{"id":"W4382400983","doi":"10.2196/46471","title":"Data Quality– and Utility-Compliant Anonymization of Common Data Model–Harmonized Electronic Health Record Data: Protocol for a Scoping Review","year":2023,"lang":"en","type":"review","venue":"JMIR Research Protocols","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Bundesministerium für Bildung und Forschung; Deutsche Forschungsgemeinschaft","keywords":"Computer science; Data quality; Context (archaeology); Data mining; Data science; Health care; Information retrieval; Quality (philosophy); Data anonymization; Protocol (science); Selection (genetic algorithm); Process (computing); Information privacy; Medicine; Metric (unit); Machine learning; Computer security","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","metaepi_narrow","open_science","research_integrity"],"consensus_categories":["metaresearch","open_science"],"category_scores_codex":[0.04431796,0.000788246,0.004050476,0.0007573867,0.0005673891,0.0007305681,0.1665693,0.0004967999,0.00001430871],"category_scores_gemma":[0.05047862,0.000674027,0.0001373577,0.003454352,0.0004908334,0.003429078,0.4477715,0.00233859,0.00002309817],"about_ca_system_candidate":true,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004017826,"about_ca_system_score_gemma":0.01004002,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001420192,"about_ca_topic_score_gemma":0.0003063261,"domain_scores_codex":[0.9820965,0.004838187,0.003919411,0.004926999,0.002123839,0.002095076],"domain_scores_gemma":[0.9060525,0.002986569,0.002385601,0.08773848,0.000557045,0.0002797908],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00003005252,0.000166398,7.879033e-7,0.3723768,0.00004208504,0.000001696973,0.000002204949,5.485571e-8,1.450158e-7,0.0002477615,0.2189164,0.4082156],"study_design_scores_gemma":[0.0006061823,0.0002639378,2.184493e-7,0.3904516,0.00001025954,0.000006184297,0.00000122992,0.07087799,5.26861e-7,0.007504792,0.5299625,0.0003146107],"study_design_candidate":"systematic_review","study_design_consensus":null,"genre_codex":"protocol","genre_gemma":"protocol","genre_scores_codex":[6.692409e-10,0.1662827,0.06959538,0.004251626,0.000008419021,0.7512771,0.008011367,0.0005577843,0.00001555379],"genre_scores_gemma":[3.994447e-10,0.3900285,0.06137815,0.00004949139,0.00003958204,0.5429823,0.005420814,0.00007674981,0.00002430426],"genre_candidate":"protocol","genre_consensus":"protocol","teacher_disagreement_score":0.407901,"threshold_uncertainty_score":0.999963,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.9009634784333649,"score_gpt":0.712507786119804,"score_spread":0.1884556923135609,"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."}}