{"id":"W4404727371","doi":"10.1177/17407745241297947","title":"Detecting irregularities in randomized controlled trials using machine learning","year":2024,"lang":"en","type":"article","venue":"Clinical Trials","topic":"Statistical Methods in Clinical Trials","field":"Mathematics","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Impact; Hamilton Health Sciences; McMaster University; Population Health Research Institute; University of Toronto","funders":"Canadian Institutes of Health Research","keywords":"Interquartile range; Artificial intelligence; Machine learning; Concordance; Outlier; Clinical trial; Computer science; Randomized controlled trial; Receiver operating characteristic; Medicine; Anomaly detection; Algorithm; Statistics; Data mining; Mathematics; Surgery; Pathology","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch","research_integrity"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.3023281,0.001241831,0.003262639,0.008484594,0.001148857,0.004874317,0.003307329,0.002516828,0.001077826],"category_scores_gemma":[0.5978215,0.001176936,0.003472878,0.007290061,0.004493872,0.003282083,0.002657214,0.002465968,0.0002561803],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003045012,"about_ca_system_score_gemma":0.004539804,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001384901,"about_ca_topic_score_gemma":0.001559805,"domain_scores_codex":[0.6338654,0.2452959,0.05922977,0.03314326,0.02695873,0.001506889],"domain_scores_gemma":[0.1653284,0.6821479,0.09568528,0.04255569,0.01305952,0.001223293],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.005567985,0.0005904189,0.601899,0.007764247,0.0122327,0.001648679,0.002423845,0.09060585,0.003100161,0.01324222,0.005881494,0.2550433],"study_design_scores_gemma":[0.002058838,0.005131219,0.2351639,0.003827725,0.005402154,0.003012268,0.001169424,0.5848234,0.01087388,0.1243526,0.0237379,0.0004467081],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2662272,0.01308374,0.7048511,0.003021067,0.0006815823,0.005990024,0.002271175,0.001977269,0.001896816],"genre_scores_gemma":[0.7530363,0.0007980235,0.2397229,0.0009928315,0.0003476645,0.003293604,0.001494275,0.00009831619,0.0002160527],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9974832,"threshold_uncertainty_score":0.8603537,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.8686945249905104,"score_gpt":0.6957301326060387,"score_spread":0.1729643923844717,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}