U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content







AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window






Chapter 30


Part III



The Management of Colostomies  

David Henry Poer, M. D.  


Although colostomy was a well-established procedure in civilian surgical practice at the beginning of World War I, military surgeons of all armies were slow to employ it.1 The chief reason was that during the early months of this war, as in all previous wars, nonintervention was the general rule in abdominal injuries. This was partly because of the poor results which had attended previous efforts at intervention and partly because the majority of surgeons did not regard major surgery as practical near the front. Since abdominal surgery, to have any hope of success, had to be performed within a reasonably short time after injury, this pessimistic point of view automatically excluded the surgical management of these injuries.  

Many surgeons, even in 1914, did not accept these premises, and as the war progressed the concept became more and more general that recovery could be expected in abdominal injuries if operation was performed. As mobile hospital facilities, equipped and staffed for surgery, were set up near the front, the practice of prompt surgery for these injuries became increasingly frequent, and in 1917 and 1918 operation became the general rule in both the British and the American Armies.  

Colostomy, however, was never adopted as a routine technique in wounds of the large bowel. Instead, it was usually reserved for the most seriously wounded patients, with extensive wounds of the colon. These casualties formed a distinctly unfavorable group, and the high case fatality rate which attended the use of colostomy was no more than might have been expected. 

Wallace, who wrote the section on abdominal injuries for the official British medical history of World War I,2 did not mention colostomy as a method of dealing with wounds of the cecum. He considered suture preferable whenever it was feasible in wounds of the transverse and vertical colons, and he also stressed the importance of drainage. Colostomy, he indicated, was an undesirable alternative, which was most likely to be necessary in injuries of the splenic flexure. Its widest application, in his opinion, was in left-sided wounds of the  

1(1) History of the Great War Based on Official Documents. Medical Services Surgery of the War. London: His Majesty's Stationery Office, 1922, vol. I, pp. 476-579. (2) The Medical Department of the United States Army in the World War. Washington: U. S. Government Printing Office, 1927, vol. XI, pt. 1, pp. 443-490.  
2Wallace, C. S.: Wounds of the Abdominal Viscera. In History of the Great War Based on Official Documents. See footnote 1 (1).  


loin, in which sepsis tends to develop late, and in rectal injuries.  Colostomy was employed only 53 times in the 155 injuries of the large bowel upon which Wallace based his discussion. It was, however, widely used in base hospitals in the management of fecal fistulas which developed after failure of suture repair. 

The results which Wallace cited explain why colostomy was not a popular procedure among British surgeons in World War I. The case fatality rate in wounds of the colon, he pointed out, was higher when it was employed; 73.5 percent as compared with just over 50 percent in cases managed by suture. These figures, Wallace added, must not be taken to indicate that colostomy is a necessarily dangerous procedure. They merely indicated that fatal sepsis was likely to occur when colostomy was employed, because infection was already present when the operation was performed.  

Lee, who wrote the section on abdominal injuries in the official United States medical history of World War I,3 set down the following general principles for the management of wounds of the colon:

1. Suture should be performed whenever it is possible, in order to secure a satisfactory closure.  

2. Colostomy should always be employed in preference to resection.  

3. Colostomy is preferable in all injuries with large, ragged openings, particularly injuries of the cecum, descending colon, and sigmoid.  

4. Drainage should be used routinely in proved or suspected retroperitoneal injuries. It should always be used when there is any doubt of the integrity of the suture line.  

5. Colostomy may be required in extensive lacerations of the lower rectum and in rectal injuries deep in the pelvis, in which suture is not possible. According to Lee, the results of suture repair were better than those of colostomy, in which the case fatality rate was 70 percent.  

Jolly's4 extensive personal experience as a field surgeon in the Spanish Civil War included 970 abdominal injuries. In his book on field surgery in total war5 he described invaginating suture and resection as alternative methods of treatment in wounds of the colon. He emphasized the importance of drainage but did not mention colostomy.  


The attitude of surgeons toward colostomy in World War II is an interesting contrast to the opinion held of it early in World War I and in the Spanish Civil War, which ended only about 3 years before the United States entered World War II. Its lifesaving properties were recognized almost at once, and, together with exteriorization of the damaged segment of bowel, it became the routine method of management for injuries of the large bowel. This policy,  

3Lee, Burton, J.: Wounds of the Abdomen. In the Medical Department of the United States Army in the World War. See footnote 1 (2), p. 337.  
4Jolly, Douglas W.: Field Surgery in Total War. New York: Paul B. Hoeber, Inc., 1941. 


which saved many thousands of lives, can fairly be regarded as one of the most important advances ever made in the treatment of battle-incurred injuries of the colon.  

Colostomy, with and without exteriorization of the damaged bowel, was first used by British surgeons during the air bombardment of London in 1940 and 1941, at the repeated suggestion of Ogilvie.6 

The new policy was apparently dictated by two considerations. The first was the success of this method of management of intestinal tumors in civilian practice. The affected segment of bowel was brought outside the abdomen at the first operation and resected later at a second-stage operation. The second consideration was the prohibitively high case fatality rate associated with immediate repair of injuries of the colon and replacement of the bowel into the abdominal cavity. As Ogilvie remarked later, in a report of the British experience with abdominal injuries in the Western Desert,7 it would be rash to claim that any of the patients who died after colostomy could have

6In a personal communication to the author of this chapter, dated 24 May 1955, Sir Heneage Ogilvie pointed out that much of this teaching was informal. He had first publicly advocated the use of colostomy for wounds of the large bowel at a lecture given at the Postgraduate Hospital in Hammersmith, England, late in 1939. This lecture was one of a series prepared for the emergency and dealing with the surgical treatment of wounds of warfare. How to manage wounds of the colon was vigorously discussed by the numerous surgeons present, and at this meeting, as well as at later meetings, a certain amount of opposition to the policy of colostomy was expressed.  
A lecture given by Sir Heneage in Kenya, when he was consulting surgeon to the East African Force which was conducting the attack on Abyssinia from the south, was reproduced in the June 1941 issue of the East African Medical Journal. The substance of the recommendations concerning the management of wounds of the colon was as follows:  
"Holes in the free portions of the colon may be closed by a purse-string suture if the bowel wall round them is perfectly healthy, but in most cases a local repair is unsafe and resection is wise. A free mobilisation is often required before the resected ends can be brought together without tension. Nearly all writers on military surgery advise end-to-end suture after resection, with or without caecostomy, but this advice is unsound in the light of modern teaching and practice, and the high mortality in colon wounds in all published statistics suggests that it is not good treatment. After mobilisation sufficient to allow the colon above and below the injury to be approximated without tension, the injured segment should be removed and the cut ends above and below it, closed with crushing clamps, should be brought to the surface at one end of the incision or wound. The clamps are removed after 24 hours, and the continuity of the bowel is restored later by crushing the spur between the openings.  
"There has been a marked tendency in recent years to look on the large intestine with increasing respect or even fear. Its walls are thin, their blood-supply is poor, and the peritoneal coat is interrupted by a broad mesentery and distorted by fat blisters, the contents are highly infective and mechanically traumatic, its luminal pressure alters with explosive suddenness. No sutures can be placed accurately in its wall, and what are thought to be lemberts often pass through all coats and carry infection. All surgeons have experienced trouble with leakage and sepsis at the suture line after resection and many feel today that no stitch should ever be put through a colon that is not both empty and sterile. A colostomy or a resected loop should not be stitched to the skin, but the skin opening should be made to fit the colostomy; a segment of bowel should not be excised and the ends sutured unless the contents have been diverted above by an excluding colostomy for at least 2 weeks. Devine has shown the way to success in cancer of the colon by his operation of exclusion and Lahey has made resection of the colon for ulcerative colitis and of the rectum for cancer safe by stage operations in which the bowel is brought to the surface. In war injuries, the way to safety is the same, for caecostomy provides only partial relief of tension at the injured site and proximal colostomy is no better unless it is done some weeks before."
The same advice was repeated in various memorandums circulated in the British East African Force in 1941 and in the British Middle East Force in 1942 and 1943, as well as in a field manual published for the use of British troops during the invasion of the Continent in June 1944. Meantime, Ogilvie had published his comprehensive report Abdominal Wounds in the Western Desert (March 1944, see footnote 7).  
The personal communication from Sir Heneage Ogilvie from which the information in this footnote has been taken concludes with the following paragraph:  
"Under the influence of these antiseptics [the effective intestinal antiseptics which were later developed], with improved surgical conditions and with the development of skilled surgery in the forward units, the policy was somewhat modified in that closure of small perforations in the proximal colon was countenanced if the surgeon was a man of considerable experience and if the case could be retained for the first 10 days for observation. However, as far as I know, up to the end of the war it was considered that exteriorisation of all colon wounds was the safest policy and the one to be recommended to all surgeons undertaking the abdominal surgery of warfare for the first time."
7Ogilvie, W. H.: Abdominal Wounds in the Western Desert. Surg., Gynec. Obst. & Obst. 78: 225-238, March 1944.


