Implementing UK Oncology Nursing Society-Informed Digital Symptom Triage With Episode-Based Review in Routine NHS Acute Oncology: Service Evaluation

Patients receiving systemic anticancer therapy (SACT) can deteriorate clinically between scheduled appointments; yet, acute oncology services often rely on reactive helplines with limited longitudinal symptom visibility.

The aim of this study was to evaluate the feasibility, safety, and workflow integration of OncsCare, a digital symptom triage platform mapping patient-reported symptoms to UK Oncology Nursing Society (UKONS) acuity tiers with episode-based clinician review.

This 10-week service evaluation (July to September 2025) implemented OncsCare within a UK tertiary acute oncology service. Patients completed daily symptom check-ins mapped to UKONS-informed green/amber/red tiers. Alerts were grouped into episode-level triage events using prespecified rules (48-hour window, symptom-domain continuity) to represent the operational workload. Outcomes included engagement, alert distribution, escalation pathways, timeliness of clinical response, and safety signals via structured case-finding.

Thirty-two patients participated (none withdrew) in this study. Daily check-in completion rate was 91.7% (1444/1574 expected patient-days). From 362 amber/red alerts, 62 episodes were generated; 38.7% (24/62) were clinically actionable, resulting in telephone management (n=12, 50%), acute care assessment (n=9, 37.5%), emergency referral (n=2, 8.3%), or admission (n=1, 4.2%). Median time to first clinical response for in-hours red alerts was 47 minutes. Predefined safety case-finding identified no intervention-attributable safety signals. Patients reported increased home reassurance (n=17, 85%), and clinicians reported improved situational awareness without increased workload.

UKONS-informed digital triage with episode-based review demonstrated feasibility, with no intervention-attributable safety signals identified in this small single-center evaluation. This operational model addresses alert fragmentation and supports prospective multicenter evaluation.