Watt, Torquil, Barbesino, Giuseppe, Bjorner, Jakob, Bonnema, Steen, Bukvic, Branka, Drummond, Russell, Groenvold, Mogens, Hegedüs, Laszlo, Kantzer, Valeska, Lasch, Kathryn, Marcocci, Claudio, Mishra, Anjali, Netea-Maier, Romana, Ekker, Merel, Paunovic, Ivan, Quinn, Terence, Rasmussen, Åse, Russell, Audrey, Sabaretnam, Mayilvaganan, and Smit, Johannes
Background and purpose: Thyroid diseases are common and often affect quality of life (QoL). No cross-culturally validated patient-reported outcome measuring thyroid-related QoL is available. The purpose of the present study was to test the cross-cultural validity of the newly developed thyroid-related patient-reported outcome ThyPRO, using tests for differential item functioning (DIF) according to language version. Methods: The ThyPRO consists of 85 items summarized in 13 multi-item scales and one single item. Scales cover physical and mental symptoms, well-being and function as well as social and daily function and cosmetic concerns. Translation applied standard forward-backward methodology with subsequent cognitive interviews and reviews. Responses ( N = 1,810) to the ThyPRO were collected in seven countries: UK ( n = 166), The Netherlands ( n = 147), Serbia ( n = 150), Italy ( n = 110), India ( n = 148), Denmark ( n = 902) and Sweden ( n = 187). Translated versions were compared pairwise to the English version by examining uniform and nonuniform DIF, i.e., whether patients from different countries respond differently to a particular item, although they have identical level of the concept measured by the item. Analyses were controlled for thyroid diagnosis. DIF was investigated by ordinal logistic regression, testing for both statistical significance and magnitude (Δ R > 0.02). Scale level was estimated by the sum score, after purification. Results: For twelve of the 84 tested items, DIF was identified in more than one language. Eight of these were small, but four were indicative of possible low translatability. Twenty-one instances of DIF in single languages were identified, indicating potential problems with the particular translation. However, only seven were of a magnitude which could affect scale scores, most of which could be explained by sample differences not controlled for. Conclusion: The ThyPRO has good cross-cultural validity with only minor cross-cultural invariance and is recommended for use in international multicenter studies. [ABSTRACT FROM AUTHOR]