Diagnostic Accuracy of Ultrasound for Detecting Gastric or Duodenal Ulcers in Pediatric Patients.

OBJECTIVE Direct ultrasound imaging findings alone have low sensitivity for diagnosing duodenal (65%) and gastric ulcers (40%). This retrospective study evaluated the efficiency of ultrasound in detecting gastric/duodenal ulcers in pediatric patients through direct and indirect findings. METHODS We evaluated 244 children who underwent ultrasound and subsequent endoscopy within 4 weeks for direct and indirect imaging findings indicative of gastric/duodenal ulcers. Positive direct imaging findings revealed gastric or duodenal wall thickness >8 or 5 mm, respectively, and indirect findings revealed inflammatory changes, hyperechogenicity, and presence of lymph node around ulcers. Correspondingly, we calculated the sensitivity and specificity for diagnosing gastric/duodenal ulcers and used the Fisher's exact and Mann-Whitney U tests to compare the frequency of findings and gastroduodenal wall thicknesses in pediatric patients with gastric/duodenal ulcers. RESULTS Overall, 6 and 24 were diagnosed with gastric and duodenal ulcers, respectively. The sensitivities of direct and indirect findings were 60.0% (18/30) and 80.0% (24/30), respectively; the corresponding specificities were 98.1% (210/214) and 97.2% (208/214). The frequency of direct and indirect sonographic findings differed significantly between patients with gastric or duodenal ulcers (18/30 versus 24/30, P = .002). Gastric and duodenal wall thicknesses were greater in patients with gastric (6.6 ± 2.6 mm versus 3.6 ± 1.4 mm; P = .003) or duodenal ulcer (5.0 ± 1.4 mm versus 2.2 ± 1.0 mm; P <.0001), respectively, than in those without. CONCLUSIONS The frequency of indirect finding was greater than that of direct finding in pediatric patients with gastric/duodenal ulcers. Therefore, sonographers should carefully evaluate indirect findings around the stomach or duodenum.