Breaking barriers in oncology data capture: Use of standardized oncology-specific note templates to enhance accuracy and completeness of structured data at the point of care.

619 Background: Structured data (SD) plays a pivotal role in leveraging the potential of electronic health records (EHRs) for improved clinical decision support (CDS), artificial intelligence (AI)-driven predictive models, precision medicine, and the utilization of real-world data (RWD) for research, quality improvement, registry development, and operational efficiency. To enhance disease (dx) and treatment (Rx) SD collection at the point of care (POC) with minimal disruption to provider workflows, we integrated a minimalistic Epic SmartForm into a standardized note template (Temp). In a prior study, this approach demonstrated 61% adoption by providers and improved documentation quality. This study builds on previous findings by evaluating the accuracy of SD entered or verified by treating providers at the POC. Methods: Between Feb 2023 and Mar 2024, 30,018 outpatient encounters occurred at our solid tumor oncology clinics, with 18,224 (61%) utilizing the Temp for note documentation. Among 34 providers, 20 (13 physicians, 7 advanced practice providers [APP]) met the criteria for Temp regular users (>20% of notes authored using the Temp). A stratified random sample of 700 notes written by Temp regular users was manually reviewed for the accuracy of four pre-specified SD elements: (1) phase of Treatment (Rx phase); (2) presence of metastatic dx (met dx); (3) current dx status (dx status); and (4) reason for treatment plan changes (Rx change). Pearson’s chi-squared tests were used to assess associations. Results: Of the 700 notes reviewed, 444 (63%) were authored by physicians and 256 (37%) by APPs. Data completeness and accuracy are summarized in Table 1. Inaccurate SD was uncommon (3-6%). Met dx and Rx phase data were nearly complete and highly accurate. However, dx status and Rx change data were missing in 14% and 25% of notes, respectively. Physician-authored notes demonstrated superior completeness and accuracy for dx status and Rx change compared to APP notes (p = 0.001). Follow-up notes contained more complete and accurate SD for met dx, dx status, and Rx change compared to initial consult notes (p < 0.001). Conclusions: The standardized note template with an embedded SmartForm facilitated high-quality SD collection at the point of care, with minimal inaccuracies. While data for certain elements (e.g., dx status and Rx change) requires improved completeness, differences in physicians vs APPs highlight opportunities for optimization and targeted training to enhance the SD quality. SD Element Accurate Not Accurate Not Applicable Not Entered Total n (%) Met dx 661 (94) 28 (4) 9 (1) 2 (0) 700 Rx phase 666 (95) 22 (3) - 12 (2) Dx status 560 (80) 39 (6) 1 (0) 100 (14) Rx change reason 490 (70) 23 (3) 15 (2) 172 (25)