CHAPTER XIX
Wounds of the Jejunum and Ileum (1,168 Casualties)
W. Philip Giddings, M. D., and John R. McDaniel, M. D.
Between 1 January 1944 and 8 May 1945, the surgeons of the 2d AuxiliarySurgical Group treated 1,168 casualties with injuries of the jejunum and ileum,of whom 345 (29.5 percent) died (table 50). These 1,168 injuries represent 37.0percent of the 3,154 abdominal injuries observed over the period of time stated.
TABLE 50.- Essential data in 1,168 wounds of jejunum and ileum
Type of wound | Cases | Frequency | Deaths | Case fatality rate | |
In 3,154 abdominal injuries | In 1,168 jejuno-ileal injuries | ||||
|
| Percent | Percent |
|
|
Univisceral | 353 | 11.2 | 30.2 | 49 | 13.9 |
Multivisceral | 815 | 25.8 | 69.8 | 296 | 36.3 |
| 1,168 | 37.0 | 100.0 | 345 | 29.5 |
NATURE OF INJURY
Included in these injuries of the jejunum and ileum are six in which the abdominal wall was not penetrated. All were incurred in vehicular accidents and all resulted in rupture of the small bowel. Another patient, who sustained a severe transfixing thoracoabdominal wound, was injured when he was impaled on the stump of a small tree by the blast of an exploding shell. With these seven exceptions, all the wounds in the series were caused by high-explosive fragments or missiles from small arms. There were no bayonet or stab wounds.
The lesions varied from pinpoint perforations and small contusions to extreme maceration and destruction of the greater portion of both ileum and jejunum. For convenience of discussion, the injuries have been classified under four headings: (1) Injury to the intestinal wall, (2) injury to the mesentery, (3) perforating injury, and (4) transections (table 51). Wide extremes of severity were noted in each group, and as a rule 2 or even 3 lesions were seen in combination.
242
TABLE 51.- Frequency ofvarious types of injury in 1,168 wounds of jejuno-ileum
Injury | Cases | Injuries | Average frequency per case |
Injury to intestinal wall | 27 | 31 | 1 |
Injury to mesentery1 | 30 | 30 | 1 |
Perforation2 | 1,083 | 4,589 | 4 |
Transection | 213 | 361 | 2 |
1Severe enough to require resection.
2The arbitrary figure of 5was used when perforations were recorded as "multiple," withoutstatement of the precise number. Whenever the precise number was recorded, itwas used.
Injury to the intestinal wall-Trauma tothe intestinal wall included contusions and lacerations of the serosal orseromuscular coats which did not perforate the mucosa. This type of injury wasnot frequent (table 51). Contusions, which were seldom more than 2 cm. in diameter, varied fromareas of slight ecchymosis to circumscribed areas of gangrene. A contusion ofthe bowel wall implied, by its very nature, that the missile which had caused ithad reached the end of its flight and was traveling at such low velocity thatits impact against the intestine was sufficient to stop it. In this type ofinjury, the missile either was embedded in the wall of the bowel or lay free inthe peritoneal cavity.
Lacerations of the intestinal wall were caused bytangential impact, and no inferences are, therefore, permissible concerning thevelocity of the causative missile. They usually occurred in association withperforations of the lumen in other portions of the bowel. Theoretically, a laceration might be of anylength. Practically, in the absence of a perforation, it was seldom more than 2 cm. long. Some lacerations were merely breaks in the serosa.Usually there were evidences of contusion about the margins of a laceration, andthe laceration-contusion type of injury was considered a potential site forfuture perforation.
Injury to the mesentery-Injuries of the mesenteryoccurred both at the mesenteric attachment and at a considerable distance fromthe bowel. They varied from small hematomas or peritoneal lacerations to rentswhich extended across the mesentery to its root. Some damage to the adjacentmesentery nearly always occurred in transections of the bowel and inperforations involving the mesenteric border. Bleeding had usually ceased whenthe casualty was first seen, but in some instances bleeding was still active.
Perforations of the mesentery other than those adjacent toinjuries of the bowel were frequent. They were usually simplethrough-and-through holes. Hematomas were sometimes present also, but in themajority of cases there was no indication that a vessel of major consequence wasinvolved. The incidence of mesenteric damage sufficiently extensive tonecessitate intestinal resection was remarkably low; in all 1,168 injuries ofthe jejunum and ileum there were only 30 cases in which resection was requiredbecause of vascular impairment
243
from mesenteric injury. The circumstances of this type ofinjury, including the effects of energy transmitted to the tissues and thenatural tendency to clot formation in a lacerated blood vessel, would beexpected to produce thrombosis. Thrombi were commonly found protruding from theends of severed vessels in the mesentery, even when the vessels were ofconsiderable size, but the thrombotic process was invariably restricted to theimmediate area of damage. There was no instance in the series of extensivemesenteric thrombosis (in the usual clinical sense of the term) in which thepathologic process could be regarded as caused primarily by trauma to themesentery.
