{"id":"W3213880547","doi":"10.1016/j.jclinepi.2021.11.001","title":"Believing in black boxes: machine learning for healthcare does not need explainability to be evidence-based","year":2021,"lang":"en","type":"review","venue":"Journal of Clinical Epidemiology","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":113,"is_retracted":false,"has_abstract":false,"ca_institutions":"Schwartz/Reisman Emergency Medicine Institute; University of Toronto","funders":"Canadian Institutes of Health Research","keywords":"Health care; Evidence-based practice; Evidence-based medicine; Real world evidence; Computer science; Medicine; Alternative medicine; Economics; Pathology","routes":{"ca_aff":true,"ca_fund":true,"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":["metaresearch","metaepi_narrow","research_integrity"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.09056282,0.0006484656,0.01130788,0.0006776213,0.0001817555,0.00008959671,0.002927485,0.001175545,0.00003577991],"category_scores_gemma":[0.4412723,0.0004429453,0.003307767,0.001123876,0.0002993174,0.0005932054,0.0006586881,0.004023385,0.00001965418],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005421545,"about_ca_system_score_gemma":0.003359228,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003300598,"about_ca_topic_score_gemma":0.0004399377,"domain_scores_codex":[0.9584046,0.02444491,0.01398937,0.001425411,0.0004985276,0.001237163],"domain_scores_gemma":[0.7735104,0.214801,0.008022286,0.001323839,0.001413443,0.0009290304],"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":[0.0001502857,0.000230817,0.005341535,0.007190407,0.000107223,0.0001628347,0.0002312728,0.003348892,4.560598e-7,0.007823633,0.0004554889,0.9749572],"study_design_scores_gemma":[0.000397895,0.002831351,0.0002734164,0.03770582,0.0002039442,0.00007306077,0.0002127911,0.02319166,0.000009648922,0.005940775,0.9283947,0.0007649549],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.0001731705,0.7024844,0.2370581,0.05729645,0.001980911,0.0009605226,0.000008699441,0.00003206143,0.000005614806],"genre_scores_gemma":[0.0007400512,0.924507,0.06275554,0.01068739,0.001096078,0.00008373316,0.000008607641,0.00004598907,0.00007558095],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.9741922,"threshold_uncertainty_score":0.9998022,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6211548340976271,"score_gpt":0.5750288304756365,"score_spread":0.0461260036219906,"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."}}