{"id":"W4389615820","doi":"10.23889/ijpds.v8i1.2153","title":"De-identification of Free Text Data containing Personal Health Information: A Scoping Review of Reviews","year":2023,"lang":"en","type":"review","venue":"International Journal for Population Data Science","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Manitoba; George & Fay Yee Centre for Healthcare Innovation; Manitoba Health","funders":"Canadian Institutes of Health Research","keywords":"Health Insurance Portability and Accountability Act; Identification (biology); Computer science; Protected health information; Information retrieval; MEDLINE; Personally identifiable information; Digital library; Data science; World Wide Web; Medicine; Confidentiality; Public health; Nursing; Health education","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.1179316,0.002912121,0.01014686,0.04757902,0.002729577,0.008884093,0.00529525,0.005956098,0.008141364],"category_scores_gemma":[0.3955746,0.002750113,0.008460402,0.03916986,0.004727517,0.01333713,0.006639732,0.004578628,0.002065458],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.009899961,"about_ca_system_score_gemma":0.04620852,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00661184,"about_ca_topic_score_gemma":0.01296552,"domain_scores_codex":[0.8518749,0.06363352,0.05323634,0.00700869,0.02267695,0.001569712],"domain_scores_gemma":[0.4428506,0.4358991,0.05691919,0.01069229,0.05182886,0.001809875],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"systematic_review","study_design_gemma":"systematic_review","study_design_scores_codex":[0.00009494887,0.00002380582,0.0004864519,0.8736622,0.001198931,0.0001621444,0.001518036,0.0001723779,0.000310608,0.002326585,0.007749103,0.1122948],"study_design_scores_gemma":[0.00003167249,0.0000426332,0.0005628764,0.9571356,0.002185999,0.0001660349,0.0005560673,0.00007186697,0.0002106478,0.001235491,0.03776786,0.00003334494],"study_design_candidate":"systematic_review","study_design_consensus":"systematic_review","genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.0007425002,0.9848749,0.003739962,0.003828664,0.001224412,0.003013481,0.0009383156,0.00005717646,0.001580695],"genre_scores_gemma":[0.009515936,0.964971,0.01054648,0.003045401,0.0007224187,0.009390373,0.001113136,0.00006251783,0.0006326167],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.1179316,"threshold_uncertainty_score":0.6236892,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.7026193400916434,"score_gpt":0.6308973187983484,"score_spread":0.07172202129329497,"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."}}