been saved by suture alone, but there was no doubt that certain of the deaths which occurred after suture of the injured bowel could have been avoided by colostomy. In his opinion, the exteriorization of the damaged colon was one of seven decisive points in the management of abdominal injuries. By this time (1943) the British experience had been sufficiently large for him to be able to analyze the conditions which made closure of a colostomy stoma difficult and to suggest improved techniques.  

The American experience.-Colostomy was extensively employed in injuries of the colon during the North African campaign in 1942-43, at first, it would seem, without knowledge of the earlier British experience. Circular Letter No. 20, 22 June 1943, Office of the Surgeon, North African Theater of Operations, consisted chiefly of comments from hospitals of the Communications Zone, North African Theater of Operations, United States Army, on the management of battle casualties in forward areas during the fighting in Tunisia. One of the criticisms had to do with failure to perform immediate colostomy in retroperitoneal injuries of the rectum.  

Colostomy became official practice for injuries of the large bowel when Circular Letter No. 178, dated 23 October 1943, was issued from the Office of The Surgeon General of the Army. The instructions were that-

* * * in large bowel injuries, the damaged segment will be exteriorized by drawing it out through a separate incision, preferably in the flank. In order to facilitate subsequent closure the two limbs of the loop should be approximated by suture for a distance of about 2? inches and then returned to the abdomen, leaving the apex exteriorized with a short length of rubber tubing or other suitable material beneath it. If the segment cannot be mobilized, the injury should be repaired and a proximal colostomy done.  

Thereafter in World War II, surgery of the colon in the United States Army was based on three general principles: (1) Exteriorization of the wounded portions of the bowel (fig. 31), to avoid intraperitoneal contamination; (2) complete diversion of the fecal stream away from distal wounds of the colon and rectum; and (3) colostomy to effect incomplete diversion of the fecal stream for purposes of gaseous decompression as well as possible future complete diversion.

Soldiers with wounds of the colon were for the most part treated in field hospitals, located near division clearing stations. They were brought to them and were prepared for surgery by resuscitative measures within a remarkably short time. Figures of the 2d Auxiliary Surgical Group (p. 256) show that the average timelag between wounding and operation in 1,222 wounds of the colon and rectum was 10.9 hours. The speed with which treatment was instituted in itself saved many lives. Strict adherence to the policy of exteriorization of the bowel and colostomy shortened the time necessary to care for wounds of the colon and saved additional lives by making earlier treatment possible for other wounded men.  

During the early American participation in the war, revision of the colostomy stoma was undertaken, when it was indicated, in fixed hospitals overseas, after which the casualties were returned to the Zone of Interior for reconstructive surgery and closure of the opening in the bowel. As experience


FIGURE 31.-Loop of damaged sigmoid exteriorized through separate inguinal incision 12 days after wounding. When the edema now present subsides, a satisfactorily functioning colostomy will remain until it is closed.

increased and additional trained surgeons were sent overseas, closure of the colostomy was frequently effected in numbered general hospitals in the communications zone. This was a common British practice. Ogilvie related that men who had been wounded in the abdomen at El Alamein marched with the British Eighth Army into Tunis. Similarly, many United States soldiers who had sustained wounds of the colon and had been treated by colostomy were able to return to active military duty in the theaters in which they had been wounded.  

For many reasons, however, it was more expedient to close most colostomies in general hospitals in the Zone of Interior, and the greater part of the data upon which this chapter is based concern that experience.  


Basic Data  

During the winter of 1945-46, questionnaires dealing with various aspects of colostomy were sent to all general hospitals in the United States. Although


many hospitals had been closed by this time and their records had been forwarded to a general collection center, the replies received furnished information on 2,378 soldiers with wounds of the colon who had been treated by colostomy in forward hospitals overseas. As might have been expected, since a large majority of these patients had already undergone disposition proceedings, numerous details were missing on many of the records.  

For convenience, and because of the deficits in the data just mentioned, these 2,378 cases are discussed in the following groupings:  

1. Two thousand one hundred and two cases in which closure had been done in Zone of Interior hospitals and on which information had been secured when the final analysis of cases was begun.  

2. Four hundred and sixty-four cases included in the two thousand one hundred and two cases just mentioned and reported from ten general hospitals in the Fifth and Sixth Service Commands and from McCloskey General Hospital in the Eighth Service Command (series A). These data were unusually complete and, for that reason, are frequently discussed separately.  

3. One hundred and eighty cases in which the colostomy had been closed in eight overseas hospitals (series B).  

4. Ninety-six cases from Rhodes General Hospital in the Second Service Command (series C) . These data were received after the rest of the material had been analyzed but are discussed separately at appropriate places.  

Techniques of Colostomy  

As might be expected, when so many surgeons had operated on so many patients and under such widely different circumstances, the 2,378 colostomies which make up this composite series represented every known technique. In approximately 300 cases, no details were stated. Loop colostomy with a single opening was performed in 40 percent of the remaining cases, the Paul-Mikulicz colostomy with spur in 30 percent, the Devine technique or some modification thereof and mucous fistulas with single or multiple openings in 10 percent each, and the tube or tangential colostomy and the loop colostomy with double openings in 5 percent each. The colostomy was located on the left side in 50 percent of the stated cases, in the transverse colon in 30 percent, and on the right side in 17 percent. The remaining operations were in the sacral region.  

