Testing patient-reported outcome measurement equivalence in multinational clinical trials: An exemplar using the 12-item Multiple Sclerosis Walking Scale

Abstract

<jats:sec><jats:title>Background</jats:title><jats:p> Although multinational clinical trials frequently use patient-reported outcomes to measure efficacy, measurement equivalence across cultures and languages, a scientific requirement, is rarely tested. Clinically accessible accounts are rare; exemplars are needed. </jats:p></jats:sec><jats:sec><jats:title>Objective</jats:title><jats:p> To develop and test a Turkish version of the Multiple Sclerosis Walking Scale (MSWS-12v2) as a clinical exemplar for examining measurement equivalence. </jats:p></jats:sec><jats:sec><jats:title>Methods</jats:title><jats:p> The MSWS-12v2 Turkish (MSWS-12v2T) was developed using recognised methods for linguistic equivalence. Rasch measurement theory was used to examine measurement performance (multiple tests of targeting, scale performance, and person measurement) and measurement equivalence (differential item functioning). UK data ( n = 3310) were used for comparisons and differential item functioning testing. </jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p> One hundred and twenty-four people from two Turkish centres completed the MSWS-12v2T. Rasch measurement theory evidence supported MSWS-12v2T as reliable (person separation = 0.96) and valid (thresholds ordered; no concerning item misfit, bias, or person misfit). However, four items demonstrated significantly different performance between UK and Turkish samples. These item differences significantly affected scores (person measurements) at the group-level ( p &lt; 0.001). Individual person differences were less pronounced. </jats:p></jats:sec><jats:sec><jats:title>Conclusions</jats:title><jats:p> Linguistic equivalence does not guarantee measurement equivalence; independent testing is required. Rasch measurement theory enables sophisticated and unique examinations of cross-cultural measurement equivalence and we recommend this be tested routinely in pivotal multiple sclerosis clinical trials. </jats:p></jats:sec>