{"id":"W4392963405","doi":"10.1016/j.ajem.2024.03.019","title":"Methodological issues on precision and prediction value of ChatGPT in emergency department triage decisions","year":2024,"lang":"en","type":"letter","venue":"The American Journal of Emergency Medicine","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"McMaster University","funders":"","keywords":"Triage; Emergency department; Value (mathematics); Medical emergency; Medicine; Computer science; Machine learning; Nursing","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":[0.5414782,0.0009729221,0.003998717,0.004186623,0.003577075,0.01100218,0.008845749,0.01238457,0.004651973],"category_scores_gemma":[0.8731983,0.001763495,0.003991829,0.004669979,0.01094685,0.007213509,0.004333995,0.01432126,0.001832754],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.005484692,"about_ca_system_score_gemma":0.006509115,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01323349,"about_ca_topic_score_gemma":0.0131096,"domain_scores_codex":[0.3988937,0.4792415,0.05591706,0.01863332,0.0435975,0.003717021],"domain_scores_gemma":[0.03927057,0.9075378,0.01140399,0.01739736,0.0231769,0.00121344],"domain_codex":"methods","domain_gemma":"methods","domain_candidate":"methods","domain_consensus":"methods","study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.01015858,0.0006498887,0.3824482,0.00336497,0.01182289,0.002169957,0.009419975,0.01205691,0.0008258134,0.08380226,0.180889,0.3023916],"study_design_scores_gemma":[0.006191649,0.002361444,0.2044762,0.01136577,0.008440251,0.007605163,0.005481766,0.1435233,0.00645588,0.4893772,0.113655,0.001066371],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"commentary","genre_gemma":"commentary","genre_scores_codex":[0.09392627,0.01505734,0.09835432,0.7200618,0.02907965,0.00223741,0.004070557,0.0004366895,0.03677613],"genre_scores_gemma":[0.6828946,0.001860871,0.080514,0.1939844,0.03314885,0.003395863,0.000760572,0.0002314582,0.003209291],"genre_candidate":"commentary","genre_consensus":"commentary","teacher_disagreement_score":0.4585218,"threshold_uncertainty_score":0.565439,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4042195558886605,"score_gpt":0.5310878016806659,"score_spread":0.1268682457920055,"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."}}