Implementing routine electronic patient-reported health literacy screening for new adult oncology patients to improve patient care at the individual and population level.

378 Background: Low health literacy (HL) profoundly affects oncology care and outcomes. In April 2023, the Dana-Farber Cancer Institute (DFCI) Patient Reported Data (PRD) Program implemented an EHR-integrated version of the validated BRIEF HL Screener (BRIEF). BRIEF identifies patients who may need tailored communication and educational resources, while also offering an understanding of population-level needs and the influence of HL on outcomes. Methods: PRD, with guidance from patient advocates and HL experts, implemented BRIEF as part of our New Patient Intake Questionnaire assigned to patients before their first visit. BRIEF was selected for its brevity, validation, and scoring. New adult oncology patients (or proxy) were offered BRIEF in seven languages via the patient portal or tablet in clinic. Responses were automatically stored in the EHR. An analysis of response data and patient characteristics was conducted to examine variation in rates of low HL among respondent groups stratified by patient characteristics and to identify potential drivers of low HL using chi-squared tests. Results: Between April 23, 2023, and May 24, 2024, 70% (22,101/31,793) of assigned patients responded to BRIEF. Among respondents, 29% (6,487) reported “limited” or “marginal” HL based on BRIEF scoring. Responses were immediately available in the EHR for care team use with “limited” or “marginal” scores highlighted to support clinical decisions. Analysis identified multiple patient characteristics related to increased likelihood of reporting low HL (see table). Conclusions: Implementation of BRIEF is a critical step towards improving communication and achieving better outcomes for oncology patients. While our reach is high at 70%, we continue to close gaps in screening completion and use analysis of demographic data to identify patients at risk of low HL even if they do not complete BRIEF. Future analysis will include adjusted odds ratios to determine feature importance in predicting low HL. To address low HL using BRIEF, we are socializing data among clinical leaders to increase use of patient-reported HL scores among staff and collaborating with patient education specialists to ensure that materials are at an appropriate literacy level, ultimately aiming to ensure accessibility for all patients regardless of literacy level. Drivers of low HL. Feature Reference Group Unadj. Odds Ratio N Primary Language: Non-English Primary Language: English 8.59 985 Education Level: 8 th Grade or less Education Level:College Degree 25.06 240 Education Level: Some high school Education Level:College Degree 6.04 281 Education Level: High School Education Level:College Degree 2.99 4,008 Age: 70 – 79 Age: 60 – 69 1.39 4,264 Age: 80 – 89 Age: 60 – 69 3.26 1,279 Age: 90 + Age: 60 – 69 14.23 125 Relationships between features and HL are significant with p-values < 0.001.