Abstract
BACKGROUND AND PURPOSE: To evaluate longitudinal patient-reported distress in cervical cancer patients treated with definitive chemoradiation (CRT). MATERIALS AND METHODS: Between 2011 and 2016, consenting cervical cancer patients treated with definitive CRT who completed ≥ 2 revised Edmonton Symptom Assessment System (ESAS-r) questionnaires at clinical visits, including baseline, were included. A linear mixed model was used to assess the longitudinal trend in ESAS-r. A minimal clinically important difference (MCID) for total ESAS-r score was defined as a change of 3-points for improvement and 4-points for deterioration. The proportion of patients with an MCID over time was described using moving averages. To test for changes, mixed effects logistic models were fitted, each of which included patient-specific random intercepts and random slopes. RESULTS: 67 patients were eligible for analysis (736 ESAS-r assessments). Median (range) follow-up was 24 months (range: 15-45) and compliance at 12 months was 60% (40/67). There was a significant decrease in ESAS-r scores over time. Baseline ESAS-r was strongly predictive of ESAS-r at follow-up (p < 0.001). The proportion of patients with an MCID for improvement from baseline significantly increased over time (p < 0.001) and the proportion with an MCID for deterioration significantly decreased over time (p < 0.001). No predictors for distress were found. CONCLUSIONS: Long-term cervical cancer survivors experience distress that significantly improves over time to an extent expected to be clinically meaningful for patients. Implementing cervical cancer specific patient-reported outcome tools into practice could better inform patient needs.