The Pedi-EAT-10 tool is a simple, caregiver - administered questionnaire in order to evaluate the high risk of dysphagia symptoms (penetration and aspiration) in the pediatric population [24, 26]. The original Pedi-EAT-10 was adapted from the adult version of Eating Assessment Tool 10 (Eat − 10) [33]. Even the adult version in Greek is already validated [34]; this is the first study that presents the translation in the Greek language of the pediatric version. According to the literature, the first standardised Pedi-EAT-10 questionnaire was developed in English [24], and it’s the first validation to another language was the Arabic [9]. Consequently, this study is the second in a row validation of Pedi-EAT-10 questionnaire and specifically in Greek language. At this point, it is important to note that there is no current valid in Greek language questionnaire for children with feeding and/or swallowing disorders in Cyprus, and this research was designed to provide the first one and that was the Pedi - EAT − 10.
The results of this study have proven that the translated tool is discriminative, valid, and reliable in a screening process for coping the occurrence of feeding and/or dysphagia symptoms in the Greek – Cypriot pediatric population. Specific hierarchical stages of translation and cultural adaptation were followed to demonstrate equivalence between the Greek, Arabic, and English versions of the Pedi-EAT-10 [9, 35]. Also, high internal consistency, consistent construct validity, and excellent test-retest reliability, which were observed in the results, indicated the accuracy and the equivalence between the Greek version and the existing two once questionnaires [9, 35].
In this study, the mean age for children with feeding and/or swallowing disorders was 6.82 ± 3.01 years, while the mean age for the other group of children without feeding and/or swallowing disorders was 7.51 ± 2.54 years, and 61.7% were males, in the whole sample. The mean age of children with feeding and/or swallowing disorders in Arabic and English studies was 2.48 ± 1.79 years, 52.8% were males [9], and 4.8 ± 1.4–9.7 years, 51% were males [35], respectively. In addition, the mean age for children without feeding and/or swallowing disorders in Arabic and English studies was 2.81 ± 1.79 years and 4.1 ± 1.6–7.8, respectively. Therefore, the present study included a more comprehensive age range and recruited elder children than the other two versions.
This study categorized the group of children with feeding and/or swallowing disorders into four groups, with respect to the medical diagnoses: Acquired Disorders (6.3%), Cerebral palsy (7.2%), Developmental Disorders (25.7%), and Syndromes (15.8%). The Arabic version of Pedi-EAT-10 similarly categorized its sample [9]. It included children with Cerebral palsy (46%), Gastrointestinal causes (25%), Airway problems (18%), Syndromes (6%), and others (5%) [9]. The original English version, Pedi - EAT − 10, focused only on children with Cerebral palsy [35]. In 2017, a study concentrated on the validation of Pedi-EAT-10 to predict aspiration in children with esophageal atresia [26]. Another study completed in 2022 aimed to assess oropharyngeal dysphagia in children with eosinophilic esophagitis using several clinical assessment methods, one of which was the Pedi - EAT – 10 [36].
The current study categorized the children with feeding and/or swallowing disorders into three major subcategories (1. Oral sensory feeding disorder, 2. Oral motor feeding disorder and 3. Oropharyngeal dysphagia). Compared to the above-mentioned studies, this research not only included a wide range of medical diagnoses but also categorized feeding and swallowing difficulties more specifically according to the existing literature [37, 38]. Τhis study is the only one that carried out these further categorizations in the sample and this has the effect of demonstrating greater value of the screening tool for subsequent evaluation.
Furthermore, the construct validity was demonstrated by the Greek version of Pedi-EAT-10, and the characteristics of pharyngeal dysphagia were identified using the PAS scale [26] and the pharyngeal residue score [39]. A significantly positive correlation between Pedi-EAT-10 total and PAS score indicated, with r = 0.45 and p < 0.005. These outcomes are in agreement with the following studies: Pedi - EAT − 10Arabic [9], original Pedi-EAT – 10English [35], Pedi-EAT-10English in children with neurological impairments [25], Pedi-EAT-10 English in children with esophageal atresia [26]. This fact proved that the Pedi-EAT-10 and PAS scale combined together have an excellent potentiality in identifying aspiration and penetration. Moreover, the only study that correlated the Pedi-EAT-10 and pharyngeal residue was from Adel et al [9] and it is in agreement with the results of this study. Also, Duncan et al [40] was noted a similar correlation measure with this study but using different tools. They correlated Pedi - EAT − 10, Infant Gastroesophageal Reflux Questionnaire Revised (I-GERQ-R) and aspiration/penetration symptoms on VFSS and the correlated measures are in agreement with this study (Pedi-EAT-10 and (I-GERQ-R) r = 0.369 and aspiration/penetration r = 0.351) [40].
