{"id":"W4404368553","doi":"10.1016/j.jeph.2024.202791","title":"Extracting social determinants of health from inpatient electronic medical records using natural language processing","year":2024,"lang":"en","type":"article","venue":"Journal of Epidemiology and Population Health","topic":"Food Security and Health in Diverse Populations","field":"Health Professions","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"South Health Campus; University of Calgary; Alberta Health Services","funders":"Canadian Institutes of Health Research","keywords":"Health records; Medical record; Natural language processing; Electronic health record; Natural (archaeology); Data science; Psychology; Computer science; Medicine; Geography; Political science; Health care","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01189162,0.0001404218,0.0009484651,0.0002756453,0.001118218,0.000004734952,0.0001039803,0.0003251237,0.0001385789],"category_scores_gemma":[0.002350838,0.0001169033,0.0001039833,0.0002220298,0.00007233231,0.0002802677,0.00004210225,0.001749482,0.000002264349],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005039981,"about_ca_system_score_gemma":0.00273922,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008803968,"about_ca_topic_score_gemma":0.002531212,"domain_scores_codex":[0.9922341,0.003555267,0.003069806,0.0002129131,0.0002690777,0.0006588659],"domain_scores_gemma":[0.9945152,0.002248197,0.002770234,0.00008231325,0.0001362855,0.0002477642],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001377678,0.00005007566,0.7275871,0.001757261,0.00003563078,0.00000595314,0.02015723,0.00003253544,0.000003257249,0.002246255,0.0009769943,0.2470099],"study_design_scores_gemma":[0.0008867818,0.0005940979,0.8973256,0.005059087,0.00004937414,0.00008014453,0.006596375,0.07588247,6.892885e-7,0.009242962,0.004084522,0.0001979199],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9343094,0.050499,0.001267331,0.01201141,0.001586196,0.0002647098,0.0000246597,0.00002426273,0.00001298707],"genre_scores_gemma":[0.9918978,0.001334363,0.00224624,0.003504202,0.0009377795,0.000002836653,0.00004763934,0.0000158534,0.00001325369],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.246812,"threshold_uncertainty_score":0.9977965,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2551996765807935,"score_gpt":0.5727994810308287,"score_spread":0.3175998044500352,"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."}}