{"id":"W3014193862","doi":"10.23889/ijpds.v5i1.1156","title":"The effect of number of healthcare visits on study sample selection in electronic health record data","year":2020,"lang":"en","type":"article","venue":"International Journal for Population Data Science","topic":"Medical Coding and Health Information","field":"Health Professions","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kellogg's (Canada)","funders":"National Center for Advancing Translational Sciences; National Center for Chronic Disease Prevention and Health Promotion","keywords":"Medicine; Sample (material); Health care; Electronic health record; Sample size determination; Family medicine; Gerontology; Statistics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.555949,0.001021598,0.001758269,0.002525769,0.003334374,0.004620255,0.004877896,0.003595515,0.005125287],"category_scores_gemma":[0.7814553,0.001284509,0.005331922,0.004297786,0.006229148,0.007832476,0.005923924,0.003497991,0.0009125424],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003799816,"about_ca_system_score_gemma":0.005121114,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003442511,"about_ca_topic_score_gemma":0.004433843,"domain_scores_codex":[0.1701004,0.712057,0.05302436,0.01923143,0.04326659,0.00232019],"domain_scores_gemma":[0.07559468,0.840637,0.04059359,0.03091063,0.01081496,0.001449185],"domain_codex":"methods","domain_gemma":"methods","domain_candidate":"methods","domain_consensus":"methods","study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.005813401,0.0006269926,0.8101327,0.005361896,0.005804047,0.0003729006,0.007318988,0.002380316,0.0008481946,0.008342355,0.01498187,0.1380163],"study_design_scores_gemma":[0.002327428,0.006331677,0.862141,0.01383207,0.005301076,0.001912137,0.003599318,0.02409801,0.005533521,0.02340875,0.05110598,0.0004091099],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5531654,0.03439476,0.2867385,0.05133109,0.005187228,0.02360999,0.007922602,0.001168334,0.03648212],"genre_scores_gemma":[0.8792655,0.001554415,0.08635162,0.01392651,0.001137724,0.01499874,0.001362272,0.0003779858,0.001025236],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.444051,"threshold_uncertainty_score":0.547594,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.338917648568192,"score_gpt":0.5860153991181046,"score_spread":0.2470977505499126,"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."}}