{"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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","sts","research_integrity","insufficient_payload"],"consensus_categories":["metaresearch","research_integrity"],"category_scores_codex":[0.04742986,0.0002093609,0.00074473,0.0004489013,0.005328974,0.0001113929,0.002495971,0.002546619,0.005093185],"category_scores_gemma":[0.2676233,0.0001512764,0.00007978319,0.0004959804,0.0001537939,0.0002031479,0.0002869499,0.01929039,0.0006306818],"about_ca_system_candidate":true,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003553964,"about_ca_system_score_gemma":0.03171027,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.0456203,"about_ca_topic_score_gemma":0.1650927,"domain_scores_codex":[0.9856009,0.004545402,0.002412705,0.0003343099,0.005434022,0.001672681],"domain_scores_gemma":[0.971505,0.01963673,0.002570207,0.001137868,0.003064201,0.002085989],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000007577522,0.000004623259,0.0200462,0.0002604973,0.00005446602,0.0001731298,0.0007993761,2.892623e-7,6.953311e-8,0.0002634897,0.9765795,0.001810817],"study_design_scores_gemma":[0.000531078,0.00003953825,0.04904293,0.001352215,0.00002511234,0.00001611528,0.0004319151,0.00005285637,1.372557e-7,0.0003854397,0.9479811,0.0001416193],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"commentary","genre_gemma":"commentary","genre_scores_codex":[0.003054403,0.0001058188,0.00009930649,0.9807458,0.004265254,0.0008943657,0.0009923308,0.00001892384,0.009823815],"genre_scores_gemma":[0.01664502,0.0004519035,0.00008054725,0.9471613,0.02771751,0.00009168253,0.0008811175,0.00004020565,0.006930737],"genre_candidate":"commentary","genre_consensus":"commentary","teacher_disagreement_score":0.2201934,"threshold_uncertainty_score":0.9987483,"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."}}