association reported for individual diseases. P-value based combined tests for association are flexible and essentially as powerful as the approach based on aggregating the individual subject data.