BEGIN:VCALENDAR VERSION:2.0 PRODID:-//wp-events-plugin.com//7.2.3.1//EN TZID:America/Los_Angeles X-WR-TIMEZONE:America/Los_Angeles BEGIN:VEVENT UID:281@escience.washington.edu DTSTART;TZID=America/Los_Angeles:20251007T163000 DTEND;TZID=America/Los_Angeles:20251007T172000 DTSTAMP:20251002T212340Z URL:https://escience.washington.edu/events/uw-data-science-seminar-rachael -rosen/ SUMMARY:UW Data Science Seminar: Rachael Rosen DESCRIPTION:Please join us for a UW Data Science Seminar featuring recent U W Rehabilitation Science PhD graduate Rachael Rosen on Tuesday\, October 7 th from 4:30 to 5:20 p.m. PT. The seminar will be held in IEB G109.\n\n"Mo deling Patient Outcomes After Amputation: Comparing Machine Learning and T raditional Approaches"\n\nAbstract: Major lower limb amputation (LLA) in o lder adults often leads to poor outcomes\, including hospital readmission and death within months of discharge. This research used national Medicare data (2018–2021) to model two outcomes after LLA: 180-day hospital read mission and post-discharge mortality. Each was evaluated using both tradit ional statistical models and machine learning (ML) methods to compare pred ictive performance and clinical utility. For readmission\, a competing ris ks model was compared to a random survival forest. ML improved predictive accuracy and captured non-linear patterns not accounted for by traditional methods. High-risk subgroups were characterized by intensive care unit st ays\, high comorbidity burden\, prolonged hospitalization\, and end-stage renal disease. For mortality\, both a Cox model and a survival random fore st performed similarly in terms of prediction. However\, violation of the proportional hazards assumption made the ML model more appropriate. Mortal ity risk was highest for individuals discharged to hospice\, nursing facil ities\, or home without services\, and lower for those discharged with reh abilitation or home health care. ML methods offered greater flexibility an d identified risk patterns missed by traditional approaches. These results demonstrate how ML can enhance outcome prediction in clinical populations and highlight opportunities to support discharge planning for patients un dergoing major amputation.\n\n Biography: Rachael Rosen\, PhD\, is a reh abilitation researcher with expertise in health services research\, prosth etics and orthotics\, and machine learning applications in clinical outcom e modeling. She recently completed her PhD in Rehabilitation Science at th e University of Washington\, where her dissertation examined access to pos t-acute care\, hospital readmission\, and mortality following major lower limb amputation. Her work used national Medicare claims data to identify d isparities and evaluate predictive models across sociodemographic and clin ical subgroups.Before transitioning to research\, Dr. Rosen practiced clin ically as a certified prosthetist- orthotist (CPO) in Level I trauma cente rs and pediatric hospitals in Washington and Mississippi. These experience s shaped her interest in policy-driven research after observing how admini strative and insurance-related barriers limited rehabilitation access for patients with complex care needs. Since 2018\, she has served as a researc h scientist in the Department of Rehabilitation Medicine at UW\, contribut ing to the development of outcome measures and leading projects focused on mobility\, equity\, and post-acute care delivery.\n\n\nThe 2025-2026 semi nars will be held in person\, and are free and open to the public. ATTACH;FMTTYPE=image/jpeg:https://escience.washington.edu/wp-content/uploa ds/2025/10/Screenshot-2025-10-02-at-2.15.16 PM.png END:VEVENT BEGIN:VTIMEZONE TZID:America/Los_Angeles X-LIC-LOCATION:America/Los_Angeles BEGIN:DAYLIGHT DTSTART:20250309T030000 TZOFFSETFROM:-0800 TZOFFSETTO:-0700 TZNAME:PDT END:DAYLIGHT END:VTIMEZONE END:VCALENDAR