Background: Water-soluble contrast swallow examination is routinely carried out after oesophagectomy to detect leakage of the cervical oesophagogastric anastomosis. This study evaluated the diagnostic accuracy and clinical value.
Methods: Patients with oesophageal carcinoma who underwent oesophagectomy with gastric conduit formation and a hand-sewn cervical anastomosis between 1989 and 2003 were reviewed on outcome of routine aqueous contrast swallow examination (RACSE) and appearance of clinical anastomotic leakage.
Results: An RACSE was carried out in 207 (82%) of 252 patients on postoperative day 8 (range 3-15). In 45 patients, no RACSE was executed, mainly because of a prolonged stay in intensive care unit. In 18 (9%) of 207 cases, the RACSE could not be interpreted by the radiologist. In 19 (53%) of 36 patients who developed a clinical leakage, the leak had already manifested clinically before the routine contrast examination was planned. Taken together, the false-positive rate was 8%, the false-negative rate 48%, sensitivity 52%, specificity 92%, positive predictive value 46% and negative predictive value 93%. No significant differences were found between the accuracy of RACSE in end-to-end or end-to-side cervical anastomoses.
Conclusion: Given the very low sensitivity and low positive predictive value and given the fact that in 53% of patients with a clinical leak, the leakage had appeared clinically before the contrast swallow examination was routinely planned, we propose to abandon the routine contrast swallow examination after oesophagectomy to detect cervical anastomotic leakage. Alternatively, anastomotic integrity can be tested by drinking small amounts of water with simultaneous observation of the cervical wound.