Annals of Otology, Rhinology & Laryngology 122(11):717-724. © 2013 Annals Publishing Company. All rights reserved.
Reliability and Validity of the Italian Eating Assessment Tool Antonio Schindler, MD; Francesco Mozzanica, MD; Anna Monzani, SLP; Eleonora Ceriani, SLP; Murat Atac, MD; Nikolina Jukic-Peladic, MD; Claudia Venturini, MD; Paolo Orlandoni, MD Objectives: We sought to evaluate the reliability and validity of the Italian EAT-10 (Italian Eating Assessment Tool; I-EAT-10). Methods: The study consisted of 4 phases: item generation, internal consistency and reliability analysis, normative data generation, and validity analysis. Discussion of the EAT-10 with 30 patients and its back-translation were accomplished. The recruited population included 172 patients (40 with dysphonia and 132 with dysphagia) and 269 asymptomatic subjects for testing of internal consistency, and 94 patients with dysphagia and 158 asymptomatic subjects for test-retest reliability analysis. Normative data were gathered from the 269 subjects. The scores of patients and asymptomatic subjects were compared. The I-EAT-10 and flexible endoscopic evaluation of swallowing (FEES) scores in 94 patients were correlated. The I-EAT-10 scores made before and after successful swallowing rehabilitation in 38 patients were compared. Results: Excellent internal consistency (Cronbach’s alpha values of 0.90 and 0.93) and strong test-retest reliability (intraclass correlation coefficients of 0.95 and 0.98) were found in patients and asymptomatic subjects. The I-EAT-10 mean (±SD) score of the normal cohort was 0.6 ± 1.1. The asymptomatic subjects and dysphonic patients scored lower than the dysphagic patients on the Kruskal-Wallis test (p = 0.001). The I-EAT-10 and FEES scores were mildly correlated. The mean I-EAT-10 score improved from 9.8 ± 10.3 to 5.8 ± 6.7 after swallowing rehabilitation (p = 0.04). Conclusions: The I-EAT-10 is a reliable, valid, symptom-specific outcome tool. Key Words: dysphagia, outcome, reliability, self-assessment, validity.
anxiety and progressive dysphagia was associated with depression.
Introduction
Oral-pharyngeal dysphagia is a common symptom that has high rates of mortality and morbidity, which significantly increase the costs of treatment. Its prevalence in the general population ranges between 16% and 22%.1-5 The prevalence of dysphagia is high in common diseases, exceeds 50% in stroke patients, and may be as high as 84% in patients with Parkinson’s disease.6,7 Moreover, swallowing disorders are diagnosed commonly in elderly and bedridden patients.
Whereas in the past the research focused mainly on swallowing biomechanics and bolus transit analysis, in recent years quality of life (QOL) measures and patient-reported symptom tools are increasingly attracting research attention. The QOL and patient-reported assessment tools play an important role, because the patient’s perspective will influence clinical decision-making. These measures may also be used to systematically document the effectiveness of a given treatment. In addition, QOL assessment and patient-reported symptoms could be used to monitor the longitudinal course of individual patients’ outcomes.11
Besides the general health status implications related to dysphagia,8 swallowing disorders may also lead to social and emotional impairment. Ekberg et al9 reported that 41% of patients with dysphagia complained of panic or anxiety during mealtimes and that 36% avoided eating with others because of swallowing problems. Eslick et al10 reported an independent relationship with emotional disorders in which intermittent dysphagia was associated with
Various instruments have been proposed in order to assess the level of handicap of a patient with a swallowing disorder: the Swallowing Quality of Life Survey (SWAL-QOL), the M. D. Anderson Dysphagia Inventory (MDADI), and the dysphagia
From the L. Sacco Department of Biomedical and Clinical Sciences, University of Milan (Schindler, Mozzanica, Monzani, Ceriani, Atac), Gruppo Italiano Studio Disfagia (GISD; Schindler), and the Società Italiana di Nutrizione Artificiale e Metabolismo (SINPE; Venturini, Orlandoni), Milan, and the Unit of Nutritional Therapy, INRCA-IRCCS (Istituto Nazionale di Riposo e Cura per Anziani–Istituto di Ricovero e Cura a Carattere Scientifico), Ancona (Jukic-Peladic, Venturini, Orlandoni), Italy. Correspondence: Antonio Schindler, MD, Ospedale L. Sacco, Via GB Grassi 74, 20157 Milan, Italy; email: antonio.schindler@unimi. it
717
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
718
Schindler et al, Validation of Italian EAT-10 table 1. study flow diagram
Phase 1: back-translation and item development Phase 2: internal consistency and reliability analysis Phase 3: normative data generation Phase 4: clinical and criterion validity analysis
symptom inventory developed by Wallace et al.12-14 Although all of these instruments have been validated, none has been widely accepted in clinical practice. The SWAL-QOL consists of 43 items, and even though it has good internal consistency, reliability, and clinical validity, it is time-consuming, as it is subdivided into 11 sections; moreover, it may be difficult for a patient to complete, as there are different instructions for the different sections.12 The MDADI consists of 20 questions grouped into 3 domains (emotional, functional, and physical). Because it was developed to assess the dysphagia-related QOL in patients with head and neck cancer, its applicability to other populations is limited.13 Wallace et al14 developed a 17-item questionnaire designed to assess symptom severity in patients with dysphagia. The instrument is based on a visual analog scale, and the maximum possible score is 1,700. The Wallace questionnaire exhibits good clinical validity and reliability, but the visual analog scoring system makes it cumbersome and time-consuming.15 In order to develop a more rapidly administered and easily scorable patient-centered dysphagia outcome tool, Belafsky et al11 developed the Eating Assessment Tool (EAT-10) in 2008. The EAT-10 was designed to assess symptom severity, QOL, and treatment efficacy. It consists of 10 questions, is easily administered, and can be completed in less than 2 minutes. An elevated EAT-10 score indicates a selfperception of a high level of dysphagia severity. The clinician only needs to add up the numbers of each question, and no visual analog scales or formulas are required to calculate a raw score. However, the omission of specific domains or sections prevents the stratification of subsets of disability into different categories. The validity and reliability of the EAT-10 have been demonstrated in a large cohort of patients with a wide variety of causes of dysphagia. To our knowledge, no symptom-specific patientcentered outcome tool exists in the Italian language. The MDADI has been culturally adapted and was validated in Italian,16 but as noted above, the fact that it was designed for oncological dysphagia only limits its application to other populations. The aim of the current study was the cultural adaptation of the EAT-10 for an Italian population and the evaluation
718
of its internal consistency, reliability, and criterion and clinical validity. The importance of this study lies in the fact that a score for symptom severity and QOL provides important information for the clinician and could facilitate both the diagnostic workup and decisions about treatment options. Moreover, an Italian version of the EAT-10 will facilitate national and cross-cultural and cross-country studies. Materials and Methods
The study was carried out according to the Declaration of Helsinki, and it was approved by the Institutional Review Boards of the L. Sacco Hospital in Milan and the INRCA in Ancona. All subjects enrolled in the study gave their written informed consent; all data were collected prospectively. The study consisted of 4 different phases (Table 1): item generation (phase 1), internal consistency and reliability analysis (phase 2), normative data generation (phase 3), and validity analysis (phase 4). Only patients with normal cognitive function (for subjects older than 65 years of age, a Mini Mental State Examination score of more than 24) and those with preserved reading skills were included in the study. Phase 1: I-EAT-10 Generation. The 5-step crosscultural adaptation process of translation and backtranslation as suggested in international guidelines was followed.17 The items of the original EAT-10 questionnaire were first translated into Italian by 2 bilingual phoniatricians experienced in dysphagia management (step 1: forward translation). Discussion of the translated manuscript with 2 Italian speech and language pathologists with extensive experience in dysphagia rehabilitation ensured the unanimity of the interpretation of the translated manuscript (step 2: synthesis). Thirty patients (10 male and 20 female) with a median age of 63.5 years (range, 32 to 68 years) who reported symptoms of dysphagia were enrolled in a pilot study (step 3: pilot study). The causes of dysphagia were head and neck cancer in 7 cases, Parkinson’s disease in 8 cases, stroke in 6 cases, and amyotrophic lateral sclerosis in 9 cases. Each patient autonomously completed the EAT-10 in its first translation and discussed the wording and meaning of each item with the senior clinician (A.S.). The wording of the questionnaire was modified in light of suggestions given by the patients (step 4: expert panel). This new and final version of the Italian EAT-10 (I-EAT-10; Table 2) was than translated back into English by a qualified professional translator (step 5: backward translation). This back-translation was compared to the original manuscript by the professional translator; no items of incongruent translation were noted, as
