Testing culturally relevant EQ-5D-5L bolt-ons for Chinese patients with lymphoma and multiple myeloma: third quantitative phase of a mixed-methods study

Abstract Objective As part of a large mixed-methods study, this third quantitative phase aimed to test the psychometric performance of seven culturally relevant EQ-5D-5L bolt-ons in Chinese patients with hematologic neoplasms (lymphoma and multiple myeloma) and to provide further evidence for bolt-on selection in China. Methods From November 2024 to February 2025, a prospective survey was conducted at a tertiary hospital in Shanghai. Baseline data were collected from 208 patients using self-completed paper questionnaires, and 155 patients completed the 1–2-week follow-up to assess test-retest reliability. Psychometric analyses included ceiling effects, informativeness, test-retest reliability, convergent and divergent validity, known-group validity, and explanatory power. Results Sleep problems and tiredness/lack of strength showed the highest relative informativeness among bolt-ons. All EQ-5D-5L + bolt-on versions showed good test-retest reliability (ICC > 0.75). Emotional control correlated moderately with anxiety/depression in both groups, and tiredness/lack of strength correlated with pain/discomfort in multiple myeloma patients. Adding tiredness/lack of strength and emotional control improved the explanatory power for EQ VAS scores in both disease groups, as did climate adaptation in multiple myeloma patients. The bolt-ons distinguished age and self-rated health groups but not tumor stage in lymphoma patients. Conclusion Tiredness/lack of strength, sleep problems, and appetite improved the psychometric properties of the EQ-5D-5L in hematologic neoplasm patients, with tiredness/lack of strength performing most consistently. These results agree with our earlier phases in the general population and among caregivers, where tiredness/lack of strength also ranked first. The consistency across three distinct Chinese populations supports further testing of this bolt-on in clinical and research settings. We caution, however, that none of the bolt-ons currently has a value set; their use in formal QALY calculation awaits valuation research.