Psychometric validation of the EORTC QLQ-OES18 in patients with advanced or metastatic esophageal squamous cell carcinoma.

Abstract

Background: The EORTC QLQ-OES18 has previously demonstrated clinical validity; however, there are limited published psychometric data for patients with advanced esophageal squamous cell carcinoma (ESCC). We evaluated the measurement properties of the QLQ-OES18 in a clinical trial population of patients with advanced or metastatic ESCC.

Methodology: Analyses used data from RATIONALE 302 (NCT03430843), a randomized phase 3 study of tislelizumab versus investigator-chosen chemotherapy as second-line treatment for patients with advanced or metastatic ESCC. Psychometric validation of the QLQ-OES18 included tests of reliability, construct validity, ability to detect change, and estimation of anchor-based meaningful within-patient change (MWPC) thresholds-the latter two being exploratory given that the trial was not powered to detect efficacy in patient-reported outcome endpoints.

Results: In total, 512 patients were randomized to either tislelizumab or chemotherapy; the average age was 61.5 years, and 84.4% were male. Three of the 4 QLQ-OES18 multi-item scales (dysphagia, eating, and pain) and the index scale met the prespecified criterion for acceptable internal consistency as well as acceptable test-retest reliability. Associations between baseline QLQ-OES18 scores and convergent/discriminant validators were generally as expected (i.e., the QLQ-OES18 pain score had a strong positive correlation with the QLQ-C30 pain score). For known-groups validity, 88.6% of analyses demonstrated the hypothesized direction of effect, suggesting that the expected differences in baseline QLQ-OES18 scores between prespecified groups were observed. Ability to detect change analyses indicated that several QLQ-OES18 domain scores demonstrated sensitivity in detecting possible treatment effects, although many patients reported minimal symptoms at baseline, which limited the ability to detect significant improvement.

Conclusion: Overall, a collection of psychometric evidence indicated that the EORTC QLQ-OES18 reliably and validly measured symptom severity in the RATIONALE 302 population. Specifically, the dysphagia domain consistently demonstrated robust psychometric properties. Limitations in data reduced the interpretability of MWPC thresholds and are discussed in detail.