From patient needs to clinical action: Implementing scalable, customizable ePRO-driven supportive care in head and neck cancer.

Abstract

445 Background: Supportive care needs are frequently under-detected in routine oncology, and even when identified, fragmented workflows impede coordinated response. We developed and deployed a modular, EHR-integrated electronic patient-reported outcome (ePRO) and clinical decision support (CDS) system for proactive, guideline-concordant supportive care in ambulatory head and neck cancer (HNC) and evaluated real-world implementation. Methods: Guided by the Consolidated Framework for Implementation Research (CFIR 2.0) and informed by 46 formative interviews (20 clinicians, 26 patients), we built a patient-facing assessment comprising a core symptom module (ESAS-r) and an HNC-customized module of PRO-CTCAE items plus financial (FACIT-COST) and malnutrition (MST) screens. Responses prompted logic-based algorithms and clinician-facing CDS with queued referral orders. The system was deployed across HNC clinics over 20 months (April 2024–December 2025), with ePROs delivered automatically before visits. In this initial evaluation, we assessed reach, unmet need, clinician CDS engagement, and patient and clinician experience. Results: Of 1,195 eligible patients, 833 (70%) completed ePROs (median age 68; 71% male). Non-responders were more likely non-English-speaking, Hispanic/Latino, and from more disadvantaged neighborhoods (all p<.05). Among responders, 625 (75%) reported burden warranting ≥1 referral. Malnutrition risk screened positive in 169 (20.3%) and financial hardship in 94 (11.3%). Among 76 ePRO users surveyed, most reported it helped communicate what mattered most to their care team (83%) and added value to visits (66%); however, 30% reported their clinician rarely or never discussed results with them. Documented CDS engagement was low (5–12% monthly rates). Clinicians found the tool acceptable and appropriate but cited build issues, such as low alert visibility and ordering burden, as primary barriers. Conclusions: A modular EHR-integrated ePRO+CDS system achieved strong, low-burden reach and surfaced substantial multi-domain unmet need, yet identified need did not reliably translate into clinical action. Clinician feedback pointed to barriers in decision-support design and workflow, not overall value. Closing this gap through more visible decision support and team-based or navigator-driven routing is the focus of the next phase of implementation to integrate patient-reported need into routine care delivery.