Abstract
Thyroid cancer survivors experience distinctive, persistent burdens that diminish health-related quality of life (HRQoL). Utilities from patient preference-based measures are needed for quality-adjusted life-year estimation and decision-making. Generic instruments lack thyroid-specific content, limiting applicability in this population. We sought to develop a thyroid-cancer-specific utility measure, Thyroid Cancer Quality of Life Index (TCQOLI) to support clinical research, cost-effectiveness analyses, and policy applications.
We conducted a multicenter, multiphase, mixed-method, cross-sectional study. Phase 1 defined the TCQOLI domains and items using input from multidisciplinary experts and patients with thyroid cancer. Phase 2 evaluated the instrument via cognitive interviews (
= 50) and a mailed/phone-assisted psychometric survey in adults with thyroid cancer (
= 163), followed by confirmatory factor analysis (CFA) and reliability/validity analyses. Phase 3 valued health states in a separate sample (
= 103) using interviewer-administered visual analog scales (VAS; 0-100) and standard gamble (SG). Levels of morbidity in each health domain with VAS and SG were used for assessing preferences for three clinical marker states. We derived a weighted dead-to-full-health lower VAS anchor, estimated a VAS to SG power mapping solution to apply to the model overall, and constructed additive, multiplicative, and unweighted indices. Agreement of the indices with direct VAS was summarized by Pearson
, mean absolute error (MAE), overall standard deviation (OSD) of differences, and intraclass correlation (ICC).
Ten candidate domains were finalized; because one domain, reproduction concern, had the weakest psychometrics and the lowest model weight, the primary instrument uses nine domains (a 10-domain version was also evaluated). CFA supported a general HRQoL factor plus a voice/swallow factor with acceptable composite reliability and model fit. The instrument-level ceiling effect was low (3.8%) with no floor effect. The 9-domain additive multiattribute utility theory index correlated with direct VAS (
≈ 0.74-0.75) and showed the best agreement. MAE/OSD was 0.045/0.095 after SG mapping and a good to excellent ICC of 0.74.
TCQOLI is a psychometrically robust, thyroid-cancer-specific, preference-based measure with patient-anchored valuation, suitable for health-economic evaluations. This concise index supports comparative-effectiveness research in thyroid cancer and informs resource allocation across clinical and policy settings.