Determining minimal important change for the thyroid-related quality of life questionnaire ThyPRO

• Published in: Endocrine Connections Vol. 10; no. 3; pp. 316 - 324
• Main Authors: Nordqvist, Selma Flora, Boesen, Victor Brun, Rasmussen, Åse Krogh, Feldt-Rasmussen, Ulla, Hegedüs, Laszlo, Bonnema, Steen Joop, Cramon, Per Karkov, Watt, Torquil, Groenvold, Mogens, Bjorner, Jakob Bue
• Format: Journal Article
• Language: English
• Published: England Bioscientifica Ltd 01.03.2021; Bioscientifica
• Subjects: interpretability; minimal important change; patient-reported outcomes; thypro; thyroid-related quality of life; ThyPRO; patient-reported outcomes; interpretability; thyroid-related quality of life; minimal important change
• DOI: 10.1530/EC-21-0026
• ISSN: 2049-3614

Objective ThyPRO is the standard thyroid patient-reported outcome (PRO). The change in scores that patients perceive as important remains to be ascertained. The purpose of this study was to determine values for minimal important change (MIC) for ThyPRO. Methods A total of 435 patients treated for benign thyroid diseases completed ThyPRO at baseline and 6 weeks following treatment initiation. At 6 weeks follow-up, patients also completed Global Rating of Change items. For each 0–100 scale, two MIC values were identified: An MIC for groups, using the receiver operating characteristic (ROC) curve method and an MIC for individual patients, using the Reliable Change Index. Results ROC analyses provided group-MIC estimates of 6.3–14.3 (score range 0–100). Evaluation of area under the curve (AUC) supported the robustness for 9 of 14 scales (AUC > 0.7). Reliable Change Index estimates of individual-MIC were 8.0–21.1. For all scales but two, the individual-MIC values were larger than the group-MIC values. Conclusions Interpretability of ThyPRO was improved by the establishment of MIC values, which was 6.3–14.3 for groups and 8.0–21.1 for individuals. Thus, estimates of which changes are clinically relevant, are now available for future studies. We recommend using MIC values found by ROC analyses to evaluate changes in groups of patients, whereas MIC values identified by a dual criterion, including the reliability of changes, should be used for individual patients, for example, to identify individual responders in clinical studies or practice.