The location of the wound and of the colostomy was not stated in a large proportion of the records. In series A, the wound was on the left side in about a third of the 416 stated cases, and the colostomy was on this side in about 58 percent of the 338 stated cases.  


(Link to table)

Preclosure Problems in Zone of Interior Hospitals  

The eventual medical policy in the management of wounds of the colon was, as already stated, to close the stoma in almost all instances in Zone of Interior hospitals. Closure, however, involved considerably more than the actual operative act. It required very careful preoperative preparation, the routine of which was based upon the problems presented in the special case. Among these problems the following were the most important:  


Associated wounds.-Well over half of the patients received in Zone of Interior hospitals with unclosed colostomies had other major injuries, which in many instances were multiple (see following tabulation of associated visceral injuries in 179 patients in series A). Wounds of the small intestine were particularly frequent; series A, in which they made up considerably over half of all associated wounds, is typical. Fractures of the long bones represented as much as 40 percent of associated injuries in many hospitals. A few patients in each hospital were paraplegic.







Small intestine:







Not stated














Urinary tract


Spinal cord




The majority of the associated injuries were well on the way to healing when the patients reached the United States, but the associated injuries sometimes required attention more urgently than did the colostomy. If fractures, for instance, had to be treated by balanced suspension skeletal traction, closure of the colostomy stoma was not feasible immediately; it was desirable that the patient be ambulatory before abdominal surgery was undertaken. Nerve suture and plastic surgery, on the other hand, were usually postponed until after the colostomy had been closed.  

Organic disease was not usual in young, sturdy soldiers, but three patients at McCloskey General Hospital had pulmonary tuberculosis. There were also occasional instances of this and other diseases at other hospitals.  

Nutritional status.-Almost all of the patients who arrived in the Zone of Interior with unclosed colostomies had lost weight (fig. 32). At one typical hospital, the weight loss ranged from 10 to 80 pounds and averaged 36 pounds. In numerous instances, the patients were at least partly responsible for their own poor nutritional status. They deliberately restricted their food intake in order to decrease the number of daily dressings required. Hypoproteinemia was found in about 10 percent of the serum-protein determinations, but other-


FIGURE 32.-Right-sided colostomy. The cecum and ascending colon were resected at the first operation after wounding, and the ends were implanted into the incisions. After resection, the ends were brought out, and an ileostomy and colostomy were created. Note the extreme emaciation which accompanies right-sided colostomy. This patient also had a fecal fistula in the buttocks. After a long period of careful preparation, intestinal continuity was restored by ileotransverse colostomy. Full recovery followed.  

wise there was little evidence of serious alteration in the basic constituents of the blood except in patients with wounds on the right side, complicated by drainage from the small intestine.  

Complications related to the colostomy.-The series A statistics may be taken as typical of the collected series in respect to complications related to the colostomy. In 275 of the 464 cases, herniation of various degrees had occurred; in some instances the bowel actually protruded through itself (that is, through the colostomy stoma). Fecal fistulas were present in 70 patients, wound infection of some degree in 38, obstruction of some degree in 34, fistulas into the buttocks in 31, osteomyelitis in 25, wound dehiscence in 10, and peritonitis in 7. Complications of wounding unrelated to the colostomy included urinary fistula in 7 cases, empyema in 2, thrombophlebitis in 2, and a fistula of the biliary tract in 1.


FIGURE 33.-Fecal fistula into extraperitoneal segment of rectosigmoid through rifle wound of buttocks; probe indicates depth of fistula. A double-barreled loop colostomy created in the descending colon immediately after injury required revision and complete separation of the stoma to prevent continued contamination of the fistula. Spontaneous healing of the fistula followed. The colostomy was later closed by direct anastomosis within the peritoneal cavity.  

In a small number of cases, the colostomy was not functioning satisfactorily when the patients were received in Zone of Interior hospitals, and prompt revision or relocation was required. A colostomy placed posteriorly in the sacral region, for instance, was relocated anteriorly, in the descending or sigmoid colon. When there was persistent contamination of the buttocks or a persistent perineal fistula (fig. 33), secondary colostomy was necessary to accomplish defunctionalization of the lower rectal segment. The secondary opening was preferably placed proximally in the transverse colon (fig. 34), a site which many surgeons preferred for primary colostomy in injuries of the pelvic colon.  

Colostomies performed by surgeons working under the stress of field conditions were sometimes not performed with the care and precision possible in civilian hospitals. Holder and Lewison,8 who had a large experience in the closure of colostomy overseas, listed the conditions which made reconstructive surgery difficult as follows:  

1. The colostomy had been placed so near the costal margins or the iliac crest that osteomyelitis was an almost inevitable complication (fig. 35).  

8Holder, H. G.; and Lewison, E. F.: Management of Colostomies Performed for War Injuries. Ann. Surg. 126: 253-261, September 1947.


2. The colostomy had been placed too close to a coexisting cystostomy.  

3. The colostomy had been created through the exploratory incision, without adequate mobilization of the bowel. This error, which was most often evident in colostomies of the transverse colon, was likely to be followed by retraction of the stoma, infection of the abdominal wall, intraperitoneal infection, sinus formation, and wound dehiscence. Wound infection, in fact, was almost inevitable because of the practice in military surgery of removing the clamps from the bowel immediately. This was necessary precaution, for, in the rush of work, they could easily be overlooked, and obstruction could occur as a result; but when it was carried out in a colostomy created through the exploratory incision, feces promptly spilled over the fresh wound.  

4. The colostomy was created through the wound of entry or exit (fig. 36). This technique was almost invariably followed by infection, which frequently went on to spreading cellulitis.  

5. The spur in a Mikulicz colostomy was sometimes too short because the loops had been insufficiently mobilized. As a result, the bowel retracted, and the loops became rotated or partially obstructed.  

6. Edema or necrosis sometimes complicated loop colostomies because glass rods or heavy rubber tubing had been used to support an inadequately mobilized  

FIGURE 34.-Double stoma in transverse colon, with complete division of bowel. The mucous membranes, however, are still in contact, and the distal segment of the colon remains contaminated. Note the extensive abdominal incisions, the result of multiple surgery.  


FIGURE 35.-Extensive injuries of descending colon, for which resection was necessary at first operation after wounding. Implantation of proximal end into wound of entrance in flank; distal end has been brought out through inguinal incision. The splenic flexure was liberated by intra-abdominal manipulation to permit primary anastomotic closure.


FIGURE 36.-Lumbar colostomy following exteriorization of cecum and ascending colon in wound of flank. The lumbar region is not a location of choice for colostomy, but in this instance function was satisfactory. Closure was effected at an extensive intra-abdominal operation, by ileotransverse colostomy.  

loop. If the incision through which the bowel had been brought out was too short, constriction of the bowel, with partial dysfunction of the colostomy, was a likely result.  

7. The openings in loop colostomies were sometimes unnecessarily large (fig. 37). Often they had been created in the transverse axis of the bowel rather than the longitudinal axis.  

8. Complete diversion of the fecal stream had sometimes not been accomplished (fig. 38). This was essential in wounds of the rectum and extensive wounds of the rectosigmoid, with associated injury to the bladder, urethra, and bony pelvis.


