Psychometric Validation of the QLQ-NMIBC24 in Low Grade, Intermediate Risk Non-Muscle Invasive Bladder Cancer

Purpose This study aimed to evaluate the psychometric properties of the EORTC QLQ-NMIBC24 and to propose distribution-based minimal clinically important difference (MCID) thresholds in patients with low-grade, intermediate-risk NMIBC (LG-IR-NMIBC). Patients and Methods Patients with LG-IR-NMIBC from two phase 3 trials (ATLAS, n = 270; ENVISION, n = 240) completed the EORTC QLQ-C30 (ENVISION only) and QLQ-NMIBC24 questionnaires at baseline, Week 6, and Month 3. Psychometric evaluation of the QLQ-NMIBC24 in the LG-IR-NMIBC population included internal consistency, item convergence, known-groups validity, test-retest reliability and responsiveness to clinical change. Results Item-item (0.40–0.96) and item-scale correlations (0.44–0.96) indicated good convergent validity across most domains. Internal consistency was acceptable for all multi-item domains (Cronbach’s α 0.78–0.93) except Malaise, which showed lower reliability (Cronbach’s α 0.34–0.67). Known-groups comparisons supported the instrument’s ability to distinguish patients by physical function (effect sizes up to 1.63) and sex-related domains. Test-retest reliability was generally good for Urinary Symptoms and Sexual Function (ICC 0.79–0.82). Selected domains (Urinary Symptoms, Sexual Intimacy, Malaise) demonstrated responsiveness to clinical change. Distribution-based MCID estimates varied across domains (4.37–16.29), providing potential thresholds for meaningful changes in HRQoL scores. Anchor-based analyses using QLQ-C30 change scores were explored, but correlations with QLQ-NMIBC24 domains (r = –0.37 to 0.01) did not meet the minimum threshold (0.37). Conclusion The QLQ-NMIBC24 demonstrates valid, reliable, and interpretable measurement properties in patients with LG-IR-NMIBC. Although sensitivity varied across domains, the findings support its use in clinical trials and potentially routine practice. Distribution-based MCID thresholds provide guidance for interpreting meaningful HRQoL changes, though further studies using anchor-based methods are warranted.