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
719
Schindler et al, Validation of Italian EAT-10
719
table 2. Italian version of Eating Assessment Tool (I-EAT-10)
Circle the appropriate response. Per ogni affermazione segni la risposta che ritiene più appropriata. To what extent are the following scenarios problematic for you? 0 = No problem 0 = Per nulla 4 = Severe problem 4 = Molto Quanto ritiene problematiche le seguenti situazioni? 1 My swallowing problem has caused me to lose weight. 0 1 2 3 4 Il mio problema di deglutizione mi ha fatto perdere peso. 2 My swallowing problem interferes with my ability to go out for 0 1 2 3 4 meals. Il mio problema di deglutizione interferisce con la possibilità di uscire per andare a mangiare. 3 Swallowing liquids takes extra effort. 0 1 2 3 4 Deglutire liquidi mi richiede uno sforzo extra. 0 1 2 3 4 4 Swallowing solids takes extra effort. Deglutire solidi mi richiede uno sforzo extra. 5 Swallowing pills takes extra effort. 0 1 2 3 4 Deglutire pillole mi richiede uno sforzo extra. 6 Swallowing is painful. 0 1 2 3 4 Deglutire è doloroso. 7 The pleasure of eating is affected by my swallowing. 0 1 2 3 4 Il piacere di mangiare è condizionato dalla mia deglutizione. 8 When I swallow food sticks in my throat. 0 1 2 3 4 Quando deglutisco il cibo rimane appiccicato in gola. 0 1 2 3 4 9 I cough when I eat. Tossisco quando mangio. 10 Swallowing is stressful. 0 1 2 3 4 Deglutire è stressante. Total EAT-10: Punteggio totale:
every item was semantically identical to the original English-language manuscript. Phase 2: Internal Consistency and Reliability Analysis. Clinical data were obtained from 172 consecutive patients seen for swallowing disorders (132 patients) or voice disorders (40 patients). The median age of the participants was 70.4 years (range, 36 to 95 years). The data obtained from this group of patients were also used for clinical validity analysis of phase 4 of the study. Two-hundred sixty-nine asymptomatic control subjects (62 male and 96 female) with no symptoms or signs of dysphagia and with no medical history of voice, swallowing, reflux, airway, neurologic, rheumatologic, hematologic, or neoplastic disorders were included in the study for internal consistency analysis. Their median age was 63.8 years (range, 18 to 90 years). The data obtained from this group of patients were also used for phase 3 and for clinical validity analysis of phase 4 of the study. The internal consistency of the I-EAT-10 was assessed with Cronbach’s alpha coefficient. Values greater than 0.7 but less than 0.8 were considered to show satisfactory consistency, those greater than 0.8 but less than 0.9 were considered to show good consistency, and those greater
than 0.9 were considered to show excellent consistency. For this analysis the I-EAT-10 scores obtained in the group of 172 patients and the 269 asymptomatic subjects were used. Of the 2 groups of patients and asymptomatic subjects recruited for internal consistency analysis, 94 patients with a mean age of 67.7 years (range, 36 to 90 years) affected by dysphagia and 158 asymptomatic subjects with a mean age of 70.1 years (range, 36 to 90 years) were randomly selected for reliability analysis. For this purpose the Italian I-EAT-10 was distributed and was completed twice by patients enrolled in the study. Each patient with dysphagia was asked to visit the clinic 2 weeks after having completed the I-EAT-10 the first time; each asymptomatic subject was contacted 2 weeks after the first meeting with the investigators. A variance of 2 days before or after the requested 2 weeks was accepted in case of a patient’s or subject’s need. The 2-week interval was selected because no substantial change in the swallowing process was expected to take place within this period. No access to their first questionnaire responses was allowed to subjects when they were completing the I-EAT-10 the second time. The test-retest reliability was assessed through a Pearson
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
720
720
Schindler et al, Validation of Italian EAT-10 Table 3. clinical and demographic characteristics of study populations
Age (y) Study Population Type of Study Disease Median Range
132 patients with Internal consistency; clinical oropharyngeal dysphagia validity 40 patients with voice Internal consistency; clinical validity disorders 94 patients with Criterion validity (correlation oropharyngeal dysphagia between I-EAT-10 and FEES scores) 38 patients with Criterion validity (comparison oropharyngeal dysphagia of I-EAT-10 scores pre/post successful swallowing rehabilitation)
Stroke (n = 55) Head and neck cancer (n = 24) Parkinson’s disease (n = 18) Amyotrophic lateral sclerosis (n = 17) Traumatic brain injury (n = 23) Zenker’s diverticulum (n = 5) Vocal fold nodules (n = 13) Vocal fold polyp (n = 11) Vocal fold cyst (n = 7) Reinke’s edema (n = 4) Muscle tension dysphonia (n = 5) Hemispheric stroke (n = 40) Brain stem stroke (n = 7) Posterior fossa neurosurgery (n = 5) Head and neck cancer (n = 13) Parkinson’s disease (n = 15) Amyotrophic lateral sclerosis (n = 11) Myasthenia gravis (n = 3) Hemispheric stroke
Sex (M:F)
70.8
36-95
62:70
69.1
37-88
12:28
67.7
36-90
40:54
77.1
54-95
18:20
I-EAT-10 — Italian Eating Assessment Tool; FEES — flexible endoscopic evaluation of swallowing.
test and the intraclass correlation coefficient (ICC), both for the total score and for the scores of single questions included in the I-EAT-10. The correlation was considered strong for values greater than 0.5, moderate for values ranging between 0.3 and 0.5, and weak for values less than 0.3. Phase 3: Normative Data. The 269 asymptomatic control subjects recruited for the internal consistency analysis were included in the study performed to establish normative data. Somewhat as in the original study of Belafsky et al,11 the upper limit of normal for the I-EAT-10 was settled by adding to the mean value of the total symptom survey score 2 standard deviations of the normal cohort. Phase 4: Clinical Validity. For the I-EAT-10 clinical validity assessment, the total I-EAT-10 scores, as well as the scores of single questions for the 132 patients with dysphagia and the 40 patients with dysphonia recruited for the internal consistency analysis, were compared with the normative data obtained from the 269 asymptomatic individuals. For criterion validity assessment, 94 patients who had dysphagia of various origins and severities and were different from the patients recruited for the previous phases underwent flexible endoscopic evaluation of swallowing (FEES). A Storz FNL-10RP2 fiberscope (Karl Storz, Tuttlingen, Germany) was
used for the FEES, in which 3 trials of 5 mL of semisolids, solids, and a cookie were used and the worst score was considered for calculation. To assess dysphagia severity after the FEES examination, we used the Pooling Score (PS),18 the Penetration-Aspiration Scale (PAS),19 and the Dysphagia Outcome and Severity Scale (DOSS).20 The clinician involved in the FEES examination and the PS, PAS, and DOSS scoring was blinded to the I-EAT-10 scores in order to ensure the stability and accuracy of the data. We analyzed the correlation between the I-EAT-10 score and the DOSS, PS, and PAS scores after the FEES evaluation. In order to further assess the criterion-related validity of the I-EAT-10, we recruited a novel cohort of 38 patients who successfully underwent swallowing rehabilitation. All 38 patients had dysphagia due to a hemispheric stroke that had occurred 2 to 3 weeks earlier. All of them underwent daily swallowing rehabilitation. Each patient underwent a bedside clinical examination with an experienced speechlanguage pathologist before and after swallowing rehabilitation. The American Speech LanguageHearing Association National Outcome Measurement System (ASHA NOMS) was used to establish the severity of dysphagia. The clinician involved in swallowing rehabilitation, bedside examination, and ASHA NOMS scoring was blinded to the I-
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
721
721
Schindler et al, Validation of Italian EAT-10 Table 4. Test-retest reliability in 94 patients with dysphagia and 158 control subjects selected for test-retest reliability analysis
Test* Retest* I-EAT-10
Question
Patients
Controls
Patients
Patients
Controls
1 1.6 ± 1.6 (0-4) 0.1 ± 0.1 (0-1) 1.5 ± 1.5 (0-4) 0.1 ± 0.2 (0-2) 2 2.1 ± 1.8 (0-4) 0.2 ± 0.6 (0-3) 2.1 ± 1.7 (0-4) 0.2 ± 0.5 (0-3) 3 2.1 ± 1.4 (0-4) 0.1 ± 0.4 (0-2) 1.9 ± 1.4 (0-4) 0.1 ± 0.4 (0-2) 4 2.1 ± 1.5 (0-4) 0.2 ± 0.5 (0-3) 2.1 ± 1.4 (0-4) 0.2 ± 0.5 (0-3) 5 1.9 ± 1.4 (0-4) 0.2 ± 0.4 (0-2) 2.0 ± 1.5 (0-4) 0.2 ± 0.5 (0-2) 6 0.4 ± 0.9 (0-4) 0.3 ± 0.6 (0-3) 0.3 ± 0.8 (0-4) 0.3 ± 0.7 (0-3) 7 1.8 ± 1.5 (0-4) 0.1 ± 0.2 (0-1) 1.7 ± 1.6 (0-4) 0.1 ± 0.1 (0-1) 8 1.9 ± 1.3 (0-4) 0.1 ± 0.1 (0-1) 1.9 ± 1.2 (0-4) 0.1 ± 0.3 (0-3) 9 1.9 ± 1.3 (0-4) 0.1 ± 0.3 (0-2) 1.8 ± 1.3 (0-4) 0.1 ± 0.2 (0-1) 10 1.1 ± 1.5 (0-4) 0.1 ± 0.2 (0-1) 1.1 ± 1.4 (0-4) 0.1 ± 0.2 (0-2) Total score 17.1 ± 9.1 (2-38) 0.9 ± 1.7 (0-9) 16.5 ± 9.1 (2-39) 0.9 ± 1.8 (0-9) ICC — intraclass correlation coefficient; CI — confidence interval. *Mean ± SD with range in parentheses. †p < 0.001.