Perforating trauma of the intestine-By far the commonesttype of injury to the small intestine was perforation into the lumen. Thisoccurred in 1,083 of the 1,168 injuries, and multiplicity was the rule, theaverage number of perforations per case being four (table 51).
Small perforations occasionally seemed to have been caused byindriven fragments of bone, especially when the missile had entered the abdomenthrough the ilium. Generally speaking, the size and shape of the rent in thebowel were determined by the corresponding characteristics of the missile, aswell as by its velocity and the direction of its flight. Through-and-throughperforations and complete transections were both frequent.
Under some conditions of injury, isolated perforations might be found at aconsiderable distance from the major trauma, though as a rule multiple injurieswere in close proximity to each other, and damage was ordinarily confined to asegment 1 or 2 feet in length, or even less. These phenomena were explained bythe fact that in the usual case, intestinal damage was caused by a singlemissile, the effect of which was confined to its line of flight. Scatteredperforations were caused by missiles which traversed major diameters of theabdomen; by separate missiles with different wounds of entry; or by multiplefragments which, though they entered through a common wound, pursued divergentcourses within the abdomen.
Individual lesions varied from tiny holes to gashes 6 inchesor more in length. When the perforations were of any size, the findings weresimilar. The tissues about the edges were contused and ecchymotic. The marginswere jagged. Mucosa usually pouted from the wound. Bleeding from the intestinalwall was sometimes free, and considerable amounts of blood were seen in theperitoneal cavity when the bowel wall was the only apparent source ofhemorrhage. Small perforations were sometimes almost completely sealed off bypouting mucosa, so that there was a minimum of soiling of the cavity. In suchcases, there was a bluish discoloration of the bowel, from intraluminalbleeding. Less commonly, blood and intestinal contents were extensively spilledfrom small perforations, as in the following case:
Case 1.1-A 29-year-old infantryman was wounded by a machinegunbullet which entered the left lower abdomen. The ilium was fractured at thewound of entrance and the greater trochanter of the left femur was fractured atthe wound of exit. When the patient was
1This case appears in table 54 as case18.
244
admitted to a field hospital 20 hours later, he was in severe shock and wasanoxic and disoriented.
Operation was performed 23.5 hours after wounding, after the administrationof 1 unit of plasma and 5 pints of blood. Although the injury was limited to asingle small perforation of the jejunum, which was readily repaired by suture, ageneralized, fulminating peritonitis was already present, and the surgeon notedon the operating sheet that the spillage of intestinal contents was the mostextensive he had ever seen.
After operation, the patient remained extremely toxic.Oliguria developed and progressed to anuria. Death occurred on the eighthpostoperative day. Post mortem examination revealed an acute, fibrinopurulentgeneralized peritonitis, with a right-sided subphrenic abscess.
Transection.-Transections of the intestine, which were fairlycommon (table 51), represented the extreme form of perforations of the jejunumand ileum. The most characteristic feature of these injuries was that they werenearly always accompanied by extensive spillage.
State of the peritoneum-The small bowel is well vascularized, and thefrequency of hemoperitoneum in injuries of the jejunum and ileum was consistentwith that fact. In some instances, 2,000 cc. of blood, or more, were in theperitoneal cavity, usually as the result of massive bleeding from the mesentery.At the other extreme, bleeding was occasionally minimal.
The peritoneal cavity was usually contaminated to some degree with contentsof the small bowel, though in an occasional case no gross soiling was apparent.The degree of contamination was usually compatible with the size and number ofthe perforations. The peritoneal reaction was fairly constant in relation to thetime at which surgery was undertaken. When operation was performed within 6hours of injury, there was usually no visible reaction. Violent exudativeperitonitis was the rule only in patients seen after a relatively long timelag(12 hours or more). In casualties who survived longer than 24 hours withoutoperation, early walling off of the perforation with loops of adherent bowel waslikely to have occurred. These findings, of course, were not invariable. Severegeneralized peritonitis was sometimes seen in patients treated early, andlocalization was also occasionally seen in early cases. Definitive peritonitiswas recorded at operation in only 50 of the 353 univisceral wounds of the jejuno-ileum,a figure which is undoubtedly too low.
