Manual compared with mechanical cervical oesophagogastric anastomosis: a randomised trial

Objective: To compare the short and medium term result of hand-sewn and sta pled anastomoses after oesophagectomy. Design: Randomised study. Setting: Teaching hospital, Italy. Subjects: 41 patients who required oesop hagectomy between February 1993 and December 1996. Interventions: Oesophagectomy and left cervical gastroplasty. Main outcome measures: Mortality and morbidity. Result: 21 patients were randomised to have the anastomosis hand-sewn, and 20 to have it stapled. The two groups were comparable. 3 patients died in h ospital (2 in the hand-sewn and 1 in the stapled group), and the remainder were followed up a mean of 21 months (range 6-34). There was one clinical l eak in the hand-sewn group compared with 3 in the stapled group, and 1 furt her radiological leak in the stapled group. 2 patients in the hand-sewn and 3 in the stapled group developed strictures. Conclusion: Though the numbers are too small to be assessed statistically, we think that these result are sufficient to persuade us that oesophagogast ric anastomoses should be hand-sewn rather than stapled.