EAT-10 scores in order to ensure the stability and accuracy of the data. Each patient completed the IEAT-10 before and after swallowing rehabilitation. Statistical Analysis. Statistical tests were performed with SPSS 18.0 statistical software (SPSS, Inc, Chicago, Illinois). The internal consistency was assessed with Cronbach’s alpha coefficient. Pearson product-moment correlation and ICC were used to evaluate the test-retest reliability of the I-EAT-10 by comparing the baseline and retesting responses. For the comparison of the I-EAT-10 scores of patients with dysphonia or dysphagia and the scores of the asymptomatic subjects, the nonparametric KruskalWallis test was used, and the Mann-Whitney test was used for post hoc analysis. Bonferroni corrections were performed, and a more stringent alpha level for each comparison was set (p = 0.016) in order to control for the increased risks of type 1 errors due to the large number of comparisons made with the Mann-Whitney test. Pearson product-moment correlation was used to assess the correlations between the I-EAT-10 scores and the PS, PAS, and DOSS scores and for the age-related correlation. The ASHA-NOMS scores made before and after swallowing rehabilitation were compared through a Table 5. I-EAT-10 scores of ASYMPTOMATIC CONTROL SUBJECTS
Age (y)
80 Total
No.
Sex (M:F)
8 1:7 26 12:14 37 18:19 177 73:104 21 8:13 269 112:157
*Mean ± SD with range in parentheses.
I-EAT-10 Score* 0.1 ± 0.1 (0-1) 0.3 ± 0.8 (0-4) 0.3 ± 0.8 (0-4) 0.7 ± 1.1 (0-9) 0.8 ± 1.0 (0-3) 0.6 ± 1.1 (0-9)
Pearson Correlation 0.88† 0.94† 0.90† 0.87† 0.87† 0.75† 0.80† 0.77† 0.85† 0.82† 0.95†
ICC (95% CI)
Controls
0.87 (0.82-0.92) 0.94 (0.91-0.96) 0.90 (0.86-0.93) 0.87 (0.81-0.91) 0.87 (0.81-0.91) 0.74 (0.63-0.82) 0.80 (0.72-0.86) 0.77 (0.67-0.84) 0.85 (0.78-0.90) 0.82 (0.75-0.88) 0.95 (0.93-0.97)
Pearson Correlation 0.98† 0.96† 0.99† 0.91† 0.94† 0.98† 0.97† 0.99† 0.96† 0.99† 0.98†
ICC (95% CI)
0.97 (0.95-0.98) 0.96 (0.94-0.97) 0.98 (0.97-0.99) 0.91 (0.87-0.94) 0.92 (0.88-0.93) 0.95 (0.92-0.96) 0.97 (0.94-0.98) 0.98 (0.97-0.99) 0.97 (0.96-0.99) 0.98 (0.97-0.99) 0.98 (0.97-0.99)
Wilcoxon test. The distribution of gender-related IEAT-10 scores, as well as the scores obtained before and after treatment, were compared by use of the Mann-Whitney test. Results
Data on age, sex, and causes of dysphagia and voice disorders of the patients recruited in the different phases of the study are reported in Table 3. All of the patients and control subjects included in the study managed to complete the I-EAT-10 without any assistance. The time required to complete the questionnaire never exceeded 4 minutes. Internal Consistency and Reliability. The Cronbach’s alpha coefficients for the I-EAT-10 total score for both the 172 patients and the 269 asymptomatic subjects were excellent: respectively, 0.90 and 0.93. The I-EAT-10 scores obtained for the test-retest reliTable 6. I-EAT-10 scores of dysphagic and dysphonic patients and control subjects
I-EAT-10 Question
Dysphagic Patients (n = 132)
1 1.4 ± 1.6 (0-4) 2 1.9 ± 1.7 (0-4) 3 1.9 ± 1.4 (0-4) 4 1.8 ± 1.5 (0-4) 5 1.7 ± 1.5 (0-4) 6 0.3 ± 0.8 (0-4) 7 1.5 ± 1.5 (0-4) 8 1.7 ± 1.4 (0-4) 9 1.7 ± 1.3 (0-4) 10 1.1 ± 1.4 (0-4) Total score 14.9 ± 9.9 (0-38)
Dysphonic Patients (n = 40)
0.1 ± 0.3 (0-2) 0.1 ± 0.4 (0-3) 0.2 ± 0.6 (0-2) 0.2 ± 0.7 (0-2) 0.6 ± 0.8 (0-2) 0.2 ± 0.5 (0-2) 0.1 ± 0.4 (0-1) 0.2 ± 0.6 (0-2) 0.3 ± 0.7 (0-2) 0.1 ± 0.3 (0-1) 1.5 ± 1.7 (0-5)
Data are mean ± SD with range in parentheses.
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
Control Subjects (n = 269)
0.1 ± 0.1 (0-1) 0.1 ± 0.4 (0-3) 0.1 ± 0.3 (0-2) 0.1 ± 0.4 (0-3) 0.2 ± 0.5 (0-4) 0.1 ± 0.5 (0-3) 0.1 ± 0.1 (0-2) 0.1 ± 0.1 (0-2) 0.1 ± 0.2 (0-2) 0.1 ± 0.1 (0-1) 0.6 ± 1.1 (0-9)
722
Table 7. Results of Mann-Whitney comparisonS AMONG control group and TWO different groups of patients
I-EAT-10 Question
1 2 3 4 5 6 7 8 9 10 Total score
722
Schindler et al, Validation of Italian EAT-10
Dysphagic vs Dysphonic 0.001* 0.001* 0.001* 0.001* 0.001* 0.009* 0.001* 0.001* 0.001* 0.001* 0.001*
Dysphagic vs Control 0.001* 0.001* 0.001* 0.001* 0.001* 0.043 0.001* 0.001* 0.001* 0.001* 0.001*
Dysphonic vs Control 0.989 0.074 0.282 0.463 0.001* 0.050 0.700 0.001* 0.001* 0.898 0.001*
Data are p values. *Statistically significant. (Because of Bonferroni correction, signifi cance level was set at p = 0.016.)
ability analysis in both the patients and asymptomatic subjects are reported in Table 4; both measures of central tendency (mean, standard deviation, and range) and of correlational analysis (Pearson test and ICC with 95% confidence interval) are reported. A minor decrease of the mean I-EAT-10 score in the retest condition was visible in both groups; however, the Pearson correlation and ICC scores were very high for each group (r = 0.95 in patients and r = 0.98 in asymptomatic subjects; ICC, 0.95% and 0.98%, respectively). Normative Data. The mean age of the 269 asymptomatic subjects was 63.8 ± 17.2 years (range, 18 to 90 years). Forty-two percent of the subjects were male. The mean I-EAT-10 scores corresponding to different age categories of men and women are reported in Table 5. The mean I-EAT-10 score for the normal cohort was 0.6 ± 1.1. The mean plus 2 standard deviations yielded an upper limit of normal for the I-EAT-10 score of 2.8. Clinical and Criterion Validity. For clinical validity analysis, the mean scores obtained from the dysphagic and dysphonic patients and from asymptomatic subjects are reported in Table 6. These data show consistently lower values of total and singleitem scores for asymptomatic subjects. The dysphagic patients scored higher on both the I-EAT-10 total score and the single questions included in the questionnaire. A nonparametric Kruskal-Wallis analysis of variance for the total score and for the questions included in the I-EAT-10 revealed a significant main effect for group (p < 0.001). The results of the Mann-Whitney comparison between the asymptomatic subjects and the two different groups of patients
Table 8. PAS, PS, and DOSS scores and their correlation with I-EAT-10 total score in 94 patients who underwent FEES
Test Texture
PAS Liquids Semisolids Solids PS Liquids Semisolids Solids DOSS
Mean Score ± SD (Range) 3.1 ± 2.3 (1-7) 2.1 ± 1.9 (1-7) 1.5 ± 1.5 (1-7) 4.6 ± 1.1 (4-8) 5.1 ± 1.7 (4-11) 4.8 ± 1.6 (4-11) 3.8 ± 1.4 (1-6)
I-EAT-10 Correlation 0.132 0.229* –0.083 0.312* 0.227* 0.068* –0.487*
PAS — Penetration-Aspiration Scale; PS — Pooling Score; DOSS — Dysphagia Outcome and Severity Scale. *p < 0.05.
are reported in Table 7. A significant difference was found between the dysphagic group of patients and the control and dysphonic groups. No differences were found between the dysphonic patients and the control group for questions 1, 2, 3, 4, 6, 7, and 10. In the group of dysphagic patients, age was not significantly correlated to the overall I-EAT-10 score or to any of the questions. The mean total I-EAT-10 score for men was 12.4 ± 10.2, and that for the women was 11.3 ± 10.7; the difference was not statistically significant (p = 0.76). For criterion validity, the mean PAS and PS scores for liquids, semisolids, and solids, the mean DOSS scores, and the correlation with the I-EAT-10 scores in the group of 94 patients with dysphagia of different origins and severities who underwent FEES are reported in Table 8. The I-EAT-10 total score was positively correlated with the semisolid PAS score (r = 0.229) and with the scores obtained on PS evaluation for liquids, semisolids, and solids (r = 0.312, r = 0.227, and r = 0.068, respectively). The DOSS score was inversely correlated with the I-EAT-10 total score (r = –0.487).