Evisceration.-Evisceration of the jejunum and ileum occurredin 153 instances in the whole series of 3,154 abdominal injuries. Trauma to theinvolved bowel or its mesentery required repair in 126 cases. In the remainingcases, evisceration was incidental and required no treatment other thanreduction. In only one instance was resection necessary because of strangulationof the eviscerated loop.
Associated injuries-Two hundred and fifty-two of the 1,168 casualtieswith injuries of the jejunum and ileum had major associated peripheral wounds,and 202 had minor associated wounds. Another 143 casualties had penetrating orperforating wounds of the thorax, of which 94 (8.0 percent of all injuries ofthe jejuno-ileum) were thoracoabdominal. About half (597) of all the casual-
245
ties with injuries of this portion of the small bowel thus had associatedwounds, about two-thirds of which (395) were of a major character.
DIAGNOSIS
The preoperative diagnosis of wounds of the jejunum and ileum was madechiefly on probabilities. There were no criteria by which their existence couldbe established or excluded before operation. Indriven fragments of bone and theconcussive effect of missiles passing extraperitoneally were known to causeperforation or rupture, and whenever these circumstances were present it was notregarded as safe to assume that the bowel had not been injured. Diagnosis,therefore, was usually established or, occasionally, disproved by exploration ofevery case in which the circumstances of wounding and the clinical signs andsymptoms indicated possible intra-abdominal injury.
PREOPERATIVE ROUTINE
The preoperative management of wounds of the small intestinediffered in no respect from the preoperative management of other abdominalinjuries. An essential part of the routine was the introduction of a Levin tubeinto the stomach, followed by aspiration of the gastric contents.
TREATMENT
Certain general principles were universally followed in themanagement of wounds of the jejunum and ileum. These included nontraumatichandling of all tissues, maximum protection of the bowel from exposure, the useof fine suture material, and as rapid operation as was consistent with thoroughexploration and necessary repair procedures. In many instances, the control ofhemorrhage and operations on other viscera took precedence over the repair ofthe small-bowel lesions. Time was always saved by a preliminary evaluation ofthe total damage. About a third of the surgeons in the group thought itexpedient, in the investigation of possible damage to the small bowel, to bringit out through the surgical incision and examine it outside of the abdomen.Complete examination of the jejuno-ileum and its mesentery was facilitated bythis technique, and those who practiced it felt that, if it were carried outgently and expeditiously, it was not shocking to the patient. Its use waslimited to cases in which the damage was obviously extensive and the lesionswere obscured by some degree of spillage.
Contusions and lacerations of the intestinal wall wererepaired and reenforced by a peritonealizing linear or purse-string suture, ofwhatever material the individual surgeon might prefer. With this exception, thetype of repair was related to the type of injury.
Simple through-and-through perforations of the mesentery weresutured, to reperitonealize the raw surfaces, but dissection was carried out inthe presence of large hematomas or continuing bleeding. After the hematoma hadbeen
246
evacuated and hemorrhage controlled, the mesentery was closedby suture. As already mentioned, intestinal resection because of mesentericvascular damage was required in only 30 cases in the series. In these patients,the usual criteria of viability of the bowel were observed, and resection wasperformed in accordance with recognized surgical principles.
A wide variety of techniques was used in the repair of perforating lesions ofthe bowel. The basic principle of management was to select the most conservativeprocedure compatible with secure repair and preservation of an adequate lumen.Very small perforations were usually closed by purse-string suture. Largerperforations were sutured transversely. Debridement of the traumatized edges wasfrequently omitted. If it was regarded as necessary, it was as conservative aspossible. It was sometimes found convenient to convert two perforations into asingle lesion, particularly if they were located close together in the samevertical plane. The single defect which resulted could be sutured in less timethan was required to suture two separate perforations, and in the end there wasless kinking of the bowel.
Resection and anastomosis-The indications for resection (tables 52 and53) were usually clear cut. It was clearly mandatory when segments of the bowelwere hopelessly macerated. It was also required when the bowel had been avulsedfrom its mesentery. Multiple adjacent perforations separated by areas of intactbowel always presented problems. Under these circumstances, some surgeonsrepaired each perforation separately. Others, if the involved segment was nottoo long, resected it in toto, fearing that multiple suture lines so closetogether might compromise the lumen and lead to intestinal obstruction fromkinking.
