Characterization of remote symptom alerts generated from patient-reported outcome (PRO) symptom monitoring in a real-world breast cancer population.

Abstract

e23193 Background: Women with breast cancer experience debilitating symptoms, yet the patient-reported symptom burden among patients on treatment remains largely uncharacterized. This study characterized electronic PRO (ePRO) symptom monitoring data of patients with breast cancer in real-world settings to assess differences in remote symptom alerts by clinical characteristics. Methods: We conducted a retrospective descriptive study of complete case ePRO data reported by patients at twelve US sites using Carevive PROmPT, a remote symptom monitoring platform. Participants consent to share de-identified, aggregate ePRO symptom data and sociodemographic and clinical characteristics from the electronic health record. We characterized PRO-CTCAE symptom data reported by breast cancer patients on active cancer treatment enrolled between 9/2020 and 11/2023. The PRO platform generates an algorithm-based alert if the combination of symptom severity, frequency, and interference is moderate or severe. We used descriptive statistics to characterize symptom severity and number of alerts by baseline patient-report ECOG status, age, cancer stage (Stage I-III vs. IV), and breast cancer subtype. Results: Of 646 patients, 519 (80%) generated alerts during treatment. Population characteristics were: median age 56 years-old (range 26-84), 72% White, 20% Black/African American, 62% Stage I-III, 15% HER2 positive, 47% ER and/or PR positive, HER2 negative,18% triple negative. Patients reported a total of 19,425 symptoms over a median of 12 weeks with 80% compliance rate, of which 39% resulted in alerts. Participants generated a median of 1 alert per week. About 50% of alerts occurred in the first 12 weeks of ePRO completion, with a higher proportion in Week 1-4 across all subgroups (range, 15-38%). On average, ECOG 2 or greater patients generated more alerts per week than ECOG 0-1, (mean 2.1, (SD 1.3) vs. 1.7 (SD 1.0). Younger patients ( < 50 years) generated more alerts compared to those older than 50 years (87% vs. 77%). Patients with early-stage disease reported more alerts than the Stage IV group (81% vs. 78%). In the metastatic setting, triple negative patients reported a median of 3 (range 1-8) moderate/severe symptoms per week; the other subtypes reported a median of 1 (range 1-8). Conclusions: Women on breast cancer treatment reported their symptom experience with high compliance rates in the real-world setting. Nearly 40% of symptom reports generated an alert, with a higher proportion in the first month of treatment. Younger, early-stage, and metastatic triple negative patients with breast cancer reported a higher symptom burden compared to other groups. Future work will investigate symptom experience by sociodemographic and clinical characteristics to further identify patients who may benefit from more intensive symptom reporting and management.