In the group of 38 patients with post-stroke dysphagia, the ASHA NOMS score in the pretreatment condition was 3.3 ± 0.9 and the post-treatment score was 5.4 ± 1.1. The difference was statistically significant (p = 0.01), suggesting a positive evolution of dysphagia. The mean I-EAT-10 score obtained in the pretreatment condition was 9.8 ± 10.3, and the post-treatment score was 5.8 ± 6.7; the difference was statistically significant (p = 0.04). Discussion
In this study, the psychometric properties of the IEAT-10 were studied in a group of 172 consecutive patients seen for swallowing or voice disorders and in a control group of 269 subjects. The fact that all questionnaires were completely filled out suggests
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
723
Schindler et al, Validation of Italian Eat-10
that all subjects actually understood the questions well and felt comfortable answering them. It might therefore be speculated that the I-EAT-10 is not a burdensome instrument, being easily self-administered and requiring no more than 4 minutes to complete. However, the population selected for the study included only patients and subjects with normative cognitive functions and preserved reading skills; it is not known whether subjects who do not satisfy these criteria could complete the I-EAT-10 as easily. The internal consistency of the I-EAT-10 appeared high, with overall Cronbach’s alpha coefficient values of 0.90 and 0.93, respectively, in patients and asymptomatic subjects. Belafsky et al11 reported an overall Cronbach’s alpha coefficient of 0.96. As for the reliability of the I-EAT-10, the scores obtained in the test-retest analysis for both the patients and the control group support the idea that the I-EAT-10 has a high stability and reproducibility over time. In fact, the ICC scores ranged from 0.74% to 0.98%, and the Pearson correlation scores for both the patient and control groups ranged from 0.75 to 0.99 — values that are considered optimal for both group comparison and individual measurements over time. In the original study, the test-retest Pearson correlation coefficients ranged from 0.72 to 0.91.11
As far as the normative data are concerned, this study confirms the results of the original one.11 In the whole studied population, the mean plus 2 SD was 2.8, and even though younger patients scored lower values, in all the different age groups of our study the mean plus 2 SD never exceeded 3 — the value that was considered as abnormal in the original study. As in the original study,11 patients with dysphagia
723
had higher scores on the I-EAT-10 than did patients with dysphonia or healthy subjects. It is possible to speculate that the I-EAT-10 may be a sensitive tool for identifying patients with swallowing disorders. Moreover, the significant difference between the IEAT-10 scores obtained before and after swallowing treatment suggests that the I-EAT-10 may be useful in the initial documentation of dysphagia severity and in the monitoring of the treatment response. A mild but significant correlation between the I-EAT-10 scores and some of the parameters used to assess dysphagia severity after FEES examination was found. This finding is not surprising, since the I-EAT-10 and the scores for penetration, aspiration, and pooling assess different dimensions of dysphagia. The observed modest correlations suggest that self-reported measures and clinician-driven bolus flow measures provide distinct yet complementary information on dysphagia, as has already been reported in previous research on the correlation between the SWAL-QOL and videofluoroscopic examination.21
In conclusion, the I-EAT-10 is a reliable and valid self-administered, symptom-specific outcome tool for dysphagia in adult Italian patients. The application of the I-EAT-10 in Italian patients with dysphagia of any origin is recommended in everyday clinical practice, as well as in epidemiological, efficacy, and outcomes research. Finally, even though this study represents the first validation of the original English EAT-10 in another language, the optimal results of internal consistency, test-retest reliability, and clinical validity seem promising and could allow for cross-cultural research in swallowing disorders.
References
1. Lindgren S, Janzon L. Prevalence of swallowing complaints and clinical findings among 50-79-year-old men and women in an urban population. Dysphagia 1991;6:187-92.
2. Talley NJ, Weaver AL, Zinsmeister AR, Melton LJ III. Onset and disappearance of gastrointestinal symptoms and functional gastrointestinal disorders. Am J Epidemiol 1992;136: 165-77.
3. Kjellén G, Tibbling L. Manometric oesophageal function, acid perfusion test and symptomatology in a 55-year-old general population. Clin Physiol 1981;1:405-15.
4. Bloem BR, Lagaay AM, van Beek W, Haan J, Roos RAC, Wintzen AR. Prevalence of subjective dysphagia in community residents aged over 87. BMJ 1990;300:721-2.
5. Wilkins T, Gillies RA, Thomas AM, Wagner PJ. The prevalence of dysphagia in primary care patients: a HamesNet Research Network study. J Am Board Fam Med 2007;20:14450. 6. Mann G, Hankey GJ, Cameron D. Swallowing disorders following acute stroke: prevalence and diagnostic accuracy.
Cerebrovasc Dis 2000;10:380-6.
7. Coates C, Bakheit AM. Dysphagia in Parkinson’s disease. Eur Neurol 1997;38:49-52. 8. Smithard DG, O’Neill PA, Park C, Morris J. Complications and outcome after acute stroke. Does dysphagia matter? Stroke 1996;27:1200-4. 9. Ekberg O, Hamdy S, Woisard V, Wuttge-Hannig A, Ortega P. Social and psychological burden of dysphagia: its impact on diagnosis and treatment. Dysphagia 2002;17:139-46. 10. Eslick GD, Jones MP, Talley NJ. Non-cardiac chest pain: prevalence, risk factors, impact and consulting — a populationbased study. Aliment Pharmacol Ther 2003;17:1115-24. 11. Belafsky PC, Mouadeb DA, Rees CJ, et al. Validity and reliability of the Eating Assessment Tool (EAT-10). Ann Otol Rhinol Laryngol 2008;117:919-24. 12. McHorney CA, Bricker DE, Robbins J, Kramer AE, Rosenbek JC, Chignell KA. The SWAL-QOL outcomes tool for oropharyngeal dysphagia in adults: II. Item reduction and preliminary scaling. Dysphagia 2000;15:122-33.
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
724
Schindler et al, Validation of Italian Eat-10
13. Chen AY, Frankowski R, Bishop-Leone J, et al. The development and validation of a dysphagia-specific quality-of-life questionnaire for patients with head and neck cancer: the M. D. Anderson dysphagia inventory. Arch Otolaryngol Head Neck Surg 2001;127:870-6.
14. Wallace KL, Middleton S, Cook IJ. Development and validation of a self-report symptom inventory to assess the severity of oral-pharyngeal dysphagia. Gastroenterology 2000;118:67887. 15. McHorney CA, Robbins J, Lomax K, et al. The SWALQOL and SWAL-CARE outcomes tool for oropharyngeal dysphagia in adults: III. Documentation of reliability and validity. Dysphagia 2002;17:97-114.
16. Schindler A, Borghi E, Tiddia C, Ginocchio D, Felisati G, Ottaviani F. Adaptation and validation of the Italian MD Anderson Dysphagia Inventory (MDADI). Rev Laryngol Otol
724
Rhinol (Bord) 2008;129:97-100. 17. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines. J Clin Epidemiol 1993;46:141732. 18. Farneti D. Pooling score: an endoscopic model for evaluating severity of dysphagia. Acta Otorhinolaryngol Ital 2008; 28:135-40. 19. Rosenbek JC, Robbins JA, Roecker EB, Coyle JL, Wood JL. A penetration-aspiration scale. Dysphagia 1996;11:93-8. 20. O’Neil KH, Purdy M, Falk J, Gallo L. The Dysphagia Outcome and Severity Scale. Dysphagia 1999;14:139-45. 21. McHorney CA, Martin-Harris B, Robbins J, Rosenbek J. Clinical validity of the SWAL-QOL and SWAL-CARE outcome tools with respect to bolus flow measures. Dysphagia 2006;21:141-8. [Erratum in Dysphagia 2008;23:461.]