TABLE 52.-Comparative results ofanastomotic and suturerepair in 1,117 wounds of jejuno-ileum1
Type of injury | Total series | Anastomosis2 | Suture | ||||||||
Cases | Deaths | Case fatality rate | Cases | Frequency | Deaths | Case fatality rate | Cases | Frequency | Deaths | Case fatality rate | |
|
|
|
|
| Percent |
|
|
| Percent |
|
|
Univisceral | 348 | 47 | 13.5 | 135 | 38.8 | 25 | 18.5 | 213 | 61.2 | 22 | 10.3 |
Multivisceral | 769 | 281 | 36.5 | 347 | 45.1 | 155 | 44.7 | 422 | 54.9 | 126 | 29.9 |
| 1,117 | 328 | 29.4 | 482 | 43.2 | 180 | 37.3 | 635 | 56.8 | 148 | 23.3 |
1Omitted from this table are nonperforating injuries and all injuries inwhich ileocolostomy was the only treatment.
2In 54 cases, anastomosis wasperformed without resection.
Various techniques of anastomosis were employed (table 53).The majority of surgeons preferred an open, two-layer, end-to-end methodwhenever it was practical. Closed methods were used when soiling was minimal. Afew surgeons routinely employed the closed silk technique. Side-to-sideanastomosis was usually reserved for injuries of the lower ileum or for cases inwhich, after extensive resection, there was marked discrepancy in the size ofthe lumens to be
247
Technique of anastomosis | Cases | Frequency | Deaths | Case fatality rate |
|
| Percent |
|
|
End-to-end | 377 | 88.1 | 120 | 31.8 |
Side-to-side | 34 | 7.9 | 16 | 47.1 |
Not recorded | 17 | 4.0 | 9 | 52.9 |
| 428 | 100.0 | 145 | 33.9 |
anastomosed. Two surgeons reported successful results with asingle-layer type of anastomosis. Two others preferred triple layers. Running orinterrupted sutures of catgut and interrupted sutures of fine silk or cottonwere widely used. As a general rule, the same type of suture and the same suturematerial employed for anastomosis were also used for the repair of perforations.
Four hundred and twenty-eight patients were subjected to 466resections, which included 35 double resections with 13 deaths (37.1 percent)and 3 triple resections with 2 deaths. The case fatality rate of 47.1 percent(table 53) for the 34 side-to-side anastomoses is a reflection of the severityof the trauma in the cases in which it was used, rather than an index of therisk of this technique. It was unusually employed only in massive resections.
In the 394 cases in which data on this point were available, the length ofthe resected segments ranged from 2 inches to 12 feet. In 272 cases, the lengthwas between 2 and 12 inches. In the other 122 cases, it averaged 34 inches. In 4 cases, segments more than 8 feet in length were resected.The case fatality rate in the 122 cases in which the resected segments measured12 inches and more in length was 35.2 percent (43 deaths), which is in closeagreement with the case fatality rate of 33.9 percent for all resections (table53).
The case fatality rate of resection in injuries of the jejuno-ileum (33.9percent) was higher than the rate of repair by suture (23.3 percent). The ratefor anastomosis without resection (54 cases, 35 deaths, 64.8 percent) was veryconsiderably higher. The ratio of the case fatality rate of repair byanastomosis to repair by suture was 1.79 in univisceral cases (18.5 and 10.3percent respectively) and 1.50 in multivisceral cases (44.7 and 29.9 percent,respectively). In other words, anastomotic repair carried a higher case fatalityrate than suture repair in both groups of cases, and of approximately the sameorder in each. The multiplicity factor furnishes no explanation of thesefigures, since the case fatality rate of anastomosis was higher in univisceralcases.
The results in this series therefore confirm the general impression thatanastomotic repair of the small bowel is more hazardous than repair by suture,regardless of the circumstances under which it is employed. On the other hand,the principal indication for intestinal resection in this series was alwaysextensive
248
trauma. The procedure was usually undertaken only when the bowel was inshreds and beyond hope of conservative repair. The higher case fatality rate forthis procedure is therefore probably as much a reflection of the severity of theinjury as of the relative risks of suture repair and intestinal resection.
The following case histories are presented as typical:
Case 2-A medical corpsman wasbrought into a field hospital about 2 hours after he had sustained a severepenetrating gunshot wound of the left abdomen, with an extensive evisceration ofsmall intestine through a large defect in the abdominal wall. He was in severeshock, and neither blood pressure nor pulse could be obtained. After 2,500 cc.of blood had been administered as rapidly as possible, the blood pressure roseto 52/40 mm. Hg and the pulse became perceptible, though it still could not becounted.
