Accuracy of contrast swallow study in assessment of caustic esophageal stricture length in children: Agreement study

Abstract

Background: Caustic esophageal stricture length assessment is essential for planning endoscopic management and predicting its prognosis. We aimed to assess the accuracy of contrast swallow study (CSS) in measuring stricture length in comparison to endoscopy (definitive investigation for actual length measurement).

Method: Medical records of caustic esophageal strictures between 2010 and 2020 were retrospectively reviewed. Reliability study was done to compare between radiological and endoscopic measurement of stricture length.

Result: 124 CSSs for 91 patients were analyzed. Six studies showed no stricture, single stricture was reported in 101 studies, double strictures were reported in 16 studies, triple strictures were reported in one study (136 radiological stricture). Endoscopy revealed 133 true strictures. Number of the strictures was consistent between CSS and endoscopy in 112 studies (90.3%) and different in 12 studies (9.7%). Eight endoscopies revealed strictures not reported in CSS (5.5% false negative strictures), while 10 CSSs reported 11 strictures that were not detected during the endoscopy (7.6% false positive strictures). Reliability analysis revealed interclass correlation coefficient = 0.6 (95% CI 0.5 to 0.7) indicating moderate reliability.

Conclusion: CSS is not accurate in assessing caustic esophageal stricture length. Combination of CSS and endoscopic investigation is better for proper evaluation of these patients. Level III of evidence.

Keywords: Agreement study; Caustic esophageal stricture; Endoscopic assessment; Radiological assessment; Reliability; Stricture length.