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
Reliability and Validity of the Italian Eating Assessment Tool Antonio Schindler, MD; Francesco Mozzanica, MD; Anna Monzani, SLP; Eleonora Ceriani, SLP; Murat Atac, MD; Nikolina Jukic-Peladic, MD; Claudia Venturini, MD; Paolo Orlandoni, MD Objectives: We sought to evaluate the reliability and validity of the Italian EAT-10 (Italian Eating Assessment Tool; I-EAT-10). Methods: The study consisted of 4 phases: item generation, internal consistency and reliability analysis, normative data generation, and validity analysis. Discussion of the EAT-10 with 30 patients and its back-translation were accomplished. The recruited population included 172 patients (40 with dysphonia and 132 with dysphagia) and 269 asymptomatic subjects for testing of internal consistency, and 94 patients with dysphagia and 158 asymptomatic subjects for test-retest reliability analysis. Normative data were gathered from the 269 subjects. The scores of patients and asymptomatic subjects were compared. The I-EAT-10 and flexible endoscopic evaluation of swallowing (FEES) scores in 94 patients were correlated. The I-EAT-10 scores made before and after successful swallowing rehabilitation in 38 patients were compared. Results: Excellent internal consistency (Cronbach’s alpha values of 0.90 and 0.93) and strong test-retest reliability (intraclass correlation coefficients of 0.95 and 0.98) were found in patients and asymptomatic subjects. The I-EAT-10 mean (±SD) score of the normal cohort was 0.6 ± 1.1. The asymptomatic subjects and dysphonic patients scored lower than the dysphagic patients on the Kruskal-Wallis test (p = 0.001). The I-EAT-10 and FEES scores were mildly correlated. The mean I-EAT-10 score improved from 9.8 ± 10.3 to 5.8 ± 6.7 after swallowing rehabilitation (p = 0.04). Conclusions: The I-EAT-10 is a reliable, valid, symptom-specific outcome tool. Key Words: dysphagia, outcome, reliability, self-assessment, validity.
anxiety and progressive dysphagia was associated with depression.
Introduction
Oral-pharyngeal dysphagia is a common symptom that has high rates of mortality and morbidity, which significantly increase the costs of treatment. Its prevalence in the general population ranges between 16% and 22%.1-5 The prevalence of dysphagia is high in common diseases, exceeds 50% in stroke patients, and may be as high as 84% in patients with Parkinson’s disease.6,7 Moreover, swallowing disorders are diagnosed commonly in elderly and bedridden patients.
Whereas in the past the research focused mainly on swallowing biomechanics and bolus transit analysis, in recent years quality of life (QOL) measures and patient-reported symptom tools are increasingly attracting research attention. The QOL and patient-reported assessment tools play an important role, because the patient’s perspective will influence clinical decision-making. These measures may also be used to systematically document the effectiveness of a given treatment. In addition, QOL assessment and patient-reported symptoms could be used to monitor the longitudinal course of individual patients’ outcomes.11
Besides the general health status implications related to dysphagia,8 swallowing disorders may also lead to social and emotional impairment. Ekberg et al9 reported that 41% of patients with dysphagia complained of panic or anxiety during mealtimes and that 36% avoided eating with others because of swallowing problems. Eslick et al10 reported an independent relationship with emotional disorders in which intermittent dysphagia was associated with
Various instruments have been proposed in order to assess the level of handicap of a patient with a swallowing disorder: the Swallowing Quality of Life Survey (SWAL-QOL), the M. D. Anderson Dysphagia Inventory (MDADI), and the dysphagia
From the L. Sacco Department of Biomedical and Clinical Sciences, University of Milan (Schindler, Mozzanica, Monzani, Ceriani, Atac), Gruppo Italiano Studio Disfagia (GISD; Schindler), and the Società Italiana di Nutrizione Artificiale e Metabolismo (SINPE; Venturini, Orlandoni), Milan, and the Unit of Nutritional Therapy, INRCA-IRCCS (Istituto Nazionale di Riposo e Cura per Anziani–Istituto di Ricovero e Cura a Carattere Scientifico), Ancona (Jukic-Peladic, Venturini, Orlandoni), Italy. Correspondence: Antonio Schindler, MD, Ospedale L. Sacco, Via GB Grassi 74, 20157 Milan, Italy; email: antonio.schindler@unimi. it
717
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
718
Schindler et al, Validation of Italian EAT-10 table 1. study flow diagram
Phase 1: back-translation and item development Phase 2: internal consistency and reliability analysis Phase 3: normative data generation Phase 4: clinical and criterion validity analysis
symptom inventory developed by Wallace et al.12-14 Although all of these instruments have been validated, none has been widely accepted in clinical practice. The SWAL-QOL consists of 43 items, and even though it has good internal consistency, reliability, and clinical validity, it is time-consuming, as it is subdivided into 11 sections; moreover, it may be difficult for a patient to complete, as there are different instructions for the different sections.12 The MDADI consists of 20 questions grouped into 3 domains (emotional, functional, and physical). Because it was developed to assess the dysphagia-related QOL in patients with head and neck cancer, its applicability to other populations is limited.13 Wallace et al14 developed a 17-item questionnaire designed to assess symptom severity in patients with dysphagia. The instrument is based on a visual analog scale, and the maximum possible score is 1,700. The Wallace questionnaire exhibits good clinical validity and reliability, but the visual analog scoring system makes it cumbersome and time-consuming.15 In order to develop a more rapidly administered and easily scorable patient-centered dysphagia outcome tool, Belafsky et al11 developed the Eating Assessment Tool (EAT-10) in 2008. The EAT-10 was designed to assess symptom severity, QOL, and treatment efficacy. It consists of 10 questions, is easily administered, and can be completed in less than 2 minutes. An elevated EAT-10 score indicates a selfperception of a high level of dysphagia severity. The clinician only needs to add up the numbers of each question, and no visual analog scales or formulas are required to calculate a raw score. However, the omission of specific domains or sections prevents the stratification of subsets of disability into different categories. The validity and reliability of the EAT-10 have been demonstrated in a large cohort of patients with a wide variety of causes of dysphagia. To our knowledge, no symptom-specific patientcentered outcome tool exists in the Italian language. The MDADI has been culturally adapted and was validated in Italian,16 but as noted above, the fact that it was designed for oncological dysphagia only limits its application to other populations. The aim of the current study was the cultural adaptation of the EAT-10 for an Italian population and the evaluation
718
of its internal consistency, reliability, and criterion and clinical validity. The importance of this study lies in the fact that a score for symptom severity and QOL provides important information for the clinician and could facilitate both the diagnostic workup and decisions about treatment options. Moreover, an Italian version of the EAT-10 will facilitate national and cross-cultural and cross-country studies. Materials and Methods
The study was carried out according to the Declaration of Helsinki, and it was approved by the Institutional Review Boards of the L. Sacco Hospital in Milan and the INRCA in Ancona. All subjects enrolled in the study gave their written informed consent; all data were collected prospectively. The study consisted of 4 different phases (Table 1): item generation (phase 1), internal consistency and reliability analysis (phase 2), normative data generation (phase 3), and validity analysis (phase 4). Only patients with normal cognitive function (for subjects older than 65 years of age, a Mini Mental State Examination score of more than 24) and those with preserved reading skills were included in the study. Phase 1: I-EAT-10 Generation. The 5-step crosscultural adaptation process of translation and backtranslation as suggested in international guidelines was followed.17 The items of the original EAT-10 questionnaire were first translated into Italian by 2 bilingual phoniatricians experienced in dysphagia management (step 1: forward translation). Discussion of the translated manuscript with 2 Italian speech and language pathologists with extensive experience in dysphagia rehabilitation ensured the unanimity of the interpretation of the translated manuscript (step 2: synthesis). Thirty patients (10 male and 20 female) with a median age of 63.5 years (range, 32 to 68 years) who reported symptoms of dysphagia were enrolled in a pilot study (step 3: pilot study). The causes of dysphagia were head and neck cancer in 7 cases, Parkinson’s disease in 8 cases, stroke in 6 cases, and amyotrophic lateral sclerosis in 9 cases. Each patient autonomously completed the EAT-10 in its first translation and discussed the wording and meaning of each item with the senior clinician (A.S.). The wording of the questionnaire was modified in light of suggestions given by the patients (step 4: expert panel). This new and final version of the Italian EAT-10 (I-EAT-10; Table 2) was than translated back into English by a qualified professional translator (step 5: backward translation). This back-translation was compared to the original manuscript by the professional translator; no items of incongruent translation were noted, as