FIGURE 37.-Excessive protrusion of transverse colon and eventration of mucous membrane with incomplete division of bowel. The opening in the abdominal wall is also unnecessarily large. This patient sustained an injury of the perineum, with loss of urinary control. After reduction of the internal herniation of the mucosa, the colostomy was closed by primary anastomosis without too much difficulty.  

9. Drainage of the retroperitoneal and pelvic spaces had been omitted in some cases and was not adequate in others. Drainage through the track created by the missile was almost never adequate.  

In addition to the technical errors just listed, other errors were observed. The use of sutures which were too tight and the inclusion of the mesenteric artery in the suture of the spur could cause tension, torsion, and constriction of the bowel, with resulting impairment of the circulation. Thrombosis was also a possible complication. Hernia might develop when too large a segment of the bowel was used for the colostomy. Partial obstruction could follow too tight closure. Eventration of the ileocecal area was a possibility in wounds of the cecum.  


Infection.-Infection always introduced serious preoperative difficulties. It could arise, as just observed, in dysfunction of the colostomy. It could also originate in a number of other ways. Sometimes abscesses in the chest or the abdomen failed to resolve completely. Infection frequently occurred along persistent fistulous tracts, particularly when osteomyelitis was a complication or a foreign body had not been removed. Fistulas which had healed spontaneously or which had been closed in overseas hospitals sometimes broke down and became infected because the fecal stream had been incompletely diverted and the tract had become contaminated. In cases of persistent infection, the cause had to be sought for and eliminated and adequate drainage established. This often required an extensive surgical procedure (pp. 352, 353). 

Fecal fistulas.-Fecal fistulas were always troublesome and were sometimes the cause of serious preoperative difficulties. This was particularly true of fistulas of the small bowel and fistulas in the sacral region. Urinary fistulas could also be very troublesome.  

FIGURE 38.-Sigmoid colostomy (incomplete division of bowel), with eversion of mucous membrane, performed to sidetrack fecal current from fecal fistula in buttocks.  


In some instances, because of a surgical error, the artificial anus had been constructed in the terminal ileum or the ileum had been used to form the colostomy in combination with the large bowel. The spur might consist of colon and ileum, or a single or a double mucous fistula might be created following end-to-end or side-to-side anastomosis between the ileum and the colon. Fistulas of the jejunum or the duodenum were usually the result of the original injury or of surgical attempts to repair it.  

Fistulas of the small bowel produced irritation or actual digestion of the skin. Sometimes the process was extensive. All methods used to protect the contaminated areas were likely to be unsatisfactory. Moreover, because of continuous drainage and the loss of essential elements from the gastrointestinal tract, patients with fistulas were often gravely emaciated. Immediate closure of the fistula was therefore indicated.  

Small, uncomplicated fistulas could be eliminated by inversion of the intestine into the lumen. In some complicated cases, a shunting or sidetracking procedure had to be employed, with anastomosis of the ileum to the descending or transverse colon proximal to the fistula, by either the side-to-side or the end-to-end technique. Later, the colostomy could be closed by some one of the methods shortly to be described, or, if it was indicated, resection of the intervening ileum and colon could be carried out. Improvement following closure of a fistula in the small bowel was usually prompt and was often dramatic.  

Patients with fistulas in the sacral region usually arrived in the Zone of Interior in poor condition, as the result of chronic sepsis. In perhaps half of all wounds of the rectum, the bony structure of the pelvis had been damaged also, and osteomyelitis was a frequent complication. As a rule, it was extensive, involving the sacrum, the coccyx, the wings of the ilium, or the upper portion of the femur. The joint cavities were also often involved in the infection. The fistula was persistently contaminated because the distal segment of the intestine had not been completely defunctionated, and the fecal drainage enhanced the bone infection. Finally, foreign bodies, which in this location are difficult to localize accurately and which were therefore often left in situ, frequently served as foci of infection. The complete removal of all foreign bodies, while extremely important at the first operation, was by no means as simple as it may sound. Pieces of clothing and gauze and bits of equipment which were not radiopaque were extremely difficult to visualize and identify. If all extraneous material was not removed, persistence of infection in fistulous tracts was inevitable. The vicious circle thus set up was extremely difficult to interrupt.  

If a foreign body was not present or if osteomyelitis was not a complication, a fistula in the sacral region occasionally healed spontaneously, even when there was persistent contamination of the distal segment of bowel. More frequently, prolonged treatment was necessary. It included daily irrigations of the fistula with physiologic salt solution or some other solution, the admin-


istration of a sulfonamide, and measures directed toward improvement of the general condition. These measures, although they were occasionally successful, were not desirable definitive treatment for a fecal fistula. For one thing, it was impossible to determine in which cases they would succeed. For another, even an apparently successful result of expectant treatment was not always permanent. The fistula was likely to reopen, the mucous membrane of the bowel tended to evert into the fistulous tract, and other undesirable consequences could ensue.  

Experience soon showed that even a short and direct fistulous tract which traversed bone would not heal until three conditions had been fulfilled: (1) The fecal stream had to be completely diverted; (2) all dead bone, including the coccyx and parts of the sacrum, had to be removed; and (3) all scar tissue which prevented obliteration of dead spaces also had to be removed. Paradoxically, healing of long, tortuous fistulous tracts which traversed bone sometimes occurred when only expectant measures were used, but the course of treatment was long and tedious and the results were so uncertain that surgery was always the preferred method.  

The first step in the treatment of a fecal fistula was diversion of the fecal stream, either by revision of the colostomy or by creation of a secondary colostomy placed proximally in the transverse colon. Revision was best accomplished by some modification of the Devine technique (fig. 39), in which the openings would be separated by a bridge of skin. The next steps were removal of any foreign bodies in the region of the fistula and adequate unroofing of the entire tract from its external opening down to the rectum. Unroofing was often difficult and hazardous because of the extensive thickening and resulting rigidity of the rectal wall and the adjacent tissues. Incisions into the peritoneum lateral to the rectosigmoid helped to free the rectum. Eversion of the rectal mucosa into the fistula, which had usually occurred, was corrected by excision of the involved mucosa or by inverting it into the rectal lumen. The final step of the operation was excision of scar tissue and of the coccyx and parts of the sacrum, together with removal of all foreign material such as metal, bits of clothing or equipment, and surgical gauze, which was sometimes still present since it had been applied as a first-aid measure.  

The extensive surgery required for the repair of sacral fistulas, in addition to the tissue lost because of the ravages of infection, usually caused serious losses of soft tissue. Plastic repair was therefore frequently necessary. The preferred technique was to perform a staged operation, swinging muscle flaps across the debrided tract and then using full-thickness skin grafts.  

Fistulous communications between the intestine and urinary tract were fairly frequent and had to be dealt with by various measures before closure of the colostomy could be attempted. They were sometimes kept open by obstruction caused by urinary calculi or by adhesions. The proctoscopic and cystoscopic studies which were essential had to be carried out with great care because of the induration and scar tissue present in both rectum and


FIGURE 39.-Revision of sigmoid colostomy by modified Devine technique to provide complete diversion of fecal stream from fistula in buttocks. The colostomy was closed by direct anastomosis within the abdominal cavity.  

bladder. Barium enemas and the injection of the fistula with radiopaque substances were useful in determining the limits of extension of the lesions.  

