Manual compared with mechanical cervical oesophagogastric anastomosis: a randomised trial.

To compare the short and medium term result of hand-sewn and stapled anastomoses after oesophagectomy.Randomised study.Teaching hospital, Italy.41 patients who required oesophagectomy between February 1993 and December 1996.Oesophagectomy and left cervical gastroplasty.Mortality and morbidity.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 hospital (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 leak in the hand-sewn group compared with 3 in the stapled group, and 1 further radiological leak in the stapled group. 2 patients in the hand-sewn and 3 in the stapled group developed strictures.Though the numbers are too small to be assessed statistically, we think that these result are sufficient to persuade us that oesophagogastric anastomoses should be hand-sewn rather than stapled.