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
719
Schindler et al, Validation of Italian EAT-10
719
table 2. Italian version of Eating Assessment Tool (I-EAT-10)
Circle the appropriate response. Per ogni affermazione segni la risposta che ritiene più appropriata. To what extent are the following scenarios problematic for you? 0 = No problem 0 = Per nulla 4 = Severe problem 4 = Molto Quanto ritiene problematiche le seguenti situazioni? 1 My swallowing problem has caused me to lose weight. 0 1 2 3 4 Il mio problema di deglutizione mi ha fatto perdere peso. 2 My swallowing problem interferes with my ability to go out for 0 1 2 3 4 meals. Il mio problema di deglutizione interferisce con la possibilità di uscire per andare a mangiare. 3 Swallowing liquids takes extra effort. 0 1 2 3 4 Deglutire liquidi mi richiede uno sforzo extra. 0 1 2 3 4 4 Swallowing solids takes extra effort. Deglutire solidi mi richiede uno sforzo extra. 5 Swallowing pills takes extra effort. 0 1 2 3 4 Deglutire pillole mi richiede uno sforzo extra. 6 Swallowing is painful. 0 1 2 3 4 Deglutire è doloroso. 7 The pleasure of eating is affected by my swallowing. 0 1 2 3 4 Il piacere di mangiare è condizionato dalla mia deglutizione. 8 When I swallow food sticks in my throat. 0 1 2 3 4 Quando deglutisco il cibo rimane appiccicato in gola. 0 1 2 3 4 9 I cough when I eat. Tossisco quando mangio. 10 Swallowing is stressful. 0 1 2 3 4 Deglutire è stressante. Total EAT-10: Punteggio totale:
every item was semantically identical to the original English-language manuscript. Phase 2: Internal Consistency and Reliability Analysis. Clinical data were obtained from 172 consecutive patients seen for swallowing disorders (132 patients) or voice disorders (40 patients). The median age of the participants was 70.4 years (range, 36 to 95 years). The data obtained from this group of patients were also used for clinical validity analysis of phase 4 of the study. Two-hundred sixty-nine asymptomatic control subjects (62 male and 96 female) with no symptoms or signs of dysphagia and with no medical history of voice, swallowing, reflux, airway, neurologic, rheumatologic, hematologic, or neoplastic disorders were included in the study for internal consistency analysis. Their median age was 63.8 years (range, 18 to 90 years). The data obtained from this group of patients were also used for phase 3 and for clinical validity analysis of phase 4 of the study. The internal consistency of the I-EAT-10 was assessed with Cronbach’s alpha coefficient. Values greater than 0.7 but less than 0.8 were considered to show satisfactory consistency, those greater than 0.8 but less than 0.9 were considered to show good consistency, and those greater
than 0.9 were considered to show excellent consistency. For this analysis the I-EAT-10 scores obtained in the group of 172 patients and the 269 asymptomatic subjects were used. Of the 2 groups of patients and asymptomatic subjects recruited for internal consistency analysis, 94 patients with a mean age of 67.7 years (range, 36 to 90 years) affected by dysphagia and 158 asymptomatic subjects with a mean age of 70.1 years (range, 36 to 90 years) were randomly selected for reliability analysis. For this purpose the Italian I-EAT-10 was distributed and was completed twice by patients enrolled in the study. Each patient with dysphagia was asked to visit the clinic 2 weeks after having completed the I-EAT-10 the first time; each asymptomatic subject was contacted 2 weeks after the first meeting with the investigators. A variance of 2 days before or after the requested 2 weeks was accepted in case of a patient’s or subject’s need. The 2-week interval was selected because no substantial change in the swallowing process was expected to take place within this period. No access to their first questionnaire responses was allowed to subjects when they were completing the I-EAT-10 the second time. The test-retest reliability was assessed through a Pearson
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
720
720
Schindler et al, Validation of Italian EAT-10 Table 3. clinical and demographic characteristics of study populations
Age (y) Study Population Type of Study Disease Median Range
132 patients with Internal consistency; clinical oropharyngeal dysphagia validity 40 patients with voice Internal consistency; clinical validity disorders 94 patients with Criterion validity (correlation oropharyngeal dysphagia between I-EAT-10 and FEES scores) 38 patients with Criterion validity (comparison oropharyngeal dysphagia of I-EAT-10 scores pre/post successful swallowing rehabilitation)
Stroke (n = 55) Head and neck cancer (n = 24) Parkinson’s disease (n = 18) Amyotrophic lateral sclerosis (n = 17) Traumatic brain injury (n = 23) Zenker’s diverticulum (n = 5) Vocal fold nodules (n = 13) Vocal fold polyp (n = 11) Vocal fold cyst (n = 7) Reinke’s edema (n = 4) Muscle tension dysphonia (n = 5) Hemispheric stroke (n = 40) Brain stem stroke (n = 7) Posterior fossa neurosurgery (n = 5) Head and neck cancer (n = 13) Parkinson’s disease (n = 15) Amyotrophic lateral sclerosis (n = 11) Myasthenia gravis (n = 3) Hemispheric stroke
Sex (M:F)
70.8
36-95
62:70
69.1
37-88
12:28
67.7
36-90
40:54
77.1
54-95
18:20
I-EAT-10 — Italian Eating Assessment Tool; FEES — flexible endoscopic evaluation of swallowing.
test and the intraclass correlation coefficient (ICC), both for the total score and for the scores of single questions included in the I-EAT-10. The correlation was considered strong for values greater than 0.5, moderate for values ranging between 0.3 and 0.5, and weak for values less than 0.3. Phase 3: Normative Data. The 269 asymptomatic control subjects recruited for the internal consistency analysis were included in the study performed to establish normative data. Somewhat as in the original study of Belafsky et al,11 the upper limit of normal for the I-EAT-10 was settled by adding to the mean value of the total symptom survey score 2 standard deviations of the normal cohort. Phase 4: Clinical Validity. For the I-EAT-10 clinical validity assessment, the total I-EAT-10 scores, as well as the scores of single questions for the 132 patients with dysphagia and the 40 patients with dysphonia recruited for the internal consistency analysis, were compared with the normative data obtained from the 269 asymptomatic individuals. For criterion validity assessment, 94 patients who had dysphagia of various origins and severities and were different from the patients recruited for the previous phases underwent flexible endoscopic evaluation of swallowing (FEES). A Storz FNL-10RP2 fiberscope (Karl Storz, Tuttlingen, Germany) was
used for the FEES, in which 3 trials of 5 mL of semisolids, solids, and a cookie were used and the worst score was considered for calculation. To assess dysphagia severity after the FEES examination, we used the Pooling Score (PS),18 the Penetration-Aspiration Scale (PAS),19 and the Dysphagia Outcome and Severity Scale (DOSS).20 The clinician involved in the FEES examination and the PS, PAS, and DOSS scoring was blinded to the I-EAT-10 scores in order to ensure the stability and accuracy of the data. We analyzed the correlation between the I-EAT-10 score and the DOSS, PS, and PAS scores after the FEES evaluation. In order to further assess the criterion-related validity of the I-EAT-10, we recruited a novel cohort of 38 patients who successfully underwent swallowing rehabilitation. All 38 patients had dysphagia due to a hemispheric stroke that had occurred 2 to 3 weeks earlier. All of them underwent daily swallowing rehabilitation. Each patient underwent a bedside clinical examination with an experienced speechlanguage pathologist before and after swallowing rehabilitation. The American Speech LanguageHearing Association National Outcome Measurement System (ASHA NOMS) was used to establish the severity of dysphagia. The clinician involved in swallowing rehabilitation, bedside examination, and ASHA NOMS scoring was blinded to the I-
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
721
721
Schindler et al, Validation of Italian EAT-10 Table 4. Test-retest reliability in 94 patients with dysphagia and 158 control subjects selected for test-retest reliability analysis
Test* Retest* I-EAT-10
Question
Patients
Controls
Patients
Patients
Controls
1 1.6 ± 1.6 (0-4) 0.1 ± 0.1 (0-1) 1.5 ± 1.5 (0-4) 0.1 ± 0.2 (0-2) 2 2.1 ± 1.8 (0-4) 0.2 ± 0.6 (0-3) 2.1 ± 1.7 (0-4) 0.2 ± 0.5 (0-3) 3 2.1 ± 1.4 (0-4) 0.1 ± 0.4 (0-2) 1.9 ± 1.4 (0-4) 0.1 ± 0.4 (0-2) 4 2.1 ± 1.5 (0-4) 0.2 ± 0.5 (0-3) 2.1 ± 1.4 (0-4) 0.2 ± 0.5 (0-3) 5 1.9 ± 1.4 (0-4) 0.2 ± 0.4 (0-2) 2.0 ± 1.5 (0-4) 0.2 ± 0.5 (0-2) 6 0.4 ± 0.9 (0-4) 0.3 ± 0.6 (0-3) 0.3 ± 0.8 (0-4) 0.3 ± 0.7 (0-3) 7 1.8 ± 1.5 (0-4) 0.1 ± 0.2 (0-1) 1.7 ± 1.6 (0-4) 0.1 ± 0.1 (0-1) 8 1.9 ± 1.3 (0-4) 0.1 ± 0.1 (0-1) 1.9 ± 1.2 (0-4) 0.1 ± 0.3 (0-3) 9 1.9 ± 1.3 (0-4) 0.1 ± 0.3 (0-2) 1.8 ± 1.3 (0-4) 0.1 ± 0.2 (0-1) 10 1.1 ± 1.5 (0-4) 0.1 ± 0.2 (0-1) 1.1 ± 1.4 (0-4) 0.1 ± 0.2 (0-2) Total score 17.1 ± 9.1 (2-38) 0.9 ± 1.7 (0-9) 16.5 ± 9.1 (2-39) 0.9 ± 1.8 (0-9) ICC — intraclass correlation coefficient; CI — confidence interval. *Mean ± SD with range in parentheses. †p < 0.001.