Laparotomy was performed 3 hours after wounding. The root of the mesenterywas found avulsed, this injury being the source of massive, persistent bleeding.The missile had perforated the left mesocolon and lay in the lumbar musculature.There were multiple lacerations and transections of the ileum and jejunum, andbecause of vascular impairment it was necessary to resect 10 feet of smallintestine. Several perforations of the jejunum were also sutured. The patient'scondition began to improve as soon as hemorrhage had been controlled, and by theend of the operation the blood pressure was 104/60 mm. Hg. A transfusion of 500cc. of whole blood was given on the operating table.
Recovery was uneventful. The Levin tube was removed on the sixth day, and onthe same day the patient has a spontaneous bowel movement. He was evacuated ingood condition on the 13th postoperative day. A month later, it was learned thathis improvement had continued and that he was about to be evacuated to the Zoneof Interior.
Case 32-An 18-year-old German prisoner of war was admitted toa field hospital about 20 hours after he had sustained a penetrating wound ofthe abdomen from a shell fragment. He was in severe shock that did not respondto vigorous resuscitation therapy. Operation was undertaken about 24 hours afterwounding. The peritoneal cavity was extensively contaminated with small-bowelcontent and was the site of a plastic peritonitis. The color of the bowel wasunhealthy, and it was seriously mangled in three separate areas. A total of 3feet was resected in three segmental procedures, and several other perforationswere repaired by suture. End-to-end anastomoses were performed, the highestabout 4 inches below the ligament of Treitz.
The man's condition became progressively worse duringoperation, in spite of continuous transfusions of whole blood, and deathoccurred about 5 minutes after the abdomen had been closed.
Case 4-A 42-year-oldartilleryman, wounded by a shellfragment which penetrated the abdomen through the left lower quadrant, arrivedat a field hospital in good condition. He was immediately given 500 cc. ofblood. Roentgenologic examination disclosed a metallic foreign body in the rightlower quadrant of the abdomen. Operation, performed 8 hours after wounding,revealed "multiple perforations of very large size" in the smallbowel. Resection of three separate intestinal loops, with end-to-end anastomosis,was performed in the jejunum and in the upper and the lower ileum. The largeshell fragment visualized in the roentgenogram was removed from the wall of theileum.
The postoperative course was entirely uneventful. The patient was in goodcondition and was taking liquids by mouth when he was evacuated on the eighthpostoperative day.
Enterostomy.-Enterostomy was employed only twice in the initial management ofthese 1,168 injuriesof the jejunum and ileum. In the first instance, a small perforation at theileocecal junction was managed by tube ileocecostomy, the procedure beingselected because of the peculiar anatomic site of the injury. The history of thesecond patient follows:
2This case appears in table 54 ascase 9.
249
Case 5-A German prisoner of war was admitted to afield hospital 3 days after he had been wounded by a shell fragment which hadpenetrated the right lower quadrant of the abdomen. His condition, except fordehydration, appeared good. There was tenderness in the right lower quadrant ofthe abdomen, and a thin, watery discharge exuded from the wound in this area.
At operation, the wound was converted, by lengthening it,into a modified gridiron incision. A large abscess cavity anterior to the cecumcontained a perforated loop of ileum, and tube ileostomy was performed throughthe perforation. Convalescence was fairly smooth, but undigested food particleswere observed in the discharge from the ileostomy. A note was made that theperforation was probably higher in the intestine than it had seemed to be atoperation and that an attempt at closure might have been preferable toenterostomy. The patient was evacuated on the 10th postoperative day. This caserepresents 1 of the 9 instances of fistula formation in the series (p. 251).
Management of the contaminated peritoneal cavity-Two surgeonsin the group routinely placed drains to the peritoneal space in the presence ofcontamination, whether from the small bowel or from other sources. The remainderwere opposed to drainage of the cavity. A very few surgeons believed that lavageof a severely contaminated peritoneal cavity prior to closure of the abdomenaided in the removal of gross material which could not be evacuated by othermeans, but this method was used too infrequently to permit an evaluation ofeither its efficacy or its possible risks.
According to the records, sulfanilamide crystals orpenicillin sodium were used intraperitoneally, alone or in combination, in 59 percent of the injuries of thejejunum and ileum. The actual figure is thought to be higher. Whether or notthey were used in this manner rested with the individual surgeon. No opinion wasarrived at concerning the efficacy of intraperitoneal chemotherapy andantibiotic therapy in injuries of the small intestine.
POSTOPERATIVE COMPLICATIONS
The recorded complications of injuries of the jejunum and ileum during thetime the patients were under observation in forward hospitals wereextraordinarily few. It is unlikely that serious complications were notrecorded. On the other hand, the figures are perhaps not really representative,since complications undoubtedly developed in many cases after evacuation.
