{"id":"W4283268556","doi":"10.1002/pds.5500","title":"Machine learning for improving high‐dimensional proxy confounder adjustment in healthcare database studies: An overview of the current literature","year":2022,"lang":"en","type":"review","venue":"Pharmacoepidemiology and Drug Safety","topic":"Advanced Causal Inference Techniques","field":"Mathematics","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Agency for Drugs and Technologies in Health; McGill University; McGill University Health Centre","funders":"National Institute on Aging; International Society for Pharmacoepidemiology","keywords":"Confounding; Proxy (statistics); Prioritization; Covariate; Medicine; Feature selection; Computer science; Data mining; Machine learning; Management science","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":["metaepi_narrow","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.006156906,0.0005835233,0.002649604,0.0001568311,0.0003035523,0.000005040853,0.0004048682,0.0001713628,0.00004301116],"category_scores_gemma":[0.003737605,0.0003698645,0.0003164687,0.0003043198,0.0001951134,0.0001692791,0.000768636,0.002381758,3.368995e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003484984,"about_ca_system_score_gemma":0.0003037805,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006589926,"about_ca_topic_score_gemma":0.00003722082,"domain_scores_codex":[0.9920011,0.004937094,0.001655764,0.0007474711,0.0001914128,0.0004672237],"domain_scores_gemma":[0.9872992,0.01049702,0.001468757,0.0004976313,0.000130479,0.0001068778],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00008605092,0.0001585642,0.0000611719,0.0798337,0.0001598368,0.000003541363,0.0002938266,0.00001574327,0.000001261927,0.09919962,0.0004652592,0.8197214],"study_design_scores_gemma":[0.0007795955,0.0001538369,0.00001181262,0.01469981,0.0008637186,0.00003905538,0.00004682001,0.001076064,0.000003491346,0.08569313,0.8960807,0.0005519064],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.000006692322,0.994091,0.0007238117,0.0004783899,0.0006376876,0.003059941,0.0009173722,0.00008000121,0.000005087362],"genre_scores_gemma":[0.00006880132,0.991473,0.005483456,0.0005237194,0.0001399297,0.001220015,0.0009959833,0.00005740976,0.00003769605],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.8956155,"threshold_uncertainty_score":0.9999198,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4355611869380462,"score_gpt":0.5536874237850973,"score_spread":0.118126236847051,"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."}}