PT  - JOURNAL ARTICLE
AU  - Karl Bang Christensen
AU  - Mikkel Bek Clausen
AU  - Enda King
AU  - Andrew Franklyn-Miller
AU  - Joar Harøy
AU  - Thor Einar Andersen
AU  - Per Hölmich
AU  - Kristian Thorborg
TI  - Validation of the Copenhagen Hip and Groin Outcome Score (HAGOS) using modern test theory across different cultures and languages: a cross-sectional study of 452 male athletes with groin pain
AID  - 10.1136/bjsports-2021-104412
DP  - 2021 Nov 23
TA  - British Journal of Sports Medicine
PG  - bjsports-2021-104412
4099  - http://bjsm.bmj.com/content/early/2021/11/23/bjsports-2021-104412.short
4100  - http://bjsm.bmj.com/content/early/2021/11/23/bjsports-2021-104412.full
AB  - Bkground No studies have tested the validity of the Copenhagen Hip and Groin Outcome Score (HAGOS) using modern test theory, across different cultures and languages.Objective To validate the Danish, English and Norwegian versions of HAGOS and its six subscales (Symptoms (S, Item 1–7), Pain (P, Item 1–10), activities of daily living (Item 1–5), Sport and recreation (Sport/rec, Item 1–8), Participation in physical activity (item 1–2) and quality of life (item 1–5)) by evaluating differential item functioning (DIF) and measurement invariance across the three language versions in male multidirectional team athletes with groin pain. Second, to modify subscales depending on goodness-of-fit to the item response theory models and calculate conversion tables if language DIF was observed.Methods We included individual responses to the Danish (n=157), English (n=146) and Norwegian (n=149) language versions of HAGOS from 452 athletes (median age 24 years old, range 20–28) with groin pain. Overall fit, model fit, individual item fit, local response dependence and measurement invariance was examined using confirmatory factor analysis and graphical Rasch models.Results The removal of seven misfitting items (S2, P1, P2, A4, SP1, SP5, Q3) resulted in 6 HAGOS subscales with acceptable psychometric properties. For the Symptoms, Pain and Sports subscales evidence of DIF was disclosed between the three different language-versions of HAGOS and conversion tables were created.Conclusions A revised HAGOS derived using modern test theory provides valid measurements for male multidirectional athletes with groin pain across different cultures and languages. Conversion tables must be applied to compare HAGOS scores from Danish, Norwegian and English language versions.Data are available on reasonable request. Statistical code and dataset are available from the corresponding author on reasonable request.