A total of 154 patients underwent FEES in the study period. The distribution of patients’ age, gender, and comorbidities are shown in Table 2. The average age of participants was 65.9 ± 17.03 years and included 93 males (60.38%). The majority of the cohort (88 patients, 57.1%) had 1–3 comorbidities. A total of 64 patients (41.5%) were diagnosed with neurological comorbidities.

Based on the standard FEES exam, 76 patients (49.35%) were diagnosed with oral, pharyngeal, or oropharyngeal swallowing disorders, and 78 patients (50.64%) had a normal standard FEES exam. The distribution of the patients based on their test results are shown in Fig. 2.

Flow chart of the distribution of the patients in the study

The empty capsule test was not performed in 45 patients (29.2%), as shown in Fig. 2. Of them, nine (23.68%) refused to proceed with the capsule test following the completion of the standard FEES exam. Fourteen patients (36.8%) were found to have severe oropharyngeal dysphagia by the standard FEES exam, and the medical team decided to avoid the capsule test.

Although the standard FEES exam was normal in 15 patients (39.47%), the medical team decided not to proceed with the capsule test due to severe comorbidities or high sensitivity to the endoscope in the lower pharynx. The test was well tolerated among all patients in which the endoscope was inserted into the esophagus, with only mild discomfort.

The distribution of modified FEES results are shown in Table 3. Sixty-three (57.7%) patients demonstrated pathological capsule swallowing in either the oral, pharyngeal, or esophageal phases. Of them, 55 patients (87.3%) showed esophageal swallowing pathology, 28 of which were isolated esophageal dysphagia, and 27 had combined dysphagia (Fig. 2). In Five patients (7.9%), the capsule test showed pharyngeal dysphagia, two of them combined with esophageal dysphagia. All five patients had already been diagnosed using the standard FEES exam. An additional six patients had oral dysphagia, one of them combined with esophageal dysphagia. Of those six patients, three were diagnosed by capsule swallowing only.

Following modified FEES results, patients were advised to proceed with gastroenterology evaluation (GE). Of them, 42 (75%) had completed the GE (Table 4). In 26 patients (61.9%; 26/42), an esophageal disorder such as gastroesophageal reflux, hiatal hernia, eosinophilic esophagitis, esophageal web, or motility disorder was diagnosed by one of the tests. Gastroesophageal pathologies in patients who underwent capsule examination are described in Table 5.

Among the 27 patients with a pathological capsule test, the diagnosis of 23 patients (85.1%) was verified by following GE. Among the 15 patients who had a normal capsule test and were recommended GE, three patients (20%) were found to have an esophageal pathology (Table 4).

Predictive values of modified FEES compared with GE tests are presented in Table 6. Compared with gastroscopy (n = 20), capsule swallowing for esophageal diagnosis had an accuracy of 78%, sensitivity of 90%, specificity of 65%, PPV of 73% and NPV of 86.6%. Compared with manometry (n = 9), capsule swallowing demonstrated a 78% accuracy, 80% sensitivity, 75%specificity, 80% PPV and 75% NPV. Compared with fluoroscopy (n = 7), capsule swallowing showed an accuracy of 91.67%, sensitivity of 100%, specificity of 80%, PPV of 87.5% and NPV of 100%. Compared with all gastroenterological exams, capsule swallowing demonstrated an 83.3% accuracy, 88.5% sensitivity, 75%specificity, 85% PPV, and 80% NPV (Table 6).

The area under the curve was 0.777 for gastroscopy, 0.775 for manometry, 0.900 for fluoroscopy, and 0.817 for total gastroenterological exams (Table 5).