Validity and reliability of the French version of Eating Assessment Tool (EAT-10)

To develop a French version of the Eating Assessment Tool (Fr EAT-10) and to assess its internal consistency, reliability and clinical validity.

Fifty-six patients referred in the Swallowing Clinics of CHU Saint-Pierre Hospital (Brussels) and EpiCURA hospital (Ath, Belgium) for dysphagia were enrolled and completed fiberoptic endoscopic evaluation of swallowing and videofluoroscopy. Seventy-three asymptomatic subjects were included in the study. To assess reliability, Fr-EAT-10 was completed twice within a 7-day period. Validity was assessed by comparing Fr-EAT-10 scores with the scores of dysphagia handicap index (DHI) in all individuals. Normative value of EAT-10 was calculated and the receiver operating characteristic (ROC) curve was used to determine the best Fr-EAT-10 threshold associated with aspiration.

Fifty-two patients completed the study. Cronbach's alpha was 0.95 indicating a high internal consistency. Test-retest reliability was high in the entire cohort (r_s = 0.921). The correlation between Fr-EAT-10 total scores and DHI was high (r_s = 0.827) indicating a high external validity. Patients had a significant higher score of Fr-EAT-10 than the controls (p < 0.001) exhibiting a high internal validity. The analysis of normative data reported that a score of Fr-EAT-10 > 3 should be considered as abnormal. The correlation between Fr-EAT-10 and the occurrence of aspiration is significant (r_s = 0.327, p < 0.05). According to the ROC curve; aspirations need to be highly suspected for patients with Fr-EAT-10 ≥ 17.

The Fr-EAT-10 developed in this study is a reliable and valid self-administered tool in the evaluation of dysphagia in French-speaking patients.