{"id":"W4398201968","doi":"10.1001/jamanetworkopen.2024.12687","title":"Assessing the Risk of Bias in Randomized Clinical Trials With Large Language Models","year":2024,"lang":"en","type":"article","venue":"JAMA Network Open","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":89,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University; Impact","funders":"","keywords":"Randomized controlled trial; Medicine; Systematic review; Reliability (semiconductor); Risk assessment; MEDLINE; Physical therapy; Internal medicine; Computer science","routes":{"ca_aff":true,"ca_fund":false,"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"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.7664369,0.005196115,0.01395449,0.01805233,0.002426321,0.01324982,0.006559938,0.008070878,0.008340577],"category_scores_gemma":[0.9329365,0.005178143,0.0210056,0.01328627,0.01012197,0.01531336,0.01006258,0.007557831,0.001736264],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00892562,"about_ca_system_score_gemma":0.02209356,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00237343,"about_ca_topic_score_gemma":0.004171249,"domain_scores_codex":[0.08244331,0.8007333,0.07603915,0.01152517,0.02865744,0.0006016649],"domain_scores_gemma":[0.01653009,0.9172918,0.03727919,0.02131979,0.007138494,0.0004405442],"domain_codex":"methods","domain_gemma":"methods","domain_candidate":"methods","domain_consensus":"methods","study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.01048824,0.0007127515,0.07493134,0.2299765,0.1003876,0.001526125,0.01004592,0.05167068,0.001730626,0.0885089,0.02818492,0.4018363],"study_design_scores_gemma":[0.0129271,0.005704092,0.02173889,0.1114738,0.05113326,0.002994254,0.001615683,0.3678545,0.00366649,0.3710735,0.04776387,0.002054577],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01275871,0.03227758,0.8973076,0.01801072,0.002829676,0.02815103,0.001988251,0.002487537,0.004188914],"genre_scores_gemma":[0.1852338,0.004086598,0.7425434,0.003523492,0.0009739686,0.06216904,0.000658085,0.0002537243,0.0005579148],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.2335631,"threshold_uncertainty_score":0.288025,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4904734455333462,"score_gpt":0.5865774049141176,"score_spread":0.09610395938077138,"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."}}