Ileus, distention, and vomiting apparently were almost universal amongpatients with abdominal injuries in World War I. In World War II, the routineuse of nasogastric decompression practically eliminated these dangerouspostoperative complications. Patients with injuries of the jejunum and ileumwere intubated with the Levin tube prior to surgery, and three-bottle siphonagesuction was instituted as soon as they reached the postoperative ward. The tubewas left in place for from 3 to 8 days after operation. Considerable differencesof opinion existed as to how long decompression was required after operation,but most surgeons favored a period of 3 to 6 days. Removal of the tube wasdetermined by the standard clinical criterion; namely, return of intestinalfunction as manifested by audible peristalsis, the passage of flatus
250
and similar phenomena. After the tube had been removed, oralnutrition was cautiously increased in accordance with the patient's ability totolerate food.
The Miller-Abbott tube, although it was always readily available,was only occasionally used in the management of jejuno-ileal injuries. It wasthe consensus of the surgeons of the group that there were few indications forintubation of the intestine since in most instances adequate decompression couldbe obtained by the Levin tube. This was fortunate, for in the few cases in whichthe Miller-Abbott tube was used it was difficult, under field conditions, to getit past the pylorus.
Routine postoperative care prevented the development of most othercomplications. Careful attention was given to the maintenance of the fluid andelectrolyte balance, to the hematocrit level, and to nutrition. Patients wereusually kept in Fowler's position until peritonitis had definitely subsided.
Intestinal obstruction.-Of the 1,168 patients with injuries of the jejunum and ileum, 20(1.7 percent) presented symptomsof mechanical intestinal obstruction while they were still in forward hospitals.Ten had sustained severe multivisceral wounds; peritonitis had been present atoperation in two cases; and gross contamination of the cavity had been recordedin six others. Five of the twenty patients had undergone suture repair, fouranastomosis without resection, and eleven resection and anastomosis. One patienthad had a double resection, and two others had had resections as well as simpleanastomoses of transections. In 10 of the 15 cases in which anastomosis wasperformed, suture repair was also necessary for other injuries.
Obstructive symptoms, in the 17 cases in which these data wererecorded, appeared between the 3d and the 32d day after operation. If the 2cases are omitted in which symptoms became apparent on the 32d day, the averagetime of appearance of symptoms is lowered to between the 6th and 7th days.
Sevenof the twenty patients died (35.0 percent). In the six instances in whichnecropsy was carried out, the cause of the obstruction was found to be adhesionsin three cases, and kinking of the anastomosis, edema at the anastomosis, andleakage from the anastomosis with subsequent peritonitis in one case each. Inthe single fatality in which autopsy was not performed, death followeddevelopment of a small intestinal fistula which was attributed to leakage at ananastomosis.
The method of management of the obstruction was mentioned inonly 1 of the 7 fatal cases. In this instance, although the Miller-Abbott tubedid not pass the pylorus, decompression of the obstructed bowel was successful,and death was caused by severe atypical pneumonia and hepatitis of unknownetiology. It seems safe to assume that conservative measures were also employedin the other fatal cases; had operation been performed, it would certainly havebeen recorded.
Ten of the thirteen patients who survived were treatedconservatively, reinstitution of nasogastric decompression being the principaltherapeutic measure. Data are not available concerning the length of time it wasrequired.
251
The other (three) patients were operated on when conservativetherapy proved unsuccessful. In one case, obstruction was the result ofadhesions, in one of adhesions complicated by multiple abscesses, and in one ofadhesions complicated by volvulus of the ileum.
Intestinal leakage and fistula formation-Leakage fromthe small bowel occurred in 12 cases after operation and was followed in 9instances by fistula formation. In 6 of the 12 cases, there were severe woundsof other hollow viscera. In the 8 cases in which these data were recorded, thecomplication developed between the 6th and the 26th postoperative days, theaverage time being 13 days.3
As always, leakage from the small bowel proved a serious matter. It was fatalin all 3 cases in which fistula formation did not occur, and in 2 other cases inwhich fistulas developed. In 4 of the 5 fatalities, death was attributed toperitonitis; the cause was not stated in the fifth case. In each of the threeautopsied cases, leakage was found to have occurred at a suture line, and itseems reasonable to assume that the same accident happened in the two otherfatal cases. In 1 of the 3 autopsied cases, 2 perforations of the terminal ileumresulted from erosion of the intestine by the through-and-through wire suturesused to repair a wound disruption on the 9th postoperative day; the patient diedof peritonitis on the 22d day. Data are not available concerning the origin ofthe leakage in the two other autopsied cases.