If the ostium was not too large, conservative therapy was usually given a trial in urinary fistulas, in an attempt to promote healing at the point of origin. If urinary calculi or adhesions were responsible for failure of healing, conservative measures were useless, and surgery was indicated. If the fistula communicated with the rectum in the lower segment or near the anus, a plastic procedure (the Stone operation) was almost always required. The patients in this group were transferred to Walter Reed General Hospital for the necessary surgery. 

In addition to the more usual types of fistulas, bizarre types were not uncommon. They included fistulas extending through the vertebral column, through the diaphragm and chest, into the acetabulum, and into the urethra. The latter type was easily diagnosed by the passage of gas through the penis.  


FIGURE 40.-Left inguinal colostomy with incomplete division of bowel following exteriorization of wounded descending colon. This frequently used type of colostomy functioned satisfactorily and was easily closed by direct anastomosis. Application of spur-crushing clamps was hazardous because interposition of loops of small intestine occasionally occurred.  

Preoperative Regimen  

The time interval between reception of the patient in a Zone of Interior hospital and closure of the colostomy varied widely, sometimes because of the patient's own status and sometimes, particularly during periods of intense activity, because of crowded hospital conditions. The categoric statement can be made, however, that in no instance was any harm caused by the delay. In fact, one of the strongest points in favor of handling wounds of the colon by exteriorization and colostomy was the ability of patients thus treated to travel long distances without harm and to wait indefinitely for closure of the stoma without deterioration of their status.


The relatively few patients who were received in Zone of Interior hospitals without associated injuries, free from infection, and with colostomies which were functioning satisfactorily were prepared for operation at once and were operated on as soon as possible. These patients were the exceptions. The great majority presented preoperative problems of one sort or another (p. 343) and required a more extensive preparatory regimen in addition to special measures suitable to their individual difficulties.  

As soon as patients were received in Zone of Interior hospitals, they were taught to take care of their own dressings. Those who were able to be ambulatory were required to be. Associated injuries were treated as necessary. Whenever it was practical, arrangements for furloughs were made. The psychologic stimulus of a return to their homes and families amply compensated for any delay in closure of the colostomy.  

Operation was not undertaken in any case until the colostomy was functioning satisfactorily (fig. 40) and certain other criteria had been met:  

1. The nutritional status must be restored to a level at which surgery would not be attended with any undue risk. A completely normal status was desirable but was not regarded as essential.  

2. The healing of associated injuries must have advanced to a stage at which ambulation was possible.

3. Fecal fistulas must have been healed for at least 8 to 12 weeks. 

4. The skin about the colostomy must be in optimum condition.  

Special measures.-The first essential of the general preoperative regimen was attention to the colostomy. Many patients, after trips from overseas which sometimes occupied a considerable time, were received with large amounts of inspissated fecal material in the distal loop. These accumulations were removed by irrigations of distilled water, physiologic salt solution, or solutions containing acriflavine, potassium permanganate, or penicillin. After the distal loop had been completely cleaned out by these means, irrigations were carried out three to five times weekly, with two objectives, to restore the tone of the loop and to increase the size of the opening, which had often become much smaller in the weeks or months since the colostomy had been created.  

The second essential of preoperative preparation was the protection of the skin from fecal discharges. This was particularly important if irritation was already present. The best results were accomplished with dressings of kaolin or aluminum paste.  

The third essential of the general regimen was restoration of the nutritional status to a level approaching normal. The caloric content of the diet was maintained between 3,500 and 5,000 calories daily. There was special emphasis on the protein component, which ranged between 125 and 175 gm. daily. Transfusions of plasma and whole blood were given as indicated. Supplementary vitamins, iron, and liver extract were also given as necessary.  

Immediate preoperative preparation.-Special preparations for surgery were begun several days in advance of operation. The routine varied in details


from hospital to hospital but was always based upon the following general policies:  

l. Routine laboratory tests were conducted, and operation was postponed if the chlorides, protein, and vitamin-C levels of the blood were not approximately normal. The prothrombin time was determined the day before operation, and closure of the colostomy was deferred if it was not satisfactory. Roentgenograms of the abdomen were made to demonstrate possible retained foreign bodies, and barium-enema studies were made to exclude possible intestinal obstruction.  

2. A sulfonamide drug (usually Sulfasuxidine (succinylsulfathiazole), phthalylsulfathiazole, or sulfaguanidine) was given for 5 days before operation in doses of 1 gm. six times daily. If local irritation was evident on the skin during the course of treatment, kaolin or aluminum paste was applied.  

3. Vitamin K was usually given the day before, and the day of, operation. Paregoric was ordered just before the patient was taken to the operating room. 

4. The ample diet previously provided was discontinued 3 days before operation and was replaced by a low-residue diet, which was, however, equally high in protein content. Only clear fluids were permitted during the 24 hours immediately before operation.  

5. Irrigation of both loops of bowel was continued up to the day of operation. On the morning of operation, irrigations of sterile physiologic salt solution were continued until they returned perfectly clear.  


Practically all methods of anesthesia were used to close the colostomies in this series, but fractional instillation of an anesthetic drug into the spinal canal was perhaps the most favored technique. Pentothal Sodium was sometimes used for induction but was never used for the entire procedure. Local anesthetic agents were avoided, because of the risk of spreading infection.  

Technical Considerations  

Official directives did not specify any standard procedure for the closure of a colostomy. This left the surgeons at the various general hospitals free to develop their own methods, under the general supervision of consultants. Since 106 surgeons are known to have closed the 2,198 colostomies managed in the general hospitals in the United States, the range of detailed techniques was correspondingly wide, although only a few basic techniques were employed.

During the early months of American participation in the war, the tendency was to employ extraperitoneal methods of closure. Later, intraperitoneal techniques became increasingly popular, for two reasons:  

1. In spite of careful efforts to avoid opening the peritoneal cavity, it was frequently entered during extraperitoneal operations.  


2. It was found that undesirable consequences did not follow deliberate or accidental invasion of the peritoneal cavity during the closure of a colostomy. 

In the 1,813 cases in which the technique employed in closing colostomies in the Zone of Interior hospitals was stated, the intraperitoneal approach was used 1,055 times (58.2 percent) and the extraperitoneal approach 758 times. At Rhodes General Hospital (series C), all 96 patients were treated by the intraperitoneal technique. In the 437 cases in which this information was available in series A, the intraperitoneal approach was used 186 times and the extraperitoneal 251 times.  

Information concerning the precise technique of closure was available in only 1,331 of the 2,198 operations performed in Zone of Interior hospitals. End-to-end anastomosis was employed in 642 cases (48.2 percent). The Pauchet type of anastomosis was employed in 22 cases, all at the same hospital, and other measures, including spur crushing followed by closure, were employed in the remaining 667 cases. At Rhodes General Hospital, simple closure was performed in 60 cases, end-to-end anastomosis in 42, and ileocolostomy in 6. There were 12 multiple operations in this series. The procedures used in the special group of hospitals which make up series A are presented in the following table.  