Also, a significant positive correlation was proved between Pedi-EAT-10 total score and the residue score with r = 0.171 and p < 0.005. This significant positive correlation of the current study was observed in another research between Pedi-EAT-10 and other feeding and/or swallowing scales or tools [25, 36, 41].
Additionally, the internal consistency of the Pedi-EAT-10 was slightly lower than the English version = 0.87 [35] and Arabic version = 0.986 [9], but still excellent (Coefficient alpha = 0.801).
In this study, the item-by-item analysis ranged from 0.767–0.843 for the Pedi - EAT − 10. Τhis result is approximately in agreement with the range that was reported for the Arabic version of Pedi-EAT-10 (0.546–0.94) [9], as well as, with Thoyre’s et al study [41] that indicated a significant moderate to strong correlations between the 10 items.
The test-retest reliability analysis for the Greek version of Pedi-EAT-10 total score (r = 0.998, p < 0.001) and for its 10-items (range: 0.927–0.998) was excellent. These results were also observed in Adel et al [9], Serel Arslan et al [35], and Thoyre et al study [41].
Furthermore, the Greek version of Pedi-EAT-10 demonstrated very good discriminant validity between the children with and without feeding and/or swallowing disorders. Specifically, the same statistically significant positive discrimination was observed with the Arabic version of the questionnaire [9].
The predictive validity for aspiration of the Greek Pedi-EAT-10 total score was calculated at 11.00 with the use of the ROC analysis. This type of analysis was used in other studies, and their results were in approximate agreement with the current study [25, 41]. On contrary, the aforementioned result is not in agreement with the Arabic version of Pedi - EAT − 10, and this probably is attributed to cultural variation [9]. It is also important to note that other studies, which used Pedi-EAT-10 and PAS scale, even they had underlined its predictive validity for aspiration they did not calculate the cut-off values for several reasons [26, 41] as in this study. Thus, the above are indicating that the Greek version of Pedi-EAT-10 indicated is of high sensitivity and specificity in identifying aspiration symptoms.
Regarding the accuracy of Pedi-EAT-10 total score in predicting pharyngeal residue, the ROC analysis demonstrated a cut-off point of 8.00 which is in good agreement with Adel’s [9] study.
Moreover, the total content validity index (CVI) of the Greek version of Pedi-EAT-10 was 0.96 and this result was similar to Pedi-EAT-10 Arabic [9] and almost the same to the English version CVI = 0.91 [35].
It must also not be omitted the fact that this study distinguishes itself from others because it performed more analyses than no other study calculated. Specifically, it used the split-half reliability technique to demonstrate the internal consistency and presented good internal consistency (split-half reliability coefficient = 0.789). Moreover the additional ROC analysis that was performed in this study revealed statistically significant positive discrimination between (a) the PwHC group and the group of parents having children with oral-pharyngeal dysphagia diagnosis, with the cut-off point being 5.00; (b) the PwHC group and group of parents having children with oral-motor feeding disorder diagnosis, with the cut-off point being 4.00 (sensitivity: 0.852 and specificity 0.037); (c) the PwHC group and group of parents having children with oral sensory feeding disorder diagnosis with the cut-off point being 4.00 (sensitivity: 0.745 and specificity 0.037).
4.1 Limitations
The findings of this study have a number of limitations. Firstly, a larger sample of participants is required to generalize to the pediatric dysphagia population in Cyprus. Still, it is essential to note that the sample was clearly categorized according to the type of the medical condition as well as classified into feeding and/or swallowing disorders subgroups. Additional research is also necessary to evaluate the ability to monitor the progress of intervention.