One hundred and eighteen patients undergoing low colorectal anastomoses were randomly allocated to reconstitution by either single layer interrupted extramucosal sutures or circular staple gun. In the 60 patients undergoing sutured anastomosis there were 2 (3 per cent) clinical leaks and 4 (7 per cent) radiological leaks, and no failures. Of the 58 patients who underwent stapled anastomosis there were 4 failures, 7 (12 per cent) clinical leaks, 14 (24 per cent) radiological leaks and 1 death. Stapled anastomoses were more than ten times as expensive as sutured anastomoses and there were no savings in time or numbers of associated colostomies. An interrupted extramucosal suture technique remains the ultimate standard for low colorectal anastomosis.

Ali patients presenting with acute upper gastrointestinal bleeding between November 1986 and April 1988 were admitted to a centralised joint medical/surgical unit, with a policy of early clinical and endoscopic assessment and rapid surgical intervention in those at high risk. Of the 430 patients admitted 69*5% were over the age of60 and 30% had significant additional medical conditions. 50 4% were bleeding from peptic ulcers and one third had been taking non-steroidal anti-inflammatory agents. Fifty five patients underwent surgery, which in two thirds was carried out within 24 hours of admission, usually for continued bleeding. In patients with peptic ulcer the operation rate was 21*6%. Overall mortality was 3*7%, and in those with bleeding gastric or duodenal ulcers 5-5%; surgical mortality in the later group was 15-2%. All patients who died had serious concomitant pathology and 87% were over 70 years of age. Adoption of a centralised approach to management ofhaematemasis and melaena is feasible in a District General Hospital and associated with an improved survival.

A prospective randomized clinical trial comparing rubber band ligation (RBL) with phenol injection in 269 patients with symptomatic haemorrhoids presenting to one surgical clinic over a 6 year period, has been carried out. Questionnaires were completed by 215 patients (106 RBL and 109 injection) with an average follow up of 2.75 years. A successful outcome was achieved in 89 per cent of those receiving RBL compared with 70 per cent for injection (P less than 0.001). All symptoms tended to respond more favourably to RBL, the results achieving statistical significance in patients complaining of bleeding and prolapse (P less than 0.01 and P less than 0.05 respectively). Complications from either technique were minimal. It is concluded that RBL is superior to phenol injection in the out-patient treatment of haemorrhoids.

In a randomized controlled study of 411 patients undergoing elective colorectal surgery, the efficacy and safety of two different regimens of imipenem were compared with those of a control regimen of cefuroxime plus metronidazole. Surgical infections occurred in 92 (26.4 per cent) of 349 evaluable patients. There were no major differences between the three treatment groups. It is concluded that imipenem is as effective as cefuroxime plus metronidazole in the prevention of sepsis after elective colorectal procedures.

Although the choledochoscope has been available for several years, its use had not achieved universal acceptance. Enthusiasts claim that the incidence of retained stones in the common bile duct would be reduced by using the instrument, but no controlled study by the same team has been reported. Patients undergoing biliary surgery by one surgical team in Southampton were operated upon in one of three hospitals. The choledochoscope was available for use at only one hospital, but the other two were otherwise fully equipped, and no selection of patients was made for any particular hospital. Over a period of 7 years, 707 patients underwent elective or emergency cholecystectomy. Routine choledochography was performed and stones were found or suspected in the bile ducts in 103 (14.6 per cent). The choledochoscope was used in 54 of the bile duct explorations and in 10 patients residual calculi following routine exploration were identified by the instrument. Postoperative T-tube choledochography was performed in 35 patients following choledochoscopy, 31 with choledocholithiasis at operation, and no unsuspected retained stones were demonstrated. In the 49 patients undergoing duct exploration by the same team without the choledochoscope, 32 were found to have choledocholithiasis. Thirty-six postoperative choledochograms were performed and six patients were found to have retained stones. There were no additional complications attributed to the use of the instrument. We conclude that the use of the choledochoscope should reduce the incidence of retained bile duct stones.