The time that a patient is kept in bed varies with the nature of the case. Old people, especially cancer patients, may often be allowed to sit up in a chair about the tenth day, though they may be moved on the sofa within a week. In these cases the abdominal wound must be firmly sewn up and supported by strapping. The average time for a patient to stay in bed after an operation on the stomach is from two to three weeks.

Complications

The complications that may follow gastroenterostomy are :

(1) Regurgitant vomiting.

(2) Contraction of the new orifice.

(3) Peptic jejunal ulcer.

(4) Pneumonia or other chest complications.

(5) Adhesions.

(6) Intestinal obstruction.

(7) Non-union and separation of the anastomosed viscera.

(8) Hernia of the intestine through the loop in the anterior, or through the mesenteric slit in the posterior operation.

(9) Exhaustion.

(10) Haemorrhage.

(11) Dragging on the jejunum when a dilated stomach retracts ; this may occur if the ligament of Treiz is short or displaced to the right of the stomach and there is no jejunal loop (3).

(1) Regurgitant Vomiting

Regurgitant vomiting is a complication that used frequently to follow the operation of gastroenterostomy, and when severe it was not infrequently fatal. It is now seldom and should never be seen, as it is entirely due to faulty technique.

It is essentially due to obstruction to the passage onwards of the duodenal contents, either from paresis of the intestine that has been handled too freely or paralysed by the too firm pressure of a faulty clamp; or to kinking of the bowel at the point of anastomosis; or to some obstruction by adhesions or pressure beyond the gastro-jejunal opening; or to the presence of a jejunal loop as in anterior gastroenterostomy. It will thus be seen that the complication is, as a rule, due to intestinal obstruction or to stasis.

The theories that have been put forward to account for it are :

(a) The presence of bile in the stomach, which Dastre's experiments on dogs absolutely disproved (4).

(b) The presence of a loop on the proximal side of the opening into the stomach, which is disproved by the large numbers of successful anterior gastroenterostomies that must necessarily have such a loop.

(c) By the situation of the opening not being at a dependent part of the stomach, also disproved by many of the early successful cases in which the opening was not made close to the lower border of the stomach.

(d) The presence of pancreatic fluid in the stomach, disproved by Moynihan's case, in which a ruptured intestine at the duodeno-jejunal flexure was treated by closing both ends of the rupture and performing a gastrojejunostomy, so that all the bile and pancreatic fluid regurgitated into the stomach through the pylorus for the fourteen weeks during which the patient survived the accident, Avithout there being any signs of vicious circle (5).

(e) The formation of a spur at the point of anastomosis. This, by preventing the onward passage of the stomach contents, may undoubtedly be a cause, but it will not occur if the technique described on p. 152 is followed.

(f) Acute angulation of the jejunum beyond the anastomotic opening; a well-recognised cause, readily avoided by one or two anchor sutures beyond the opening.

(g) Pouting valves of mucous membrane. This may be a cause, but it is readily avoided by the proper application of the marginal suture securing apposition of the intestinal to the gastric mucous membrane.

(h) Compression of the colon by the jejunal loop in the anterior operation (Doyen).

(i) Adhesions forming subsequent to the operation leading to constriction of the distal arm of the jejunum, as in a case under my care in 1901, which was operated on six months later and cured by the division of a band crossing the distal jejunal loop (6).

It will thus be seen that the causes of the vicious circle are avoidable, and the complication should therefore seldom, if ever, occur ; and, in fact, since recognising the cause in 1901 I have never seen a case of regurgitant vomiting in my practice.

Treatment Of The Vicious Circle

This should be preventive by accuracy of technique, and if the following points are observed the vicious circle will not occur :

(a) Accurate union of the mucous margins of the stomach and jejunum.

(b) Securing the anastomotic opening at or near the lower border of the stomach.

(c) Applying one or more anchor sutures be}Tond the point of anastomosis.

(d) Bringing the distal loop of jejunum over to the right of the spine in arranging the peritoneal toilet before closing the abdomen.

(e) Making the anastomosis in the posterior operation either without a loop or with a very short interval between the anastomosis and the jejunal flexure.

(f) In the anterior operation the loop must not be made too short so as to compress the colon.

If the technique has been faulty, and unfortunately regurgitant vomiting should occur, what can be done ?

(a) Raise the head and shoulders so as to prop up the patient in a semi-recumbent posture.

(b) Wash out the stomach and repeat it if necessary.

(c) Feed by the bowel and stop mouth-feeding for a time.

(d) Give small doses of calomel in repeated doses, followed by enemata, to try to secure a movement of the bowels.

(e) If these fail, do not wait too long before reopening the abdomen and performing entero-anastomosis-an effectual method of treatment.

(2) Subsequent Contraction Of The Anastomotic Opening

Although there may be moderate contraction of the new opening, both in cases where the stomach is greatly dilated before operation and in those where the pylorus is patent at the time of operation, yet if the anastomotic opening be made sufficiently large, not under 2 in., and the union of mucous membrane to mucous membrane be efficiently performed, contraction to a serious extent will not be likely to occur. I have always found the opening to be patent where I have at long periods subsequent to operation had to operate again for some other cause.