{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,10,13]],"date-time":"2026-10-13T10:45:39Z","timestamp":1791888339643,"version":"4.3.4"},"posted":{"date-parts":[[2026,9,21]]},"group-title":"Health Informatics","reference-count":31,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T00:00:00Z","timestamp":1789948800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,10,8]]},"abstract":"<jats:sec>\n                  <jats:title>Objective<\/jats:title>\n                  <jats:p>To develop and evaluate machine learning methods for automated identification of family relationships from electronic health records (EHRs). We employed Random Forest classifiers to identify five relationship types: Mother-Child, Father-Child, Sibling-Sibling, Twin-Twin, and Partner-Partner, using a four-stage iterative refinement process to improve precision for patient-centered outcomes research (PCOR) and healthcare applications.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Materials and Methods<\/jats:title>\n                  <jats:p>We used two large-scale Indiana datasets: the Indiana Network for Patient Care (INPC), comprising approximately 15 million unique individuals across approximately 45 million medical records, and the Indiana Natality dataset (birth certificates, 1970-2024), which served as the gold standard. Positive cases were derived by linking verified relationships from Natality records to INPC using a shared Global Identifier. Negative cases were drawn from a three-tier blocking strategy combined with sliding window restriction and similarity scoring, which together reduced the comparison space from over 10\u00b9\u2074 potential pairs to approximately 121 million candidates. Five Random Forest classifiers underwent four-stage iterative refinement: initial training, feature removal for multicollinearity, feature engineering to address systematic errors, and training-data refinement restricted to relationship pairs with at least one shared contact feature.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>The models achieved precision of 0.92 to 1.00, recall of 0.97 to 1.00, and F1 scores of 0.94 to 1.00 across all relationship types (Mother-Child 0.97, Father-Child 0.98, Sibling-Sibling 0.98, Twin-Twin 1.00, Partner-Partner 0.94). Between 15% and 78% of Natality-verified relationships lacked any shared contact information in INPC and were structurally undetectable by record linkage; restricting training and evaluation to linkable pairs improved precision for four of the five models, and training-data refinement improved F1-scores by a further 0.02-0.04 for the Mother-Child, Father-Child, and Sibling models. High-confidence predictions (probability \u2265 0.9) captured 77-99% of true positives. Age difference was the primary predictor for parent-child and twin relationships, while phone number similarity was most important for siblings.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Discussion<\/jats:title>\n                  <jats:p>The framework identifies family relationships from standard demographic fields available in most EHR systems. The iterative refinement approach shows how systematic error analysis can guide methodological improvements. Training-data refinement addresses a constraint in EHR-based family linkage: not all verified biological relationships have overlapping demographic footprints in healthcare data. Unlike rule-based approaches, the framework provides probability scores that enable configurable deployment thresholds.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Conclusion<\/jats:title>\n                  <jats:p>The Random Forest models demonstrate high performance suitable for large-scale research deployment and are ready for testing in clinical applications rather than immediate widespread clinical use. The methodology can be adapted to other EHR systems, supporting family-centered study design in patient-centered outcomes research.