{"corpus_id":191940883,"paper_sha":"0ed4da9460dab7b671d918073acb7218551d409c","doi":"10.1016/J.BRAT.2019.103412","arxiv_id":null,"pmid":31233922,"pmcid":"PMC7556331","mag_id":2947984607,"dblp_id":null,"acl_id":null,"title":"Machine learning methods for developing precision treatment rules with observational data.","year":2019,"publication_date":"2019-09-01","venue":"Behaviour Research and Therapy","journal":{"name":"Behaviour research and therapy","pages":"\n          103412\n        ","volume":"120"},"journal_issn":null,"journal_title":null,"publication_types":["JournalArticle"],"pubmed_pub_types":["Journal Article","Research Support, Non-U.S. Gov't"],"s2_fields_of_study":["Medicine","Computer Science"],"reference_count":135,"citation_count":60,"influential_citation_count":1,"is_open_access":false,"arxiv_categories":null,"arxiv_license":null,"arxiv_journal_ref":null,"mesh_headings":[{"d":"Clinical Decision Rules","mj":true,"ui":"D000081415"},{"d":"Databases, Factual","mj":false,"ui":"D016208"},{"d":"Decision Support Systems, Clinical","mj":false,"ui":"D020000"},{"d":"Humans","mj":false,"ui":"D006801"},{"d":"Information Storage and Retrieval","mj":false,"ui":"D016247"},{"d":"Machine Learning","mj":true,"ui":"D000069550"},{"d":"Observational Studies as Topic","mj":false,"ui":"D064887"},{"d":"Precision Medicine","mj":true,"ui":"D057285"},{"d":"Suicide, Attempted","mj":false,"ui":"D013406"},{"d":"Suicide Prevention","mj":true,"ui":"D000092864"}],"chemicals":null,"comments_corrections":null,"source_flags":5,"s2_open_access_pdf_url":null,"s2_open_access_landing_url":null,"s2_open_access_license":null,"s2_open_access_status":null,"pmc_open_access_pdf_url":null,"pmc_open_access_landing_url":null,"pmc_open_access_license":null,"pmc_open_access_status":null,"unpaywall_open_access_pdf_url":null,"unpaywall_open_access_landing_url":null,"unpaywall_open_access_license":null,"unpaywall_open_access_status":null,"abstract":"Clinical trials have identified a variety of predictor variables for use in precision treatment protocols, ranging from clinical biomarkers and symptom profiles to self-report measures of various sorts. Although such variables are informative collectively, none has proven sufficiently powerful to guide optimal treatment selection individually. This has prompted growing interest in the development of composite precision treatment rules (PTRs) that are constructed by combining information across a range of predictors. But this work has been hampered by the generally small samples in randomized clinical trials and the use of suboptimal analysis methods to analyze the resulting data. In this paper, we propose to address the sample size problem by: working with large observational electronic medical record databases rather than controlled clinical trials to develop preliminary PTRs; validating these preliminary PTRs in subsequent pragmatic trials; and using ensemble machine learning methods rather than individual algorithms to carry out statistical analyses to develop the PTRs. The major challenges in this proposed approach are that treatment are not randomly assigned in observational databases and that these databases often lack measures of key prescriptive predictors and mental disorder treatment outcomes. We proposed a tiered case-cohort design approach that uses innovative methods for measuring and balancing baseline covariates and estimating PTRs to address these challenges.","claims":[{"public_id":"cl_d50cfc09438377925898fe537c41b93f","status":"active","text":"A tiered case-cohort design is proposed to address treatment nonrandomization and missing predictor or outcome measures in observational databases.","confidence":0.94,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_d50cfc09438377925898fe537c41b93f"},{"public_id":"cl_418cbb07f3cd4be961ab405700dbd5ad","status":"active","text":"Ensemble machine learning methods are proposed for statistical analysis in developing precision treatment rules instead of relying on individual algorithms.","confidence":0.92,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_418cbb07f3cd4be961ab405700dbd5ad"},{"public_id":"cl_f9c60b9d98206762ffcf784624d41204","status":"active","text":"Large observational electronic medical record databases can be used to develop preliminary precision treatment rules, with subsequent validation in pragmatic trials.","confidence":0.95,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_f9c60b9d98206762ffcf784624d41204"}],"concepts":[{"public_id":"co_03813ed4b8364ef5d88c1ff0c9faeeef","status":"active","name":"predictor variables","description":"Clinical, symptom, or self-report variables used as inputs for treatment selection.","types":["feature","measure"],"aliases":["predictors"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_03813ed4b8364ef5d88c1ff0c9faeeef"},{"public_id":"co_40c8c0f5790c47be1c98bbea3507e282","status":"active","name":"ensemble machine learning methods","description":"Machine learning approaches that combine predictions from multiple algorithms.","types":["method"],"aliases":["ensemble methods"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_40c8c0f5790c47be1c98bbea3507e282"},{"public_id":"co_49cc8e35ff68eb808035fb771f5badfd","status":"active","name":"self-report