We conducted an observational case-control study to validate the Spanish version of the Pedi-EAT-10 and assess its psychometric properties and internal consistency as a diagnostic tool.

The study was conducted in adherence to the most recent revision of the Declaration of Helsinki and current law and regulations in Europe and Spain. We obtained the informed consent of the legal guardians of the participants through the signing of the corresponding form.

The study was approved by the Ethics Committee of the hospital (code CHUC_2021_128).

The Pedi-EAT-10 is an analogical verbal self-report scale that yields a direct score to assess specific symptoms of dysphagia. The validity and reliability of the original scale (in English) have been demonstrated and reported in a previous publication.15 It is a 10-item questionnaire developed by a multidisciplinary team of experts in swallowing disorders. To obtain the score, the parent must provide a subjective rating on a scale from 0 to 4 points, where 0 indicates absence of the problem and 4 presence of the problem in its severest form. The final score is obtained by adding the 10 item scores. A final score of 4 points or greater is indicative of abnormal swallowing.

The validation process of the Spanish version of the Pedi-EAT-10 started with the translation to Spanish of the original version published in English followed by back translation to English by 2 independent teams of bilingual translators. This version was revised by experts in nutrition and speech therapy to obtain and adapted version better fitting the clinical context in Spain. When it had been revised and approved, it was once again back translated to English, and the translation was found to be faithful to the original. The resulting final text can be found in Fig. 1. We obtained permission to translate and adapt the questionnaire from the lead author of the original version (SS Arslan).

Figure 1.

Spanish version of the Pedi-EAT-10 questionnaire.

(0.72MB).

We recruited 91 participants of both sexes aged 6 months to 18 years with a diagnosis of dysphagia cases) and 91 healthy individuals (controls). The necessary sample size was calculated for a power of 80% (with an alpha risk of 5% and a beta risk of 20%) in detecting a statistically significant differences of 9 points in the total score of the test Pedi-EAT-10 between the patients without dysphagia (controls) and the patients with a diagnosis of dysphagia (cases) in a bilateral test.

Participants with dysphagia (case group) were recruited prospectively and consecutively among the children referred to the paediatric nutrition clinic with a history suggestive of oropharyngeal dysphagia, usually with some form of predisposing underlying disease. the diagnosis of dysphagia was confirmed in all of them following the full evaluation by the paediatrician specialised in nutrition and the speech therapist, who carried out an exhaustive assessment including the completion of the Pedi-EAT-10, a focused history taking, questionnaires to assess sialorrhoea and dental hygiene, direct observation of prescribed intake and a volume-viscosity swallow test (V-VST). In addition, 24% underwent a videofluoroscopic swallowing study (VFSS) and 39% a fibreoptic endoscopic evaluation of swallowing (FEES), per the judgment of the clinician.

The rest of patients (control group) were recruited prospectively among children who were healthy or had minor disease that would not manifest with or contribute to dysphagia and with no manifestations compatible with dysphagia from the primary care caseloads of the North Tenerife health care area. Controls were matched for age and sex with cases and recruited in the initial phase.

To assess the construct validity of the Pedi-EAT-10, the questionnaire was completed by the parents or primary caregivers of the child under the supervision of a clinical provider (speech therapist/paediatrician). To pursue homogeneity in the answers supervising providers would give to the questions that participants may bring up during the completion of the questionnaire, precise instructions were drafted and provided to all supervising providers. In addition to providing the data collected in the screening tool (Pedi-EAT-10), participants also filled out an ad hoc questionnaire created to collect basic information such as the age and sex of the child, underlying disease or chronic treatments and information on the time it took to complete the Pedi-EAT-10 and the perceived difficulty in completing it. In the case group, we also recorded the primary diagnosis with the corresponding International Classification of Diseases 11th Revision (ICD-11) code.

All the analyses were made based on the available data, without imputation of missing data, and describing the number of values missing in each analysis. For all tests performed, the level of statistical significance was 5% (α = 0.05). The statistical analysis was performed with the software IBM SPSS version 25.0 (IBM Corp, Armonk, NY, USA).

