soon as the diagnosis is established with any degree of cer tainty, all food should be withheld and preparation should be made for imme diate operation. When the diagnosis is considered absolutely certain, the in cision is perhaps best made parallel to the border of the ribs, as most gastric ulcers perforate near the cardia and les ser curvature. In case of doubt. a me dian incision would be preferable. Care ful search should then be made, first of the anterior wall of the stomach, be ginning near the cardia, taking next the pylorus, and, finally, the posterior wall, the positions named being given in order of the relative frequency of per foration. The edges of the ulcer are in verted and one or two rows of Halsted or Lembert sutures are inserted. The excision of the ulcer is unnecessary in the great majority of cases. Special care should be taken that the entire ulcer is invaginated, and careful search should be made for a second perforation or dangerously thin area. Needless fatali ties have resulted from the neglect of both of these precautions. If it is im possible to invaginate the edges of the ulcer, they may be drawn together by sutures as a cutaneous wound would be sutured, and in ease of possible insecure suture, an omental graft may be sewed over as an additional safeguard. In case neither of these procedures is possible. because of the fixation of the stomach by adhesions or great inflammatory thickening, the abdominal cavity should be walled off by iodoform gauze and a drainage-tube introduced down to the ulcer. The resulting fistula will usually close spontaneously, or, if necessary, it may be closed by a later plastic opera tion. After suturing the ulcer the most thorough cleansing of the peritoneal cavity is of the utmost importance; thorough flushing with large quantities of warm sterile salt solution should be carried out and any suspicious spots may be wiped clean with gauze. In order that all parts of the abdominal cavity be reached it may be necessary to sepa rate peritoneal adhesions, and in some cases counter-openings may be made. Drainage is usually desirable, not only from the site of operation, but also from the pelvis.
In the after-treatment of the cases external heat, stimulating enemata, sub cutaneous use of stimulants, and intrav enous infusions of salt solution may be called for if severe shock is present. In order to give the stomach rest, nourish ment is usually given by nutrient en emata for the first five or seven days. Capsulated collections of pus, which sometimes result from peritoneal infec tion, are best opened from an external incision, as the separation of adhesions may give rise to general infection. Suc cess in this operation depends mainly upon early diagnosis and operation and thorough cleansing of the peritoneal cavity.
Patients operated upon within twelve hours from the time of perforation have excellent prospects for recovery, the mortality being about 16 per cent. in cases operated upon since 1896. Other factors of importance are the condition of the patient at the time of operation, the amount of food contained in the stomach at the time of perforation, and the skill of the operator. The entire
mortality in the cases thus far reported has been about 49 per cent.
Seventy-eight cases of operation for perforated gastric ulcer collected by per sonal assistant, Dr. Tinker, as compared with the 78 collected by Weir and Foote in 1896.
1. Age and sex: Of Keen and Tinker's 7S cases, 9 were men and 61 women; in S cases the sex was not recorded. All the men were over 25 years of age. Of the women, 41 were under 25 and only 16 over that age. The age was not stated in 4 cases.
2. The site of perforation was in the following situations in decreasing order of frequency: On the anterior wall, near the cardia, near the lesser curvature, near the pylorus, and on the posterior wall.
3. Ulcers not found: The number of cases in which the ulcer was not found at the time of operation has been very much less in recent cases than in those reported a few years ago. This has been partly due to the fact that more recent operators have had less hesitation in breaking down adhesions.
4. The mortality has been progress ively reduced. Of Weir and Foote's cases, 55 died and 23 recovered, giving a mortality of 70.51 per cent. Of the eases collected by Keen and Tinker, 25 (lied and 50 recovered: a mortality of 35.S9 per cent.
5. The mortality in relation to time of operation: An analysis of the 156 cases shows the following:— If one wishes to have any reasonable prospect of recovery, the case must be operated on within the first twelve hours, and practically the earlier the better. W. W. Keen (N. Y. Med. Jour., May 7 and 21. June 4 and 11, '93).
There is a very marked reduction in the death-rate since the introduction of operative procedures in 1SSO. Early series of cases gave a mortality of 71.51 per cent. as against 40 per cent. in the later cases. There is an immense differ ence in the mortality the earlier opera tion is done. Within twelve hours of the perforation the mortality was re duced to 19.23 per cent. The perforation was found in the anterior wall in 125 eases out of 240 collected; in the pos terior wall in 32 cases; near the lesser curvature in 61; near the cardia in 74; near the pylorus in 40 cases. These figures of operative cases are in striking contrast to Welch's observations in the post-mortem room. Of 793 cases. 523 in nearly equal proportions were in the posterior wall and in the lesser curva ture. C. E. Armstrong (Montreal Med. Jour., Aug., 1900).
Gastrolysis.— G a strolysis—i.e., freeing the stomach from adhesions—is an opera tion that has been found necessary in a considerable number of cases in which extensive adhesions have given rise to de cided disturbances of digestion or to se vere pain.
There are no positively distinctive symptoms of this condition, and conse quently the diagnosis is very difficult.
The cause of trouble has seldom been determined before operation. Strong, band-like adhesions may give rise to con striction or obstruction of the bowel.