measures","description":"Patient-reported assessments used as predictor variables.","types":["measure"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_49cc8e35ff68eb808035fb771f5badfd"},{"public_id":"co_86872be53f8a1c606139ebc0419c9a53","status":"active","name":"pragmatic trials","description":"Trials conducted in real-world clinical settings to evaluate interventions under routine conditions.","types":["study design"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_86872be53f8a1c606139ebc0419c9a53"},{"public_id":"co_8c1f53faf1b739620d0f495bb1e98dea","status":"active","name":"precision treatment rules","description":"Composite decision rules that combine multiple predictors to guide treatment selection for individual patients.","types":["method","decision rule"],"aliases":["PTRs"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_8c1f53faf1b739620d0f495bb1e98dea"},{"public_id":"co_aa8fb73457f568cd4543e258b44abdaf","status":"active","name":"baseline covariates","description":"Patient characteristics measured before treatment or outcome assessment.","types":["feature","covariate"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_aa8fb73457f568cd4543e258b44abdaf"},{"public_id":"co_b8da557bdc253890e863f453902f610a","status":"active","name":"tiered case-cohort design","description":"A sampling design that combines case-cohort elements in multiple stages to support estimation under observational data constraints.","types":["study design","sampling design"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_b8da557bdc253890e863f453902f610a"},{"public_id":"co_b96fa467c88cdcc7dea1f0a0df430d64","status":"active","name":"clinical biomarkers","description":"Biological markers measured in clinical settings and used as potential predictors.","types":["measure"],"aliases":["biomarkers"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_b96fa467c88cdcc7dea1f0a0df430d64"},{"public_id":"co_ba530bb33a8bcb39c85028a8c0e7aff9","status":"active","name":"observational electronic medical record databases","description":"Electronic health record datasets collected from routine care without random treatment assignment.","types":["dataset","data source"],"aliases":["electronic medical record databases","EMR databases"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_ba530bb33a8bcb39c85028a8c0e7aff9"},{"public_id":"co_cd6a8e8d7b02bacc3fa6e827edd3e5bf","status":"active","name":"individual algorithms","description":"Single machine learning algorithms used separately for analysis or prediction.","types":["method"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_cd6a8e8d7b02bacc3fa6e827edd3e5bf"},{"public_id":"co_ea186e62a57185218075f4d713fd073d","status":"active","name":"symptom profiles","description":"Patterns or summaries of patient symptoms used as candidate predictors.","types":["measure","feature"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_ea186e62a57185218075f4d713fd073d"},{"public_id":"co_eaee7bae82d30f27f2e4ab865bf4195d","status":"active","name":"randomized clinical trials","description":"Controlled trials in which treatment assignment is randomized.","types":["study design"],"aliases":["RCTs"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_eaee7bae82d30f27f2e4ab865bf4195d"},{"public_id":"co_ef66b5cdebb06333be2b2cf61d61a6bc","status":"active","name":"composite precision treatment rules","description":"Treatment rules formed by combining information from multiple predictors.","types":["method","decision rule"],"aliases":["composite PTRs"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_ef66b5cdebb06333be2b2cf61d61a6bc"}],"external_ids":{"DOI":"10.1016/J.BRAT.2019.103412","ArXiv":null,"PubMed":31233922,"PubMedCentral":"PMC7556331","MAG":2947984607,"DBLP":null,"ACL":null},"open_access":{"is_open_access":false,"pdf_url":null,"landing_url":"https://sah.borca.ai/papers/191940883","source":null,"pdf_url_source":null,"license":null,"reason":"pdf_url_not_indexed"},"reference_availability":{"status":"available","references_indexed":true,"full_text_available":false,"full_text_source":null,"count_basis":"semantic_scholar_metadata","extraction_status":"not_applicable","reason":null},"source":{"provider":"episteme2","base_corpus":"semantic_scholar_dump","freshness_mode":"unknown","basis":["semantic_scholar_metadata","postgres_metadata"],"limits":["paper metadata is based on indexed upstream scholarly datasets","claims and concepts are available only for extracted papers","absence of claims or concepts means no extracted graph data is available in this response"],"status":"available","degraded":false,"degraded_reasons":[],"diagnostics":{"status":"available","degraded":false,"degraded_reasons":[],"metadata_status":"available","graph_status":"available","abstract_status":"available"},"source_flags":5},"paper_id":638991,"paper_uid":"4848cdfd-bd78-484c-a930-338fbf9524fa","canonical_identity":{"paper_id":638991,"paper_uid":"4848cdfd-bd78-484c-a930-338fbf9524fa","identity_status":"available","lookup_basis":"semantic_scholar_external_id","compatibility_path":"corpus_id"},"url":"https://sah.borca.ai/papers/191940883"}