TABLE 98.-Techniques of closure of colostomy and associated operations in 464 battle-incurred wounds of large bowel (series A)1 





Simple closure




End-to-end anastomosis


Anal reconstruction


End-to-side anastomosis


Closure distal loop


Side-to-side anastomosis


Crushing of spur


Pauchet anastomosis


Reconstruction of spur






1This information is not available in 17 cases.

Information concerning the technique of closure employed in the 180 cases (series B) in which closure was accomplished overseas was available in only 76 cases. Some method of anastomosis was used in 5 cases, and spur crushing and closure were employed in the other 71.  

Extraperitoneal operations.-Simple closure of the colostomy was employed in approximately 60 percent of the cases in which the approach was extraperitoneal. This was an entirely adequate operation when the stoma was small. In such cases, the technique used at the original operation had been tube or tangential colostomy or loop colostomy without division of the posterior wall of the colon. The preferred suture material was catgut of small size. The  


suture line was inverted at least once. If spur crushing had been necessary, closure of the anterior wall of the colon was the final step of the operation. 

End-to-end anastomosis was used in about 30 percent of the cases managed by extraperitoneal operation. Side-to-side or end-to-side anastomosis was employed in the other 10 percent. These techniques all proved satisfactory in the cases-which were not too numerous-in which it was possible to mobilize the loops of bowel adequately outside of the abdomen, and they had the great advantage of eliminating all risk of postoperative intestinal obstruction at the suture line.  

Intraperitoneal operations.-When intraperitoneal closure of a colostomy was contemplated, no attempt was ever made to crush an existing spur. Instead, the limbs were dissected free, and sufficient space was thus provided for any type of anastomosis. When the peritoneal cavity was to be entered, special efforts were also made to free the area of adhesions. These dense adhesions, which often bound the bowel intimately to the anterior abdominal wall, many times had to be treated by sharp dissection. The objective of these maneuvers was to leave the repaired intestine entirely free in its normal location.  

Simple closure of the anterior wall of the bowel was a satisfactory technique if the posterior wall of the colon remained intact. If it was not intact, end-to?end anastomosis was the preferred technique, though a number of surgeons had good results with side-to-side and end-to-side anastomoses. The Pauchet technique produced excellent results in the single hospital in which it was used. In this technique, longitudinal slits are made in the proximal and distal loops after they have been fixed with a posterior suture, and side-to-side anastomosis, as in a Finney pyloroplasty, is used to reconstruct the bowel. The large stoma thus created at the line of anastomosis serves as a safeguard against possible future obstruction at this point.  

Ileotransverse colostomy was used in a small number of cases in which damage to the right half of the colon and the small intestine had been followed by infection and massive destruction of soft tissues and in which sidetracking of the fecal current had been necessary. The operation was performed by joining the ileum, above the injury, to the transverse or descending colon. Closure of the original colostomy by end-to-end anastomosis was then possible. Leakage of irritating material seldom occurred, and failure was unusual. In most cases, because of the multiple openings, resection of some portion of the intestine was necessary. Usually the portion resected was in the ascending, transverse, or descending colon. If extensive damage had made two or more colostomies necessary, it was frequently simpler, when closure was undertaken, to resect the middle segment of the intestine and restore intestinal continuity by an end-to-end anastomosis.  

Adjunct procedures.-Herniations in the fascia (p. 345) varied from slight weakening of the tissues, which permitted insignificant protrusions of the peritoneal contents, to extensive openings which required plastic repair. Unless there was some good reason for terminating the operation after


closure has been accomplished, the repair was performed at the same time the stoma was closed. Catgut, silk, or cotton was used for sutures, depending upon the preference of the individual surgeon. The results of these operations were generally good.  

Drainage was seldom necessary when the stoma was closed, and instances were on record in which it was clearly harmful. The risk of placing drains between layers of fascia was typified by the slough which occurred in two cases in a hospital in which this technique was popular.  

In a few hospitals, it was the custom to close the wound through the fascial layer at the first operation and complete the closure 2 or 3 days later.  

Postoperative Regimen  

Although practices varied from hospital to hospital, the postoperative regimen after closure of a colostomy was deliberately kept as simple as possible. In some hospitals, nothing was given by mouth for from 24 to 48 hours. In others, fluids were given promptly and soft food shortly afterward.  

The fluid balance and nutritional status were maintained by the parenteral administration of dextrose solution, physiologic salt solution, casein hydrolysate, plasma, and whole blood, according to the necessities of the individual case.  

Constant intestinal decompression by the Wangensteen technique or by means of the Miller-Abbott tube was instituted in all cases in which distention was considered a possibility. These methods were most frequently employed after operations on the right half of the colon.  

The sulfonamide which had been employed before operation was continued for 5 to 7 days afterward. When penicillin became available, it was used by the intramuscular route in all cases in which infection seemed a possibility. The dose was 100,000 units every 4 hours. Toward the end of the war, when streptomycin became available, it was used with apparently good results in gram-negative infections.  

When vitamin K had been used before operation, it was usually continued for 3 days or more after operation.  

Anal dilatation was carried out daily for 5 days after operation.  

Early ambulation was not practiced uniformly, but there were no untoward results in any of the hospitals in which it was permitted.  

Before the patient was discharged, all the laboratory determinations which had been carried out before operation were repeated. Roentgenologic studies were also carried out, to establish the patency of the anastomosis and as a final check against overlooked foreign bodies.  

Postoperative Complications  

The discussion of postoperative complications is confined to the 464 cases (series A) from the group of hospitals which supplied especially full details on


the questionnaires from which the data on closure of colostomy were secured. They are typical of the complications which occurred in all hospitals.  

In these 464 operations, leakage occurred at the site of closure in 70 cases (about 15 percent). It was usually insignificant, and healing occurred promptly under conservative management. In 23 cases, however, secondary closure was necessary, and, in a few instances, multiple operations were necessary before the intestinal wall was securely closed. One patient required seven operations.  

Local infection occurred in 39 cases (8.4 percent). It varied from slight exudation to extensive abscess formation, and in a few cases was associated with wound dehiscence. Clostridial myositis occurred in 1 case, and in 2 others the infection was apparently the result of amebic infection.  

Hemorrhage occurred in six cases. It was usually manifested by local bleeding. In one case, the formation of a large hematoma was followed by abscess formation. In two instances, bleeding occurred in patients who had received Sulfasuxidine. Although the bleeding was controlled without difficulty in all six cases in which it occurred, the potential seriousness of this complication was obvious.  

Paralytic ileus occurred only twice, but in one instance it was associated with massive distention; secondary surgery was eventually required. The negligible incidence of ileus in this series as well as in the larger total series can be attributed to the prophylactic institution of constant intestinal decompression in so many cases.  

Intestinal obstruction of some degree was apparent clinically in about 10 percent of all cases. It was usually slight and transient and subsided spontaneously; in such cases, the cause was assumed to be edema at the anastomosis or the suture line. Adhesions were usually responsible for mechanical intestinal obstruction, though in a small number of cases it occurred because so much of the bowel wall had been infolded that an actual diaphragm had been created across the intestinal lumen. Although the 10-percent incidence of postoperative intestinal obstruction was undesirably high, only 3 of the 11 deaths which occurred in Zone of Interior hospitals after closure of colostomy could be attributed to it. Advocates of the open abdominal approach contended that the risk of obstruction was less when this technique was used.  

