{"id":"W4379197097","doi":"10.1177/14604582231180581","title":"Breaking the 80:20 rule in health research using large administrative data sets","year":2023,"lang":"en","type":"article","venue":"Health Informatics Journal","topic":"Artificial Intelligence in Healthcare","field":"Health Professions","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Alberta Bone and Joint Health Institute; University of Calgary","funders":"Canadian Institutes of Health Research; Canada Foundation for Innovation","keywords":"Online analytical processing; Data warehouse; Computer science; Data cube; Data mining; Leverage (statistics); Data science; Analytics; Health care; Population; Database; Medicine","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.2875256,0.001217948,0.003493279,0.007373235,0.004386216,0.02038199,0.005187808,0.003816865,0.00208075],"category_scores_gemma":[0.4852983,0.00175954,0.00326853,0.01864529,0.01032689,0.01646243,0.01186556,0.00925553,0.002109295],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.005331894,"about_ca_system_score_gemma":0.01373036,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009246422,"about_ca_topic_score_gemma":0.006356637,"domain_scores_codex":[0.7213697,0.1670274,0.03135216,0.02564011,0.05175264,0.002857925],"domain_scores_gemma":[0.3693193,0.5256559,0.03299947,0.04694708,0.0210961,0.003982153],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.001529397,0.000889397,0.2119352,0.003564805,0.001508566,0.002344303,0.007535386,0.03189775,0.002113637,0.1798454,0.04118149,0.5156547],"study_design_scores_gemma":[0.0004424779,0.0005465749,0.0429194,0.002775582,0.0004325075,0.001626524,0.004349233,0.08899788,0.003723734,0.7707329,0.08318154,0.0002716619],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1825572,0.0113941,0.6743054,0.08156613,0.003324884,0.002664454,0.005220222,0.003009029,0.03595864],"genre_scores_gemma":[0.3525179,0.003525994,0.6198716,0.01381084,0.00224548,0.002338709,0.002998343,0.0005291901,0.002162004],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.2875256,"threshold_uncertainty_score":0.8786078,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.8455603265902553,"score_gpt":0.6929477354375245,"score_spread":0.1526125911527307,"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."}}