Distress trajectories as indicators of biopsychosocial supportive care vulnerability in a nationwide PRO-based oncology program.

Abstract

543 Background: Routine distress screening is widely implemented in oncology, yet its role as an operational trigger for comprehensive supportive care assessment remains unclear. This study sought to evaluate whether longitudinal distress trajectories identify patients at risk for psychosocial, health-related quality of life (HRQoL), and nutritional vulnerability. Methods: Adults with cancer initiating treatment across Brazil were enrolled in the biopsychosocial support (BPS) program, a nationwide patient-reported outcome-based supportive care initiative using a clinical dashboard. Participants completed longitudinal assessments of distress (Distress Thermometer), anxiety/depression (HADS), HRQoL (FACT-G), and nutritional risk (PG-SGA) every 3 months. Patients with moderate/high distress received supportive care follow-up. GEE and mixed-effects models evaluated longitudinal outcomes, and logistic regression identified predictors of distress escalation. Results: Among 3,526 patients (median age, 59; 69% female), the most common cancer types were breast (37%) and gastrointestinal (23%); 41% had stage III-IV disease. Participants were mainly from the Southeast (67%) and Central-West (27%). Distress trajectories were classified as persistently low (58%), improved (18%), persistently moderate/high (13%), worsened (6%), and fluctuating (5%) (Table). Longitudinal analyses showed that worsening distress was associated with increased anxiety/depression risk (OR=1.57, p=0.001), HRQoL impairment (OR=1.81, p=0.001), nutritional risk (OR=1.20, p=0.009), and lower FACT-G total scores (β=-2.78, p=0.001). Improving distress was associated with reduced anxiety/depression risk (OR=0.72, p=0.006) and higher FACT-G scores (β=1.05, p=0.001). Among 2,343 patients with low baseline distress, 13% transitioned to moderate/high distress; baseline anxiety/depression risk, HRQoL impairment, and nutritional risk predicted distress escalation (AUC 0.649). Conclusions: Distress trajectories may provide a practical framework for risk stratification in oncology. Integrated into a real-time clinical dashboard, repeated distress assessment may facilitate proactive identification of supportive care needs and adaptive supportive care delivery in routine practice. Longitudinal changes in psychological symptoms, quality of life, and nutritional risk according to distress trajectory. Distress Trajectory HADS Moderate/High Risk (%) Baseline, Final FACT-G Impairment (%) Baseline, Final Nutritional Risk (%) Baseline, Final FACT-G Total Baseline, Final Persistently low 3.3, 5.2 18.4, 17.1 24.4, 34.3 89.8, 90.2 Improved 33.0, 14.9 56.5, 48.5 33.2, 45.2 78.3, 81.3 Worsened 11.0, 35.7 33.8, 66.7 34.7, 55.6 83.0, 75.9 Persistently moderate/high 53.2, 49.3 69.6, 81.2 36.6, 50.7 72.4, 68.7 Fluctuating 27.1, 29.1 45.2, 54.4 31.3, 44.3 81.2, 80.7