{"id":"W2748018347","doi":"10.1503/cmaj.170807","title":"Routinely collected data: the importance of high-quality diagnostic coding to research","year":2017,"lang":"en","type":"letter","venue":"Canadian Medical Association Journal","topic":"Medical Coding and Health Information","field":"Health Professions","cited_by":54,"is_retracted":false,"has_abstract":true,"ca_institutions":"Ottawa Public Health; Children's Hospital of Eastern Ontario; University of Ottawa","funders":"National Institute for Health and Care Research; Wellcome Trust","keywords":"Data science; Computer science; Coding (social sciences); Data collection; Key (lock); Data quality; Health records; Information retrieval; Data mining; Health care; Statistics","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"about_ca":false,"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":[0.1792409,0.0005510203,0.001311963,0.003895527,0.00454808,0.008591507,0.004426322,0.02049001,0.00945932],"category_scores_gemma":[0.5777289,0.0009975919,0.0008532124,0.00323517,0.01606829,0.01639291,0.007631954,0.03325275,0.004649449],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.01059421,"about_ca_system_score_gemma":0.02782119,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01647462,"about_ca_topic_score_gemma":0.02268915,"domain_scores_codex":[0.7495754,0.1475467,0.02557638,0.008385587,0.06526199,0.003654063],"domain_scores_gemma":[0.1984132,0.6627663,0.0276146,0.02495369,0.07283358,0.0134187],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000101791,0.00004778112,0.00647096,0.0009410367,0.00002697765,0.0007423837,0.002590472,0.0001149279,0.0002646674,0.01989557,0.8642998,0.1045036],"study_design_scores_gemma":[0.0001837889,0.0002504881,0.01833859,0.01916034,0.00007165929,0.006820359,0.00851889,0.002050496,0.0007707661,0.1235366,0.8199053,0.0003927216],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"commentary","genre_gemma":"commentary","genre_scores_codex":[0.0005528399,0.002100413,0.00292596,0.9872542,0.004479949,0.00005994432,0.0001157929,0.00005797024,0.002452905],"genre_scores_gemma":[0.03714493,0.008609487,0.02627661,0.886978,0.03778794,0.0006453317,0.0005208903,0.0002442868,0.001792533],"genre_candidate":"commentary","genre_consensus":"commentary","teacher_disagreement_score":0.8207591,"threshold_uncertainty_score":0.947928,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4180962378961449,"score_gpt":0.5302135301044664,"score_spread":0.1121172922083215,"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."}}