Secondary surgery for closure of fistulas was not performedin forward hospitals. The patients were evacuated as promptly as possible togeneral hospitals, where facilities for prolonged special care were available.
UNIVISCERAL WOUNDS
The 353 univisceral injuries in this series (table 50)include, as already mentioned (p. 241), 2 nonbattle injuries. The other 351casualties were all wounded by high-explosive fragments or by bullets. The ileumwas injured more frequently (180 cases) than the jejunum (128 cases), the ratiobeing roughly 3:2. Both portions of the bowel were injured simultaneously 45times, which is only about a quarter as frequently as the ileum was woundedalone.
The case fatality rate for the 353 univisceral injuries(table 50) was 13.9 percent. The rate for wounds of the jejunum alone (13deaths) and of the ileum alone (23 deaths) was approximately the same, 10.2 and12.8 percent, respectively. The case fatality rate rose sharply to 28.9 percent(13 deaths) when both portions of the bowel were involved. The increase isreadily explained by the fact that many of the injuries represented extensivetrauma to a large segment of the midbowel and required massive resection.
The average timelag from. injury to operation in the recordedfatal cases of univisceral injury of the jejuno-ileum was twice that in therecorded nonfatal
3Attention iscalled elsewhere (p. 249) to the additional case of fistula formation observedin a German prisoner of war, who had an established fistula of the small boweland an intraperitoneal abscess when he was admitted to a field hospital 3 daysafter wounding.
252
cases, being 9.5 hours in 293 nonfatal cases and 19.1 hoursin 46 fatal cases. If 2 cases in which the timelag was unusually prolonged (72hours and 90 hours, respectively) are omitted, the figure for the fatal cases isreduced to 16.3 hours. The timelag for all 339 (recorded) univisceral injurieswas 10.9 hours, which is essentially the same as the lag (10.6 hours) for the1,057 cases in the whole series of jejuno-ileal injuries in which these data areavailable. If all patients with this type of injury could have been operated onwithin 8 hours of wounding, their chances of recovery would probably have beengreatly enhanced.
A disproportionately large number of severe associated injuries apparentlycontributed to the case fatality rate of univisceral wounds of the jejunum andileum. Associated injuries were more than twice as frequent among the fatal asamong the nonfatal cases. They were recorded in 72 (23.7 percent) of the 304patients with univisceral injuries who survived and in 27 (55.1 percent) of the49 patients who died. They included serious compound fractures of the longbones, traumatic amputations of extremities, penetrating wounds of the thoraxother than thoracoabdominal wounds, severe cranial injuries, severemaxillofacial injuries, and soft-tissue wounds which either were extensive orwere associated with severe hemorrhage.
At least 3 of the 49 fatalities in univisceral wounds of the jejunum andileum can be justifiably attributed to these associated injuries. One man with awound of the heart died on the operating table from cardiac tamponade. Another,who had also sustained a blast injury of the lungs, died 5 hours afteroperation. The third died from hemolytic streptococcic bacteriemia and pyemiafollowing infection of a massive wound of the thigh. In all three cases, thediagnosis was established at autopsy.
In 15 other cases, death apparently was the result of the combined effectsof intestinal and associated wounds, both of which were severe, as thefollowing representative case history indicates:
Case 6-An infantry manwas brought into a field hospital in severe shock 20 hours after he hadsustained an extensive shell-fragment wound of the right buttock. Operation wasperformed 4 hours later, after he had received 3,000 cc. of blood. Two smallperforations of the ileum were sutured at laparotomy, which revealed anextensive early fibrinous peritonitis.
It proved impossible to control the extensive phagedenicinfection which developed in the wound of the buttock and which ultimatelyinvolved the entire gluteal muscle group and the lumbar and posterior thighmuscles also. The patient became oliguric and uremic and died on the fifth dayafter operation, his course having been continuously downhill. At autopsy, theperitoneal cavity was found clean and free of infection, and infection andnecrosis in the thigh and buttock were reported to be the chief causes of death.
If the 18 cases in which death was attributable wholly or in part toassociated wounds are excluded from this discussion and if only the 31fatalities in which the intestinal injury played the major role (table 54) areconsidered, a more accurate impression can be obtained of the causes ofdeath in cases in which injury of the jejunum and ileum was the primarily fatalfactor. Shock or peritonitis was listed as the chief cause in 15 patients whosurvived operation
253
for periods of time varying from 5 minutes to 48 hours. These casesfell into the well-defined group of casualties, seen in forward hospitals, whowere admitted to the field hospitals in the severe shock which seemed, in partat least, to be secondary to massive peritoneal contamination (p. 127). Thetimelag was usually long, associated hemorrhage was frequent, and death occurredpromptly.