Peritonitis occurred in 8 cases, 1.8 percent of the total number in series A. In 1 or 2 instances, the infection was mild and localized. In the other cases, peritoneal involvement was generalized, and five deaths were attributed to it. In 2 of the 3 surviving patients, abscess formation subsequently occurred, in the pelvis and the subphrenic space, respectively, and surgical drainage was necessary.  

There was only one instance of thrombophlebitis. The infrequent occurrence of this complication was attributed by many surgeons to the emphasis placed upon movement about the bed immediately after operation, as well as to the practice, in many hospitals, of early ambulation.  



Eleven deaths are known to have followed closure of the colostomy stoma in the 2,198 cases cared for in Zone of Interior hospitals (0.5 percent). They were about equally distributed between the intraperitoneal and extraperitoneal techniques. Five, as just noted, were caused by peritonitis and three by intestinal obstruction. Liver failure, mesenteric vascular occlusion with gangrene of the bowel, and inanition accounted for the three remaining deaths.  

Two other deaths occurred before operation, in patients who were being prepared for surgery. In each instance, the cause was an intractably poor nutritional status and the resulting inanition. The cause of the single death which occurred in the 180 colostomies closed in overseas hospitals is not known. 

The 12 deaths which followed surgery in these 2,378 closures thus represented a case fatality rate of 0.5 percent.  


A study of the 2,378 patients who underwent closure of colostomy in this composite series indicates that, from the standpoint of the initial phase of management overseas, casualties with wounds of the colon fared better in World War II than in any previous war for which records are available. They were removed from the battlefield with remarkable speed. They were prepared for operation with efficient resuscitative measures and without loss of time. Finally, the routine practice of colostomy in wounds of the colon saved innumerable lives which would undoubtedly have been lost in any previous war. The principle of exteriorization of the damaged colon was so firmly established in World War II that no further controversy can possibly arise concerning its use in any case in which the bowel can be sufficiently mobilized to permit it. It is a simple procedure, which can be carried out safely and promptly by surgeons of limited experience. It entirely obviates the need for difficult surgical decisions and for extensive resections and anastomoses in patients whose status is poor and who must be operated on under field conditions. Certain aspects of the operation deserving special comment are given in succeeding paragraphs.  

Diversion of the fecal stream.-One of the items on the questionnaire circulated in general hospitals of the Zone of Interior concerned the possibility of accomplishing total diversion of the fecal stream with an ordinary colostomy. There was not complete agreement in the 19 replies received. Four surgeons did not think that it could be accomplished with a tube or tangential colostomy. Four thought that it could not be, or probably could not be, accomplished with a loop colostomy, one thought that it could be, and one other thought that it probably could be. Four thought that it could be accomplished by spur colostomy or probably could be, and one thought that it could not be. The four remaining surgeons thought that total diversion of the fecal stream could be accomplished with a Devine colostomy. The infer-


ence, therefore, is that only by the use of the Devine technique (fig. 39, p. 354) or some other special technique (fig. 41) in which there is actual separation of the stoma could one be absolutely certain of complete diversion of the fecal stream. Some surgeons regarded the creation of a second colostomy of the spur type, proximal to the sigmoid colostomy, as the better method of preventing contamination of the distal segment. The transverse colon was the preferred location for the second colostomy.  

FIGURE 41.-Extensive damage to descending colon, with complete severance of gut, requiring resection at first operation after wounding. The end of the colon was implanted into the abdominal incision, creating a satisfactory colostomy. Closure by primary anastomosis after reopening abdominal cavity.


There was also some difference of opinion concerning the absolute necessity for diversion of the fecal stream in the event that a fistula existed distal to the colostomy, particularly in the sacral region or the buttocks. The majority of the 62 surgeons who replied to this question considered it absolutely essential, or advisable, or desirable. The others did not consider it necessary. There was general agreement that a short, direct fistulous tract would not heal if fecal contamination continued, particularly if the tract traversed bone which was the site of an osteomyelitis. There was also general agreement that even if long, tortuous tracts healed under conservative management, sometimes without complete diversion of the fecal current, expectant treatment was not justified. The healing process could be expected to cover many months, and healing was unlikely until dead bone and foreign bodies had been removed. The consensus was that under these circumstances there were three courses of action: (1) Complete decontamination of the fistulous tract by revision of the colostomy into a modified Devine colostomy; (2) creation of a proximal colostomy in the transverse colon; or (3) closure and inversion of the fistula into the rectum, with complete unroofing of the tract, excision of everted rectal mucosa, and removal of all foreign bodies and of the coccyx or part of the sacrum.  

Technical considerations.-The trend, as the war progressed, to substitute intraperitoneal for extraperitoneal techniques of closure has already been mentioned. The untoward results which sometimes followed crushing of the spur had much to do with the change in practice. In civilian surgery, the management of the spur presents no problem, for the surgeon who creates the colostomy crushes it himself. In military practice, the surgeon who constructs the colostomy almost never crushes the spur. The hundreds of patients admitted to general hospitals in the Zone of Interior with unclosed colostomies were operated on by surgeons who, however experienced they might be, had no knowledge of the initial surgery in any case beyond what was contained on the field medical record. These data were often entirely inadequate. Even when the records were reasonably complete, the conditions were not at all the same as if the surgeon who was to crush the spur had been responsible for the creation of the colostomy or had at least observed its creation. Furthermore, the unavoidable haste with which many operations were necessarily performed in forward hospitals, combined with the limited experience of some of the operating surgeons, militated against the construction of an ideal spur, especially when it was known that a simple loop colostomy would suffice.  

In addition to these fundamental considerations, unfortunate experiences in dealing with poorly constructed spurs played a large part in the substitution of the intraperitoneal for the extraperitoneal technique. The application of a clamp or an enterotome often produced severe pain, nausea, and vomiting. Peritonitis sometimes followed the accidental crushing of large mesenteric vessels. Other complications included infection, fistula formation into the ileum or jejunum, paralytic ileus, intestinal obstruction, and prolonged edema


and cyanosis of the intestinal mucosa. In two instances in this series, benign neoplasms occurred at the site of the spur.  

Most serious of all, rotation of the loops had taken place in nearly a hundred cases in the total series, and the existing relationships could not be determined before the enterotome was applied. As a result of this combined experience, more and more surgeons abandoned attempts to crush the spur and resorted, instead, to intraperitoneal closure of the colostomy.  

The data available in this analysis do not supply adequate statistical proof concerning the relative frequency of postoperative complications after extraperitoneal versus intraperitoneal colostomy closure. The clinical impression was, however, that complications were fewer when the intraperitoneal technique was used. There was considerable logic in this point of view: 

1. The peritoneal surfaces had been vaccinated before operation, because the colostomy had practically always existed for months before closure was attempted.  

2. All the patients were young and free from organic disease. As a rule, they were in excellent condition except for the colostomy and its complications. 

3. Chemotherapy and antibiotic therapy probably played some part in the good results, by preventing invasive infection. Streptomycin was regarded as particularly useful in this respect.  

