TPS1682 Background: Real-world evidence generation in oncology is often limited by inconsistent and unstructured documentation within electronic health records (EHRs). Key clinical variables including disease status, performance status, and treatment change rationale are frequently embedded in free text, requiring retrospective manual abstraction that is inefficient and error-prone. As a result, clinically meaningful real-world endpoints, such as time to treatment change and real-world progression-free survival (rwPFS), are inconsistently captured. To address these gaps, the University of Pennsylvania developed OncAssessment (OA), an EHR-embedded SmartForm integrated directly into routine oncology clinical documentation. Unlike post hoc abstraction or standalone registries, OA enables prospective, structured data capture of key oncologic variables at the point of care by treating clinicians, and is paired alongside with electronic patient-reported outcomes (ePROs). This approach aligns clinician decision-making with patient-reported experience within routine workflows, enabling timely, scalable, and higher-fidelity real-world data (RWD) capture. Methods: This pragmatic implementation study, developed by the University of Pennsylvania in collaboration with Gilead Sciences, evaluates OA and ePRO implementation in standard clinical practice. Adult patients with stage IIIB-IV unresectable non-small cell lung cancer (NSCLC) initiating standard-of-care immunotherapy with or without chemotherapy (Cohort I) and adult breast cancer patients receiving systemic infusional therapy (Cohort II) are identified via the EHR. 200 patients per cohort are planned for enrollment over 2 years. Enrollment began July 22, 2025. For identified patients, clinicians are prompted to complete OA at each visit: documenting clinical status, tumor status, treatment changes, and rationale. ePROs are automatically distributed through the EHR to Cohort I patients prior to their visits. Structured data from OA, ePROs, and other EHR sources are auto-extracted. Primary objectives for Cohort I include OA utilization, time to treatment change decision (TTCd), comparison of TTCd with time to treatment change (TTC) derived from EHR reporting, ePRO completion rates, and concordance between patient- and clinician-assessed clinical status. Secondary objectives include rwPFS captured via OA, comparison with rwPFS determined by manual chart and radiology review, characterization of patient experience via ePROs, and exploratory analyses of discordance and factors associated with treatment change. The primary objective for Cohort II is OA utilization. Follow-up continues until progression, treatment change, or 2 years for Cohort I and 12 months for Cohort II.
Singh et al. (Thu,) studied this question.