Peritonitis was listed as the primary cause of death in 8 patients whosurvived operation for periods varying from 3 to 22 days. Peritoneal infectioncould not be controlled in these cases, but the element of persisting shock,which was present in the patients who died promptly after operation, was no partof the picture.
TABLE 54.-Causes of death in 31 primarily fatal injuries ofjejuno-ileum
Case | Timelag | Site of injury | Postoperative survival | Cause of death |
| Hours |
|
|
|
11 | 8.0 | Jejuno-ileum | 36 hours | Peritonitis; shock. |
12 | 19.0 | Ileum | 2 days | Generalized peritonitis, severe; pulmonary edema, severe. |
13 | 17.0 | Jejunum | 2 hours | Shock; peritonitis. |
4 | 12.0 | Jejuno-ileum | 24 hours | Shock. |
5 | 20.0 | ...do... | 2 days | Do. |
6 | 38.0 | Jejunum | 36 hours | Shock; severe mesenteric hemorrhage. |
17 | 27.0 | Ileum | 9 hours | Generalized fibrinopurulent peritonitis (present at operation). |
8 | 36.0 | ...do... | 5 hours | Shock; peritonitis. |
9 | 22.0 | Jejuno-ileum | 5 minutes | Do. |
10 | --- | ...do... | 5 hours | Shock. |
11 | 13.0 | Jejunum | 24 hours | Shock; peritonitis. |
12 | 48.0 | Ileum | 6.5 hours | Do. |
113 | 11.5 | Jejuno-ileum | 24 hours | Generalized peritonitis. |
14 | 37.0 | ...do... | 14 hours | Shock; peritonitis. |
115 | 10.0 | ...do... | 24 hours | Generalized peritonitis; shock. |
16 | 90.0 | Jejunum | 4 days | Generalized peritonitis. |
117 | 14.5 | Ileum | 5 days | Suppurative, generalized peritonitis, severe; mesenteric thrombosis lower third ileum. |
118 | 23.5 | Jejunum | 8 days | Acute fibrinopurulent peritonitis, severe; anuria; uremia. |
119 | 14.0 | Jejuno-ileum | 22 days | Generalized and localized peritonitis; 2 perforations ileum caused by wire sutures. |
20 | 6.0 | Ileum | 3 days | Peritonitis. |
21 | 15.0 | ...do... | 11 days | Do. |
122 | 7.0 | Jejuno-ileum | 12 days | Generalized peritonitis; bronchopneumonia. |
123 | 4.0 | Ileum | 8 days | Generalized and localized purulent peritonitis; leakage at anastomosis. |
124 | 10.0 | ...do... | 5 days | Intestinal obstruction; kinked anastomosis. |
125 | 7.0 | ...do... | 10 days | Massive pulmonary embolism. |
126 | 16.0 | Jejuno-ileum | 5 days | Oliguria; anuria; uremia. |
127 | 6.5 | Jejunum | 3 days | Diffuse purulent tracheobronchitis. |
128 | 6.5 | ...do... | 13 days | Hepatitis; atypical pneumonia; intestinal obstruction. |
129 | 5.5 | Ileum | 10 minutes | Aspiration of vomitus. |
130 | 4.0 | ...do... | 1 day | Cardiorespiratory death, unexplained clinically or at autopsy. |
31 | --- | Jejuno-ileum | 5 days | Not recorded. |
1Autopsy.
254
The incidence of peritonitis recorded at operation is not considered reliablebecause many of the operative notes were incomplete. It is of interest, however,that it was specifically mentioned as being present in 41 percent of thepatients who died, as contrasted with only 10 percent of those who survived.
Onedeath was attributed to intestinal obstruction. In one case, no cause of deathwas stated. In the six remaining cases, the fatalities were attributed to causesnot inherently related to wounds of the small intestine, such as pulmonaryembolism and anuria.
In the fatal univisceral wounds, therefore, about half of the patients diedin the immediate postoperative period of shock and overwhelming peritonealcontamination. About a quarter died later of peritonitis. The remainder, exceptfor the patient who died of intestinal obstruction, died of unpredictable andunrelated complications. The outstanding causes of death in patients who diedprimarily of univisceral injuries to the jejunum and ileum were shock, severeperitoneal contamination, prolonged timelag, and peritonitis. Interaction amongthese factors was often observed clinically. Apparently they were mutuallycomplementary, and they cannot be completely divorced from each other for thepurpose of statistical analysis.