4. With the intraperitoneal technique, the restoration of the bowel more nearly approached normal because the anastomosis could be performed more accurately and all adhesions could be obliterated. At the conclusion of the operation, therefore, the repaired bowel lay free in its normal location, and postoperative intestinal obstruction was less likely to occur.  

The experience in this series showed that injuries to the right side of the colon presented special problems, particularly if the small intestine had also been damaged. It was not felt, however, that the use of the small bowel with the colon to create a spur colostomy offered any solution of the problem. Wounds in this area were best dealt with by resection of the damaged cecum or ascending colon, with direct end-to-end anastomosis between the ileum and colon. After this had been done, a single-barreled external mucous fistula was created in the proximal colon.  

Injuries of the rectum also furnished special problems. Sacral or perineal colostomy, employed in less than 1 percent of the cases in this total series, was used chiefly when severe damage to the rectum was associated with partial or total destruction of the anal sphincter. Even as a temporary expedient, however, this was not regarded as good surgery. The nursing care in such cases was extremely difficult, irritation of the skin was almost unavoidable, and the patients themselves objected to this type of colostomy for a number of reasons, including purely esthetic reasons.  

The best plan in injuries to the fixed portions of the rectum and colon was immediate repair of the anal sphincter, combined with posterior drainage and followed by proximal colostomy, preferably in the transverse or descending


colon (fig. 42). Loop openings were separated, to avoid continued contamination of the distal loop. This precaution was particularly important if the pelvis or the sacrum had been injured or if there had been extensive damage to the rectum or bladder below the peritoneal attachment (fig. 43).  

FIGURE 42.-Proximal colostomy in descending colon for wounds of sigmoid and rectum. After the damaged colon had been resected, the terminal loop was closed and an open proximal loop was used to form the colostomy (insert).  

When a patient was received with a sacral colostomy and with damage to the anal sphincter which had not been repaired or the repair of which had been unsuccessful, a secondary colostomy was created at once, in the sigmoid or transverse colon, before plastic surgery on the anus was undertaken. Even if this colostomy had to become permanent, because of failure of repair of the anus, it was possible to close the sacral opening, and the unfortunate casualty was left in much better condition than he had been in after the first operation.  

A consideration of the errors observed in the colostomies in this series leads to the conclusion, fortified by the entire experience of World War II, that, in injuries of the colon, the damaged loop should always be brought out through a stab wound placed laterally to the main incision and that it should be fixed in this location by sutures which do not pass through the bowel wall (fig. 44).


FIGURE 43.-Proximal colostomy in transverse colon for retroperitoneal wounds of sigmoid and rectum. The shell fragment also traversed a sacral vertebra and the bladder. Insert shows final double-barreled colostomy.  

Satisfactory support can be provided by a glass tube of the kind in which catgut is sterilized or by some similar blunt object placed beneath the mesenteric border. Rubber tubing can also be used. Spur formation is not essential. In fact, observations in Zone of Interior hospitals, and attempts to crush spurs in which the relationship to surrounding structures had changed, suggested that any technique which required creation of a spur might be actually undesirable.  

The revisions and relocations of the original colostomies performed in this series would have been entirely unnecessary if the general principles set forth in this chapter had been followed in the creation of the original opening. Furthermore, some, at least, of the problems encountered at every hospital, such as retraction of the loops of bowel, undue protrusion of the bowel through exces-  


FIGURE 44.-Exteriorization of damaged sigmoid colon. No attempt has been made to create a spur. Insert shows resulting eversion of mucous membrane.  

sively large fascial openings, intestinal obstruction caused by adhesions, continued contamination of a distal fistulous tract, and retained foreign bodies, would have presented no difficulties if these principles had been followed.  

Chemotherapy and antibiotic therapy.-Although the colon is one of the most heavily contaminated structures in the body, there was considerable divergence of opinion as to the necessity for, and the value of, supplemental chemotherapy and antibiotic therapy in closure of colostomies. One or the other of these methods, and sometimes both, were used in most cases in this total series, but about a third of the surgeons who performed the operations were not convinced of the value of any of these agents, about 10 percent considered them of no value, and a few regarded them as harmful.  

Most of the surgeons who used these drugs thought that their chief value, at least as far as the then newly developed preparations Sulfathalidine (phthalyl?sulfathiazole) and Sulfasuxidine were concerned, was to insure a clean intestinal surface at the time of operation. They questioned their antibacterial effect and were inclined to attribute the striking absence of infection in these cases to the fact that sufficient time had elapsed since the first operation to permit some degree of autovaccination of the peritoneal surfaces. In spite of these doubts of the value of the routine use of these agents, it was the general practice to administer Sulfasuxidine and Sulfathalidine orally for several days before operation to all patients who had had peritonitis or intra-abdominal abscesses, as well as to those who were likely to require extensive resection of the colon.


When penicillin and streptomycin became available, they were used on the same indications.  

Almost all of the available sulfonamides except sulfanilamide were used in this series, but Sulfasuxidine and sulfaguanidine were the most popular. Sulfathalidine was not available at all hospitals. All routes of administration were employed, including the local application of the powdered drug, which was used at operation in about 10 percent of all cases. Stool examinations made in the cases in which chemotherapy was used showed that, while there was not complete destruction of all micro-organisms, the colonies were always greatly reduced in number.  

Irritation of the skin about the colostomy was an annoying consequence of sulfonamide therapy (fig. 45). It could usually be controlled by the use of  

FIGURE 45.-Right-sided colostomy. Cecum and ascending colon exteriorized at time of injury. Note irritation of skin around colostomy, which was common in right-sided colostomies.  


aluminum or kaolin paste. Secondary hemorrhage, another possible complication, was always potentially serious. Its prevention required careful prothrombin studies before operation and the administration of vitamin K in the immediate preoperative and postoperative period.  

Penicillin was used extensively after it became available in the Zone of Interior in the spring of 1944. It was thought to be of great value in the prevention of infections caused by gram-positive organisms. Streptomycin did not become available until later and then only in hospitals designated for special studies. It was regarded as of great value in the prevention and treatment of gram-negative infections.  


The most important lesson to be derived from the large number of colostomies upon which this chapter is based is that the closure of the stoma created in the management of a battle-incurred injury of the colon is a safe and relatively simple procedure. In this series, almost every possible unfavorable circumstance was present from the moment of wounding. Initial treatment was carried out in a forward hospital, under field conditions. Even in fixed hospitals, treatment was conducted under the stress and strain of military conditions. The wounds were frequently of great gravity. Other wounds were frequently associated. In spite of these facts, the case fatality rate for closure of the colostomy was fractional, complications were relatively few, and practically every patient who survived was returned to normal activities and, from the standpoint of his intestinal injury, was in excellent condition on discharge.  

In World War II, improvements in the technique of intestinal surgery, the intelligent use of replacement therapy and of chemotherapy and antibiotic therapy, and increasing experience on the part of responsible surgeons made it possible to close colostomies within relatively short periods after their creation. The wartime experience suggests that in future wars it may be possible to accomplish closure in all but the most complicated cases in overseas hospitals, without evacuating the patients to Zone of Interior hospitals. From a military standpoint, this would be a most important advance, for it would permit soldiers with wounds of the colon to return to full duty after relatively brief periods of disability and convalescence.