EAT-10 scores in order to ensure the stability and accuracy of the data. Each patient completed the IEAT-10 before and after swallowing rehabilitation. Statistical Analysis. Statistical tests were performed with SPSS 18.0 statistical software (SPSS, Inc, Chicago, Illinois). The internal consistency was assessed with Cronbach’s alpha coefficient. Pearson product-moment correlation and ICC were used to evaluate the test-retest reliability of the I-EAT-10 by comparing the baseline and retesting responses. For the comparison of the I-EAT-10 scores of patients with dysphonia or dysphagia and the scores of the asymptomatic subjects, the nonparametric KruskalWallis test was used, and the Mann-Whitney test was used for post hoc analysis. Bonferroni corrections were performed, and a more stringent alpha level for each comparison was set (p = 0.016) in order to control for the increased risks of type 1 errors due to the large number of comparisons made with the Mann-Whitney test. Pearson product-moment correlation was used to assess the correlations between the I-EAT-10 scores and the PS, PAS, and DOSS scores and for the age-related correlation. The ASHA-NOMS scores made before and after swallowing rehabilitation were compared through a Table 5. I-EAT-10 scores of ASYMPTOMATIC CONTROL SUBJECTS
Age (y)
80 Total
No.
Sex (M:F)
8 1:7 26 12:14 37 18:19 177 73:104 21 8:13 269 112:157
*Mean ± SD with range in parentheses.
I-EAT-10 Score* 0.1 ± 0.1 (0-1) 0.3 ± 0.8 (0-4) 0.3 ± 0.8 (0-4) 0.7 ± 1.1 (0-9) 0.8 ± 1.0 (0-3) 0.6 ± 1.1 (0-9)
Pearson Correlation 0.88† 0.94† 0.90† 0.87† 0.87† 0.75† 0.80† 0.77† 0.85† 0.82† 0.95†
ICC (95% CI)
Controls
0.87 (0.82-0.92) 0.94 (0.91-0.96) 0.90 (0.86-0.93) 0.87 (0.81-0.91) 0.87 (0.81-0.91) 0.74 (0.63-0.82) 0.80 (0.72-0.86) 0.77 (0.67-0.84) 0.85 (0.78-0.90) 0.82 (0.75-0.88) 0.95 (0.93-0.97)
Pearson Correlation 0.98† 0.96† 0.99† 0.91† 0.94† 0.98† 0.97† 0.99† 0.96† 0.99† 0.98†
ICC (95% CI)
0.97 (0.95-0.98) 0.96 (0.94-0.97) 0.98 (0.97-0.99) 0.91 (0.87-0.94) 0.92 (0.88-0.93) 0.95 (0.92-0.96) 0.97 (0.94-0.98) 0.98 (0.97-0.99) 0.97 (0.96-0.99) 0.98 (0.97-0.99) 0.98 (0.97-0.99)
Wilcoxon test. The distribution of gender-related IEAT-10 scores, as well as the scores obtained before and after treatment, were compared by use of the Mann-Whitney test. Results
Data on age, sex, and causes of dysphagia and voice disorders of the patients recruited in the different phases of the study are reported in Table 3. All of the patients and control subjects included in the study managed to complete the I-EAT-10 without any assistance. The time required to complete the questionnaire never exceeded 4 minutes. Internal Consistency and Reliability. The Cronbach’s alpha coefficients for the I-EAT-10 total score for both the 172 patients and the 269 asymptomatic subjects were excellent: respectively, 0.90 and 0.93. The I-EAT-10 scores obtained for the test-retest reliTable 6. I-EAT-10 scores of dysphagic and dysphonic patients and control subjects
I-EAT-10 Question
Dysphagic Patients (n = 132)
1 1.4 ± 1.6 (0-4) 2 1.9 ± 1.7 (0-4) 3 1.9 ± 1.4 (0-4) 4 1.8 ± 1.5 (0-4) 5 1.7 ± 1.5 (0-4) 6 0.3 ± 0.8 (0-4) 7 1.5 ± 1.5 (0-4) 8 1.7 ± 1.4 (0-4) 9 1.7 ± 1.3 (0-4) 10 1.1 ± 1.4 (0-4) Total score 14.9 ± 9.9 (0-38)
Dysphonic Patients (n = 40)
0.1 ± 0.3 (0-2) 0.1 ± 0.4 (0-3) 0.2 ± 0.6 (0-2) 0.2 ± 0.7 (0-2) 0.6 ± 0.8 (0-2) 0.2 ± 0.5 (0-2) 0.1 ± 0.4 (0-1) 0.2 ± 0.6 (0-2) 0.3 ± 0.7 (0-2) 0.1 ± 0.3 (0-1) 1.5 ± 1.7 (0-5)
Data are mean ± SD with range in parentheses.
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
Control Subjects (n = 269)
0.1 ± 0.1 (0-1) 0.1 ± 0.4 (0-3) 0.1 ± 0.3 (0-2) 0.1 ± 0.4 (0-3) 0.2 ± 0.5 (0-4) 0.1 ± 0.5 (0-3) 0.1 ± 0.1 (0-2) 0.1 ± 0.1 (0-2) 0.1 ± 0.2 (0-2) 0.1 ± 0.1 (0-1) 0.6 ± 1.1 (0-9)
722
Table 7. Results of Mann-Whitney comparisonS AMONG control group and TWO different groups of patients
I-EAT-10 Question
1 2 3 4 5 6 7 8 9 10 Total score
722
Schindler et al, Validation of Italian EAT-10
Dysphagic vs Dysphonic 0.001* 0.001* 0.001* 0.001* 0.001* 0.009* 0.001* 0.001* 0.001* 0.001* 0.001*
Dysphagic vs Control 0.001* 0.001* 0.001* 0.001* 0.001* 0.043 0.001* 0.001* 0.001* 0.001* 0.001*
Dysphonic vs Control 0.989 0.074 0.282 0.463 0.001* 0.050 0.700 0.001* 0.001* 0.898 0.001*
Data are p values. *Statistically significant. (Because of Bonferroni correction, signifi cance level was set at p = 0.016.)
ability analysis in both the patients and asymptomatic subjects are reported in Table 4; both measures of central tendency (mean, standard deviation, and range) and of correlational analysis (Pearson test and ICC with 95% confidence interval) are reported. A minor decrease of the mean I-EAT-10 score in the retest condition was visible in both groups; however, the Pearson correlation and ICC scores were very high for each group (r = 0.95 in patients and r = 0.98 in asymptomatic subjects; ICC, 0.95% and 0.98%, respectively). Normative Data. The mean age of the 269 asymptomatic subjects was 63.8 ± 17.2 years (range, 18 to 90 years). Forty-two percent of the subjects were male. The mean I-EAT-10 scores corresponding to different age categories of men and women are reported in Table 5. The mean I-EAT-10 score for the normal cohort was 0.6 ± 1.1. The mean plus 2 standard deviations yielded an upper limit of normal for the I-EAT-10 score of 2.8. Clinical and Criterion Validity. For clinical validity analysis, the mean scores obtained from the dysphagic and dysphonic patients and from asymptomatic subjects are reported in Table 6. These data show consistently lower values of total and singleitem scores for asymptomatic subjects. The dysphagic patients scored higher on both the I-EAT-10 total score and the single questions included in the questionnaire. A nonparametric Kruskal-Wallis analysis of variance for the total score and for the questions included in the I-EAT-10 revealed a significant main effect for group (p < 0.001). The results of the Mann-Whitney comparison between the asymptomatic subjects and the two different groups of patients
Table 8. PAS, PS, and DOSS scores and their correlation with I-EAT-10 total score in 94 patients who underwent FEES
Test Texture
PAS Liquids Semisolids Solids PS Liquids Semisolids Solids DOSS
Mean Score ± SD (Range) 3.1 ± 2.3 (1-7) 2.1 ± 1.9 (1-7) 1.5 ± 1.5 (1-7) 4.6 ± 1.1 (4-8) 5.1 ± 1.7 (4-11) 4.8 ± 1.6 (4-11) 3.8 ± 1.4 (1-6)
I-EAT-10 Correlation 0.132 0.229* –0.083 0.312* 0.227* 0.068* –0.487*
PAS — Penetration-Aspiration Scale; PS — Pooling Score; DOSS — Dysphagia Outcome and Severity Scale. *p < 0.05.
are reported in Table 7. A significant difference was found between the dysphagic group of patients and the control and dysphonic groups. No differences were found between the dysphonic patients and the control group for questions 1, 2, 3, 4, 6, 7, and 10. In the group of dysphagic patients, age was not significantly correlated to the overall I-EAT-10 score or to any of the questions. The mean total I-EAT-10 score for men was 12.4 ± 10.2, and that for the women was 11.3 ± 10.7; the difference was not statistically significant (p = 0.76). For criterion validity, the mean PAS and PS scores for liquids, semisolids, and solids, the mean DOSS scores, and the correlation with the I-EAT-10 scores in the group of 94 patients with dysphagia of different origins and severities who underwent FEES are reported in Table 8. The I-EAT-10 total score was positively correlated with the semisolid PAS score (r = 0.229) and with the scores obtained on PS evaluation for liquids, semisolids, and solids (r = 0.312, r = 0.227, and r = 0.068, respectively). The DOSS score was inversely correlated with the I-EAT-10 total score (r = –0.487).