<\/jats:p>\n                <\/jats:sec>","DOI":"10.64898\/2026.09.18.26363428","type":"posted-content","created":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T19:25:17Z","timestamp":1790018717000},"source":"Crossref","is-referenced-by-count":0,"title":["Identifying Family Relationships from Electronic Health Records: A Machine Learning Approach"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0009-0007-2316-4381","authenticated-orcid":false,"given":"Abhinav","family":"Pundir","sequence":"first","affiliation":[{"name":"Colorado School of Public Health, University of Colorado Anschutz, Aurora, CO 80045, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Andrew","family":"Hill","sequence":"additional","affiliation":[{"name":"Colorado School of Public Health, University of Colorado Anschutz, Aurora, CO 80045, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael G.","family":"Kahn","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics, University of Colorado Anschutz, Aurora, CO 80045, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bethany M.","family":"Kwan","sequence":"additional","affiliation":[{"name":"Colorado Clinical and Translational Sciences Institute, University of Colorado Anschutz, Aurora, CO 80045, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Daniel M.","family":"Lindberg","sequence":"additional","affiliation":[{"name":"Department of Emergency Medicine, Kempe Center for the Prevention & Treatment of Child Abuse & Neglect, University of Colorado Anschutz, Aurora, CO 80045, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shaun J.","family":"Grannis","sequence":"additional","affiliation":[{"name":"Department of Family Medicine, Indiana University School of Medicine, Indianapolis, Indiana, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Titus K.","family":"Schleyer","sequence":"additional","affiliation":[{"name":"Regenstrief Institute, Indianapolis, IN, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lisa M.","family":"Schilling","sequence":"additional","affiliation":[{"name":"Division of General Internal Medicine, University of Colorado Anschutz, Aurora, CO 80045, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sarah V.","family":"Kautz","sequence":"additional","affiliation":[{"name":"Colorado Clinical and Translational Sciences Institute, University of Colorado Anschutz, Aurora, CO 80045, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Toan C.","family":"Ong","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics, University of Colorado Anschutz, Aurora, CO 80045, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026101302550771000_2026.09.18.26363428v2.1","doi-asserted-by":"publisher","DOI":"10.7326\/0000605-200912150-00165"},{"issue":"1","key":"2026101302550771000_2026.09.18.26363428v2.2","doi-asserted-by":"crossref","first-page":"27","DOI":"10.1097\/MOP.0000000000001415","article-title":"Linking household members and defining relational networks using routine health data","volume":"37","year":"2025","journal-title":"Curr Opin Pediatr"},{"key":"2026101302550771000_2026.09.18.26363428v2.3","unstructured":"Polubriaginof F , Tatonetti NP , Vawdrey DK . An assessment of family history information captured in an electronic health record. AMIA Annu Symp Proc. 2015;2015:2035\u201342."},{"key":"2026101302550771000_2026.09.18.26363428v2.4","doi-asserted-by":"publisher","DOI":"10.1001\/jama.292.12.1480"},{"key":"2026101302550771000_2026.09.18.26363428v2.5","doi-asserted-by":"crossref","unstructured":"Christen P. Data matching: concepts and techniques for record linkage, entity resolution, and duplicate detection. Berlin: Springer; 2012.","DOI":"10.1007\/978-3-642-31164-2"},{"issue":"9","key":"2026101302550771000_2026.09.18.26363428v2.6","doi-asserted-by":"crossref","first-page":"2025","DOI":"10.1007\/s10995-014-1453-8","article-title":"Linkage Methods for Connecting Children with Parents in Electronic Health Record and State Public Health Insurance Data","volume":"18","year":"2014","journal-title":"Matern Child Health J"},{"issue":"32","key":"2026101302550771000_2026.09.18.26363428v2.7","doi-asserted-by":"crossref","first-page":"e26950","DOI":"10.1097\/MD.0000000000026950","article-title":"Creation of a linked cohort of children and their parents in a large, national electronic health record dataset","volume":"100","year":"2021","journal-title":"Medicine (Baltimore"},{"key":"2026101302550771000_2026.09.18.26363428v2.8","doi-asserted-by":"publisher","DOI":"10.1093\/ije\/dyx177"},{"key":"2026101302550771000_2026.09.18.26363428v2.9","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2018.04.032"},{"issue":"5","key":"2026101302550771000_2026.09.18.26363428v2.10","doi-asserted-by":"crossref","first-page":"915","DOI":"10.1093\/jamia\/ocad028","article-title":"Using electronic health record data to link families: an illustrative example using intergenerational patterns of obesity","volume":"30","year":"2023","journal-title":"J Am Med Inform