{"id":"W4416789332","doi":"10.1186/s12873-025-01402-w","title":"Accuracy is not enough: explainable boosting machine model and identification of candidate biomarkers for real-time sepsis risk assessment in the emergency department","year":2025,"lang":"en","type":"article","venue":"BMC Emergency Medicine","topic":"Sepsis Diagnosis and Treatment","field":"Medicine","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Lakehead University","funders":"Princess Nourah Bint Abdulrahman University","keywords":"Emergency department; Boosting (machine learning); Workflow; Sepsis; Risk assessment; Identification (biology); Transparency (behavior); MEDLINE","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":[0.001584829,0.00029106,0.0005666523,0.0003014434,0.0002113574,0.000005061901,0.0001670616,0.00008296281,0.0004498956],"category_scores_gemma":[0.0008196476,0.0001926307,0.0001496945,0.0005824312,0.00006250175,0.0001054931,0.00005443862,0.0001248766,0.000002775181],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001462355,"about_ca_system_score_gemma":0.0001610239,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002481266,"about_ca_topic_score_gemma":0.0003735741,"domain_scores_codex":[0.9972146,0.0001411776,0.001342135,0.0005173065,0.0004344515,0.0003502934],"domain_scores_gemma":[0.9981097,0.0004205276,0.0005309936,0.0006155908,0.0002266612,0.00009649915],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0007629113,0.001707139,0.8682749,0.001568953,0.001036295,0.000007587915,0.003977446,0.0002977715,0.02338073,0.001193929,0.09033233,0.00746002],"study_design_scores_gemma":[0.007995577,0.000962711,0.8312703,0.0006710664,0.002965283,0.000005573292,0.002273097,0.1413559,0.006831896,0.002394708,0.00289115,0.0003826791],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9756373,0.005260637,0.002656603,0.009521277,0.0006009858,0.003092163,0.0002071833,0.00003941558,0.00298446],"genre_scores_gemma":[0.9720626,0.02357807,0.001558749,0.0001771336,0.00008593157,0.0009263224,0.0003118506,0.00002589485,0.001273388],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1410581,"threshold_uncertainty_score":0.7855252,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0727895011021151,"score_gpt":0.4022741645275164,"score_spread":0.3294846634254013,"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."}}