What minimum level of language comprehension is required for reliable administration of the SAQOL-39NLg?

Abstract

ABSTRACT Background . Quality of life (QOL) is an important outcome in health care, usually captured by self-reported questionnaires. The Stroke Aphasia Quality Of Life Scale (SAQOL-39; SAQOL-39g) was developed by Hilari and colleagues specifically for people with aphasia (PWA), who often struggle with the linguistic complexity of QOL scales. It has since been translated into many languages, including Dutch (SAQOL-39NLg). Some level of (language) comprehension ability is however still required to fill out the SAQOL-39g reliably. So far, little research has been done to evaluate the minimum level of language comprehension needed to fill in the SAQOL-39g. Aim . The aim of this project was to find the minimum level of comprehension needed for PWA to complete the Dutch version of the SAQOL-39g, the SAQOL-39NLg reliably. Methods & procedures . PWA in the chronic phase were asked to fill in the SAQOL-39NLg twice. SLT students supported the PWA during administration. We hypothesized that if PWA comprehend the questions, scores at T1 and T2 should be similar and thus Intraclass Correlation Coefficients would be high. Aphasia severity and language comprehension were assessed at T1 by means of the Token Test, the word and sentence comprehension part of the Comprehensive Aphasia Test- Dutch version (CAT-NL) and the Semantic Association Test (SAT). ICC, Standard Error of Measurement (SEM) and SEM% of the results of the two SAQOL-39NLg administrations were calculated for groups of PWA grouped and ranked based on the results of the language assessments. Outcomes & results . For this study, sixty PWA were included. All of them had some level of comprehension difficulties, the majority moderate to severe. No correlation was found between results on any of the language tests, and difference scores of the two administrations of the SAQOL-39NLg. The ICC between SAQOL-39NLg administrations was above 0.73 when clustered on SAT scores, and above 0.80 for groups clustered based on Token Test and CAT-NL scores. For all repeated measurements, SEM% was between 5% and 10%, and therefore acceptable. Conclusions . We conclude that it is possible for PWA to answer the SAQOL-39NLg reliably, when he or she is able to perform language tests. This finding strongly suggests that PWA should not be excluded from scientific research, when questionnaires are used.