Assoc"},{"key":"2026101302550771000_2026.09.18.26363428v2.11","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0164667"},{"key":"2026101302550771000_2026.09.18.26363428v2.12","doi-asserted-by":"publisher","DOI":"10.1002\/gepi.21855"},{"issue":"9","key":"2026101302550771000_2026.09.18.26363428v2.13","doi-asserted-by":"crossref","first-page":"1407","DOI":"10.1093\/jamia\/ocaf102","article-title":"Identifying family structures from obituaries and matching them to patients in an electronic health record","volume":"32","year":"2025","journal-title":"J Am Med Inform Assoc"},{"key":"2026101302550771000_2026.09.18.26363428v2.14","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0248195"},{"key":"2026101302550771000_2026.09.18.26363428v2.15","doi-asserted-by":"publisher","DOI":"10.1007\/s10654-018-0356-1"},{"key":"2026101302550771000_2026.09.18.26363428v2.16","doi-asserted-by":"publisher","DOI":"10.1007\/s00038-012-0379-4"},{"key":"2026101302550771000_2026.09.18.26363428v2.17","doi-asserted-by":"publisher","DOI":"10.1007\/978-1-59745-423-0_10"},{"issue":"5","key":"2026101302550771000_2026.09.18.26363428v2.18","doi-asserted-by":"crossref","first-page":"619","DOI":"10.1053\/j.ajkd.2017.05.007","article-title":"Family Aggregation and Heritability of ESRD in Taiwan: A Population-Based Study","volume":"70","year":"2017","journal-title":"Am J Kidney Dis"},{"issue":"3","key":"2026101302550771000_2026.09.18.26363428v2.19","first-page":"264","article-title":"Linkable administrative files: Family information and existing data","volume":"8","year":"2017","journal-title":"Longit Life Course Stud"},{"key":"2026101302550771000_2026.09.18.26363428v2.20","doi-asserted-by":"publisher","DOI":"10.1093\/ije\/dym136"},{"issue":"1","key":"2026101302550771000_2026.09.18.26363428v2.21","first-page":"1686","article-title":"Multigenerational health research using population-based linked databases: an international review","volume":"6","year":"2021","journal-title":"Int J Popul Data Sci"},{"key":"2026101302550771000_2026.09.18.26363428v2.22","doi-asserted-by":"crossref","unstructured":"Weaver J , Hardin JH , Blacketer C , et al. Development and evaluation of an algorithm to link mothers and infants in two US commercial healthcare claims databases for pharmacoepidemiology research. BMC Med Res Methodol. 2023;23:246.","DOI":"10.1186\/s12874-023-02073-6"},{"key":"2026101302550771000_2026.09.18.26363428v2.23","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btab419"},{"key":"2026101302550771000_2026.09.18.26363428v2.24","doi-asserted-by":"crossref","unstructured":"Williams KS , Rahurkar S , Grannis SJ , et al. Evolution of clinical Health Information Exchanges to population health resources: a case study of the Indiana network for patient care. BMC Med Inform Decis Mak. 2025;25:97.","DOI":"10.1186\/s12911-025-02933-9"},{"key":"2026101302550771000_2026.09.18.26363428v2.25","doi-asserted-by":"publisher","DOI":"10.1093\/jamia\/ocac068"},{"key":"2026101302550771000_2026.09.18.26363428v2.26","doi-asserted-by":"publisher","DOI":"10.1016\/j.cpc.2010.04.018"},{"key":"2026101302550771000_2026.09.18.26363428v2.27","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-020-2649-2"},{"issue":"8","key":"2026101302550771000_2026.09.18.26363428v2.28","first-page":"707","article-title":"Binary codes capable of correcting deletions, insertions, and reversals","volume":"10","year":"1966","journal-title":"Soviet Physics Doklady"},{"key":"2026101302550771000_2026.09.18.26363428v2.29","doi-asserted-by":"publisher","DOI":"10.1023\/A:1010933404324"},{"key":"2026101302550771000_2026.09.18.26363428v2.30","first-page":"281","article-title":"Random search for hyper-parameter optimization","volume":"13","year":"2012","journal-title":"J Mach Learn Res"},{"key":"2026101302550771000_2026.09.18.26363428v2.31","unstructured":"Kohavi R. A study of cross-validation and bootstrap for accuracy estimation and model selection. In: Proceedings of the 14th International Joint Conference on Artificial Intelligence (IJCAI). Vol 2. San Francisco: Morgan Kaufmann; 1995:1137-43."}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.09.18.26363428","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,13]],"date-time":"2026-10-13T09:55:15Z","timestamp":1791885315000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.09.18.26363428"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,21]]},"references-count":31,"URL":"https:\/\/doi.org\/10.64898\/2026.09.18.26363428","relation":{},"subject":[],"published":{"date-parts":[[2026,9,21]]},"subtype":"preprint"}}