{"id":"W4411181503","doi":"10.1177/20552076251348850","title":"Answering real-world clinical questions using large language model, retrieval-augmented generation, and agentic systems","year":2025,"lang":"en","type":"article","venue":"Digital Health","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":33,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute for Clinical Evaluative Sciences; SickKids Foundation; Hospital for Sick Children","funders":"","keywords":"Question answering; Computer science; Natural language processing; Psychology; Information retrieval","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005998549,0.0001042544,0.0002713179,0.0001807072,0.0002485428,0.0001002374,0.00003707326,0.00007297937,0.000005353643],"category_scores_gemma":[0.0002539326,0.0001013537,0.00004438658,0.0003743778,0.00004491849,0.0001889431,0.00002749529,0.0001822074,0.000008765905],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002692563,"about_ca_system_score_gemma":0.0008235204,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001669102,"about_ca_topic_score_gemma":0.0004269325,"domain_scores_codex":[0.9984993,0.00006405666,0.0007269716,0.0002752029,0.0001423316,0.0002921477],"domain_scores_gemma":[0.9991765,0.00009215629,0.0001160915,0.000237998,0.0001500075,0.0002272353],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006733937,0.00256606,0.7302705,0.006137815,0.0003149296,0.00006369811,0.011732,0.004096611,0.001510664,0.04716785,0.01958075,0.1758857],"study_design_scores_gemma":[0.0002767895,0.0003020194,0.0112824,0.001344019,0.00006601895,0.00002217585,0.003893428,0.9794338,0.0003437307,0.0004730522,0.002332153,0.0002303561],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9719713,0.001676074,0.02178728,0.001784915,0.0008533279,0.0006116782,0.00003861394,0.00009445187,0.001182329],"genre_scores_gemma":[0.9953206,0.0004991076,0.0003353633,0.001028326,0.0003460456,0.000007246268,0.0001512484,0.00001393848,0.002298093],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9753373,"threshold_uncertainty_score":0.4133082,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2412290545891537,"score_gpt":0.5220397402381639,"score_spread":0.2808106856490102,"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."}}