{"id":"W4408685817","doi":"10.1136/bmjopen-2024-088735","title":"GPT for RCTs? Using AI to determine adherence to clinical trial reporting guidelines","year":2025,"lang":"en","type":"article","venue":"BMJ Open","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"Ottawa Hospital; University of British Columbia","funders":"Canadian Institutes of Health Research","keywords":"Medicine; Guideline; Test (biology); Randomized controlled trial; MEDLINE; Clinical trial; Family medicine; Medical physics; Artificial intelligence; Medical education; Internal medicine; Computer science; 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"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.7834989,0.002061172,0.005533136,0.01107815,0.003328174,0.01541516,0.006872159,0.00744147,0.005919969],"category_scores_gemma":[0.9370356,0.002353946,0.01335656,0.01242021,0.005460015,0.01551069,0.008202367,0.01059298,0.00177609],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.008215251,"about_ca_system_score_gemma":0.02080292,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002792461,"about_ca_topic_score_gemma":0.003567004,"domain_scores_codex":[0.100714,0.6308889,0.1980033,0.01751764,0.0511947,0.001681374],"domain_scores_gemma":[0.02599473,0.7732773,0.09735408,0.06864333,0.03307037,0.00166029],"domain_codex":"methods","domain_gemma":"reporting","domain_candidate":"reporting","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.006327037,0.0007159453,0.1245802,0.07299921,0.03492545,0.0006726434,0.02034748,0.008162161,0.001914482,0.03266611,0.1050358,0.5916534],"study_design_scores_gemma":[0.01393178,0.009679444,0.1448147,0.1276984,0.02780379,0.002167998,0.008517572,0.1378889,0.01177844,0.2116058,0.3021573,0.001955877],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09187994,0.07328986,0.4669159,0.1911023,0.0138334,0.09953724,0.01935331,0.006222022,0.03786613],"genre_scores_gemma":[0.4485736,0.004690332,0.3943458,0.03504786,0.002462123,0.1083509,0.00445767,0.0008581824,0.001213523],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.2165011,"threshold_uncertainty_score":0.2669844,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.9112409160646195,"score_gpt":0.750494443740196,"score_spread":0.1607464723244235,"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."}}