In the group of 38 patients with post-stroke dysphagia, the ASHA NOMS score in the pretreatment condition was 3.3 ± 0.9 and the post-treatment score was 5.4 ± 1.1. The difference was statistically significant (p = 0.01), suggesting a positive evolution of dysphagia. The mean I-EAT-10 score obtained in the pretreatment condition was 9.8 ± 10.3, and the post-treatment score was 5.8 ± 6.7; the difference was statistically significant (p = 0.04). Discussion
In this study, the psychometric properties of the IEAT-10 were studied in a group of 172 consecutive patients seen for swallowing or voice disorders and in a control group of 269 subjects. The fact that all questionnaires were completely filled out suggests
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
723
Schindler et al, Validation of Italian Eat-10
that all subjects actually understood the questions well and felt comfortable answering them. It might therefore be speculated that the I-EAT-10 is not a burdensome instrument, being easily self-administered and requiring no more than 4 minutes to complete. However, the population selected for the study included only patients and subjects with normative cognitive functions and preserved reading skills; it is not known whether subjects who do not satisfy these criteria could complete the I-EAT-10 as easily. The internal consistency of the I-EAT-10 appeared high, with overall Cronbach’s alpha coefficient values of 0.90 and 0.93, respectively, in patients and asymptomatic subjects. Belafsky et al11 reported an overall Cronbach’s alpha coefficient of 0.96. As for the reliability of the I-EAT-10, the scores obtained in the test-retest analysis for both the patients and the control group support the idea that the I-EAT-10 has a high stability and reproducibility over time. In fact, the ICC scores ranged from 0.74% to 0.98%, and the Pearson correlation scores for both the patient and control groups ranged from 0.75 to 0.99 — values that are considered optimal for both group comparison and individual measurements over time. In the original study, the test-retest Pearson correlation coefficients ranged from 0.72 to 0.91.11
As far as the normative data are concerned, this study confirms the results of the original one.11 In the whole studied population, the mean plus 2 SD was 2.8, and even though younger patients scored lower values, in all the different age groups of our study the mean plus 2 SD never exceeded 3 — the value that was considered as abnormal in the original study. As in the original study,11 patients with dysphagia
723
had higher scores on the I-EAT-10 than did patients with dysphonia or healthy subjects. It is possible to speculate that the I-EAT-10 may be a sensitive tool for identifying patients with swallowing disorders. Moreover, the significant difference between the IEAT-10 scores obtained before and after swallowing treatment suggests that the I-EAT-10 may be useful in the initial documentation of dysphagia severity and in the monitoring of the treatment response. A mild but significant correlation between the I-EAT-10 scores and some of the parameters used to assess dysphagia severity after FEES examination was found. This finding is not surprising, since the I-EAT-10 and the scores for penetration, aspiration, and pooling assess different dimensions of dysphagia. The observed modest correlations suggest that self-reported measures and clinician-driven bolus flow measures provide distinct yet complementary information on dysphagia, as has already been reported in previous research on the correlation between the SWAL-QOL and videofluoroscopic examination.21
In conclusion, the I-EAT-10 is a reliable and valid self-administered, symptom-specific outcome tool for dysphagia in adult Italian patients. The application of the I-EAT-10 in Italian patients with dysphagia of any origin is recommended in everyday clinical practice, as well as in epidemiological, efficacy, and outcomes research. Finally, even though this study represents the first validation of the original English EAT-10 in another language, the optimal results of internal consistency, test-retest reliability, and clinical validity seem promising and could allow for cross-cultural research in swallowing disorders.
References
1. Lindgren S, Janzon L. Prevalence of swallowing complaints and clinical findings among 50-79-year-old men and women in an urban population. Dysphagia 1991;6:187-92.
2. Talley NJ, Weaver AL, Zinsmeister AR, Melton LJ III. Onset and disappearance of gastrointestinal symptoms and functional gastrointestinal disorders. Am J Epidemiol 1992;136: 165-77.
3. Kjellén G, Tibbling L. Manometric oesophageal function, acid perfusion test and symptomatology in a 55-year-old general population. Clin Physiol 1981;1:405-15.
4. Bloem BR, Lagaay AM, van Beek W, Haan J, Roos RAC, Wintzen AR. Prevalence of subjective dysphagia in community residents aged over 87. BMJ 1990;300:721-2.
5. Wilkins T, Gillies RA, Thomas AM, Wagner PJ. The prevalence of dysphagia in primary care patients: a HamesNet Research Network study. J Am Board Fam Med 2007;20:14450. 6. Mann G, Hankey GJ, Cameron D. Swallowing disorders following acute stroke: prevalence and diagnostic accuracy.
Cerebrovasc Dis 2000;10:380-6.
7. Coates C, Bakheit AM. Dysphagia in Parkinson’s disease. Eur Neurol 1997;38:49-52. 8. Smithard DG, O’Neill PA, Park C, Morris J. Complications and outcome after acute stroke. Does dysphagia matter? Stroke 1996;27:1200-4. 9. Ekberg O, Hamdy S, Woisard V, Wuttge-Hannig A, Ortega P. Social and psychological burden of dysphagia: its impact on diagnosis and treatment. Dysphagia 2002;17:139-46. 10. Eslick GD, Jones MP, Talley NJ. Non-cardiac chest pain: prevalence, risk factors, impact and consulting — a populationbased study. Aliment Pharmacol Ther 2003;17:1115-24. 11. Belafsky PC, Mouadeb DA, Rees CJ, et al. Validity and reliability of the Eating Assessment Tool (EAT-10). Ann Otol Rhinol Laryngol 2008;117:919-24. 12. McHorney CA, Bricker DE, Robbins J, Kramer AE, Rosenbek JC, Chignell KA. The SWAL-QOL outcomes tool for oropharyngeal dysphagia in adults: II. Item reduction and preliminary scaling. Dysphagia 2000;15:122-33.
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
724
Schindler et al, Validation of Italian Eat-10
13. Chen AY, Frankowski R, Bishop-Leone J, et al. The development and validation of a dysphagia-specific quality-of-life questionnaire for patients with head and neck cancer: the M. D. Anderson dysphagia inventory. Arch Otolaryngol Head Neck Surg 2001;127:870-6.
14. Wallace KL, Middleton S, Cook IJ. Development and validation of a self-report symptom inventory to assess the severity of oral-pharyngeal dysphagia. Gastroenterology 2000;118:67887. 15. McHorney CA, Robbins J, Lomax K, et al. The SWALQOL and SWAL-CARE outcomes tool for oropharyngeal dysphagia in adults: III. Documentation of reliability and validity. Dysphagia 2002;17:97-114.
16. Schindler A, Borghi E, Tiddia C, Ginocchio D, Felisati G, Ottaviani F. Adaptation and validation of the Italian MD Anderson Dysphagia Inventory (MDADI). Rev Laryngol Otol
724
Rhinol (Bord) 2008;129:97-100. 17. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines. J Clin Epidemiol 1993;46:141732. 18. Farneti D. Pooling score: an endoscopic model for evaluating severity of dysphagia. Acta Otorhinolaryngol Ital 2008; 28:135-40. 19. Rosenbek JC, Robbins JA, Roecker EB, Coyle JL, Wood JL. A penetration-aspiration scale. Dysphagia 1996;11:93-8. 20. O’Neil KH, Purdy M, Falk J, Gallo L. The Dysphagia Outcome and Severity Scale. Dysphagia 1999;14:139-45. 21. McHorney CA, Martin-Harris B, Robbins J, Rosenbek J. Clinical validity of the SWAL-QOL and SWAL-CARE outcome tools with respect to bolus flow measures. Dysphagia 2006;21:141-8. [Erratum in Dysphagia 2008;23:461.]
Downloaded from aor.sagepub.com at BRIGHAM YOUNG UNIV on March 15, 2015
