Factors Associated With Symptom Burden Among Pediatric Patients With Cancer

The objective was to identify factors associated with self-reported symptom burden measured using Symptom Screening in Pediatrics Tool (SSPedi) in pediatric patients with cancer.

This was a secondary analysis of a cluster randomized trial enrolling pediatric patients newly diagnosed with cancer. Twenty sites were randomized to routine symptom screening versus usual care. Intervention included thrice-weekly symptom screening with SSPedi, delivery of severely bothersome scores to health care teams, and implementation of locally adapted symptom management care pathways. Primary outcome was total SSPedi scores, 0 (no bothersome symptoms) to 60 (worst bothersome symptoms), obtained at baseline, week four, and week eight in 430 patients (n = 217 intervention and n = 213 usual care). We created a mixed linear regression model evaluating design (including time point), patient/guardian, and site characteristics for their associations with symptom burden after controlling for treatment assignment.

SSPedi scores were significantly lower at weeks 4 and 8 compared with baseline (P < .0001 overall), and at intervention versus control sites (P < .0001). In the full model, males (estimate, -3.3 [95% CI, -4.6 to -2.0]; P < .0001) and sites with higher physician staffing ratios (each physician full-time equivalent per 100 new diagnoses estimate -0.20 [95% CI, -0.5 to 0.0]; P = .024) had significantly lower total SSPedi scores.

Total symptom burden was reduced by time, intervention (symptom screening and care pathways), and greater physician staffing ratio. Females had higher symptom burden. These data may inform programmatic implementation of routine symptom screening in pediatric patients with cancer.