We conducted a descriptive analysis of all the data entered in the data collection form. Tests were used to assess whether quantitative data followed a normal distribution (Kolmogorov and Shapiro-Wilks) and established that this was the case for all study variables under study. To describe quantitative variables, we used the mean and standard deviation. For categorical variables, we used absolute frequencies and percentages.

To assess whether there were significant differences between the two groups in the scores for each item and the total, we compared the means (Student t-test for independent samples), and to determine whether the age group was significantly associated with the scores, we conducted an analysis of variance (ANOVA) adjusted for multiple comparisons with the Bonferroni correction. To assess whether there were significant differences between group in qualitative variables, we used a test to compare independent proportions (χ2). We assessed the internal consistency of the instrument by means of the Cronbach alpha. We used the Pearson correlation coefficient to measure the correlation between each individual item and the total score. We assessed the sensibility and specificity by means of receiver-operating characteristic (ROC) curves.

The study included a total of 182 participants (91 controls and 91 cases with dysphagia), divided by age group to match participants and obtain comparable groups. However, the Pedi-EAT-10 forms for 4 children with dysphagia were misplaced during the study period, so these 4 cases were excluded from the analysis.

There were no significant differences in age or sex between participants in the two groups, as cases and controls had been matched (Fig. 2).

Figure 2.

Distribution of participants matched for age and sex.

(0.22MB).

In the control group, 18.7% had a controlled underlying disease that did not cause dysphagia a priori.

In the cage group, there was a majority of children with neurologic disease (46.2%, most frequently genetic encephalopathy and disease of unknown aetiology). Other causes of dysphagia were prematurity (7%), pulmonary disease (7%) or gastrointestinal disease (6%). Up to 15.4% of children in the case group had a combination of diseases affecting different systems and 6 patients had no known comorbidities (6%).

The questionnaire was easy to complete and took little time: in the dysphagia case group, 100% of parents completed it in full, leaving no items unanswered and exhibiting no reluctance to provide answers.

In the control group, we were able to assess more objectively the perceived difficulty: 82.4% rated the Pedi-EAT-10 as very easy and 17.6% as easy, and no participant rated it as being moderately difficult, difficult or very difficult. The mean time devoted to completing the questionnaire was 2.18 min (from 1 to 10 min). Thus, we could corroborate that this instrument can also be completed in little time in the Spanish version.

The researchers found no salient biases: tendency to avoid extreme responses, to answer what is perceived as socially acceptable, to choose the same answer in successive questions etc.

In the group of 87 children with dysphagia, 73 had positive results indicative of dysphagia (total score ≥ 4 points), while in the group of 91 healthy controls, only 5 (all aged 6–24 months and without diseases or treatments that could explain the results or findings suggestive of dysphagia in the focused history taking). In every age group, the total score of the questionnaire was significantly different between children with dysphagia and controls (P < .001) (Fig. 3).

Figure 3.

Mean score of the Pedi-EAT-10 by age group in cases and controls.

(0.09MB).

Using these data, we found a high sensitivity and specificity, of 83.9% and 94.5%, respectively, with a positive predictive value of 86% and a negative predictive value of 93.6%. This was reflected in the plotted ROC curve, with an excellent area under the curve of 0.943 (Fig. 4).

Figure 4.

Receiver-operating characteristic curve for the Spanish version of the Pedi-EAT-10.

(0.07MB).

In the assessment of the reliability of the instrument and its internal consistency, we obtained a Cronbach α of 0.865 (since it was within the ideal value range, 0.70−0.90, we ruled out poor consistency, redundancy and duplication). Also, for each of the 10 items, we calculated the Cronbach α that would result if that item were removed, and found values greater than 0.8 in every case (Table 1). This suggests that even if one of the items is left unanswered, the reliability and consistency of the test continue to be high.

Most of the correlations of individual item scores with the total score for the Pedi-EAT-10 were greater than 0.65, which indicates a good correlation between the total score and the answers given to each of the items, all with a P value of less than .001 (Table 1). This confirmed the adequacy of the items that comprise the Pedi-EAT-10.

We analysed each of the items of the test by age group. The different in the score for each individual item comparing the case and control groups was statistically significant in every age group (P < .001).