U.S. flag

An official website of the United States government

Skip to main content
Return to topReturn to top

Medical Science Publication No. 4, Volume 1

21 April 1954




It would be presumptuous of me to discuss with this group the classificationof hemorrhage and methods of control, but in order to present a numberof the observations made in Korea, some of the basic teachings relativeto hemorrhage will be mentioned. The conclusions reached here are drawnfrom observations made at the surgical hospital level.

Hemorrhage by definition is loss of blood from a blood vessel and, assuch, is a major contributing factor to shock. Hemorrhage may be classifiedas primary or secondary and may occur from a lacerated or severed vessel.It may be external or internal, arterial or venous, or both.


The majority of hemorrhagic problems reaching the surgical hospitalwere in those patients with wounds of major vessels of one or more of theextremities and wounds of lesser vessels of the abdominal and thoracicviscera. Often, extremity wounds were multiple and complicated by abdominalor thoracic hemorrhage. Most of those patients with wounds of the greatvessels of the thoracic and abdominal cavities bled to death before reachingthe hospital.

External. Although wounds of major vessels of the extremitieswere usually accessible to control by tourniquet, observations made froma group of 75 patients with 79 injuries to major vessels of the extremityshowed that 47 percent of this group were admitted with a tourniquet inplace and 40 percent were in shock. This percentage of patients in shockis indicative of the seriousness of hemorrhage from a major vessel evenwhen in a location where it can be controlled.

Since the severed artery tends to contract, retract and form a clot,it was interesting to speculate whether or not shock might be less prevalentin the patient with a completely severed artery than in the patient witha lacerated or partially severed artery, which is held apart and continuesto hemorrhage. Observation of 202 patients with extremity wounds involvingmajor vessels showed that 113, or 56

*Presented 21 April1954, to the Course on Recent Advances in Medicine and Surgery, Army MedicalService Graduate School, Walter Reed Army Medical Center, Washington, D.C.


percent, had a lacerated artery and 89, or 44 percent, had a severedartery. Using data from only a limited number (32) of this group arrivingin shock, an attempt was made to correlate the occurrence of shock witha specific type of arterial injury. Of those studied in shock, it was foundthat 53 percent had laceration of the artery and 47 percent had a severedartery, showing that shock occurred in the patient with a severed majorartery about as readily is in the patient with a lacerated major artery.Although several of these patients had multiple injuries, the gross hemorrhagewas from the major vessel involved.

While in general, primary, gross hemorrhage was the immediate problem,continued ooze also proved to be a problem in many instances. Artz, Sakoand Howard at the 46th Surgical Hospital (1) reported 20 deathsamong 138 severely injured patients requiring 5 or more pints of bloodeach. Of these 20 deaths, 7 died of uncontrolled hemorrhage, 4 from uncontrolledpostoperative oozing, and 3 from uncontrolled major vessel hemorrhage.

We have also observed patients continue to ooze and die postoperativelyfrom massive muscle wounds, liver wounds and pelvic wounds. Such continuedoozing was particularly evident following multiple transfusions. Dextranwas administered alone to some patients and to others in ratios of 1:1of blood and 1:2 of blood until some received a total of 2,500 to 3,000cc. of dextran. No increased bleeding was attributed to use of the dextran,although such has been reported since the war.

Internal. While most primary hemorrhage is external and is obviousor visible hemorrhage, internal hemorrhage is often the more serious problemespecially if in the abdomen or chest. Although we usually think of internalhemorrhage as abdominal or thoracic, it often occurs in the presence ofcomminuted fractures, especially fractures of the femur. Following suchhemorrhage, a patient may show signs of severe shock without evidence ofexternal bleeding. Another hidden form of internal hemorrhage is the pulsatinghematoma. Following injury to an artery, the blood may be retained betweenfascial and muscle planes or in the soft tissue forming a pulsating hematoma.The center of the hematoma communicates with the hole in the artery andcontains liquid blood while the peripheral portion of the hematoma tendsto laminate and clot, forming a false sac, the beginning of a false aneurysm.If both artery and vein are injured and the clot is contained in a similarmanner, then an arteriovenous fistula is likely to develop.

Blood flow distal to such pulsating hematomas will vary depending onthe severity of the injury to the artery, the number of collaterals destroyed,the size of the hematoma, the pressure exerted by the hema-


toma, and the peripheral resistance. When the extremity distal to thesite of vascular damage has ample blood supply and the hematoma is self-containedand uninfected, it may be expedient to allow the pulsating hematoma todevelop an arteriovenous fistula or false aneurysm which may be operatedon at a later date. This is particularly true of carotid and high subclavianartery injuries. Surgery for a pulsating hematoma in either of these locationsmay result in profuse hemorrhage before control can be obtained. Clampingof the damaged carotid artery for only the length of time required to repairit may lead to a fatality at this time.


While the pulsating hematoma of arterial origin is more likely to expandand cause pressure on adjacent structures, both the arterial and the arteriovenoustypes of pulsating hematomas may bleed secondarily, especially in the presenceof infection.

Secondary hemorrhage is usually arterial and may occur from a slippedligature, from the faulty suture line of a repaired vessel, from sloughof an injured vessel, or an overlooked vessel injury that has previouslythrombosed. Infection predisposes to secondary hemorrhage. Such hemorrhagemay occur from a severed or lacerated vessel as infection liquefies theclot. This is particularly true in the presence of pulsating hematomas.Russell (2) reported 59 cases of secondary hemorrhage developingat Tokyo Army Hospital in patients evacuated from Korea. He found thatmost secondary hemorrhage from war wounds occurred between the eighth andeighteenth days after wounding but varied from 5 to 44 days.


A study of 79 major vascular injuries in extremities showed that 47percent were admitted with a tourniquet in place which had been appliedfrom 40 minutes to 141/2hours previously, the average time being 4 hours. Tourniquets were foundapplied, covered by dressings and forgotten. Sometimes the presence ofa tourniquet or the time of application was not noted on the field medicaltag. Some tourniquets had cut deeply into the soft tissues. Others wereapplied at levels which required amputation of an extremity higher thanwas necessary for the injury alone.

It is believed that a tourniquet should be applied only tight enoughto control hemorrhage and left in place until it can be removed by a medicalofficer with blood or plasma expander available to resuscitate the patient.When packs or pressure dressings will suffice, the tourniquet should beremoved and the pressure dressing applied only tight enough to controlhemorrhage. Dressing can be carelessly applied


as tightly as a tourniquet. If applied properly, packs and pressuredressings will often control the major hemorrhage and allow some collateralvessels to function. Patients with pressure dressings must be observedcarefully to make sure that bleeding does not recur as the blood pressureincreases with resuscitation.

Now that the repair of acute vascular injuries is feasible, every effortshould be taken to repair the damaged artery. Some patients with majorartery injury will have hemorrhaged so severely that resuscitation becomesa major problem requiring prolonged use of a tourniquet. Continued useof the tourniquet during prolonged resuscitation may result in a nonviableextremity. In such cases with open wounds it may be practicable to clampthe injured artery as near the damaged site as possible preserving as muchof the artery as possible for repair. In this manner, the remaining collateralvessels are free to function while resuscitation is being accomplished.This procedure is not without danger. If the tourniquet is removed duringa critical stage of resuscitation, even after control of the damaged vessel,tourniquet shock may occur. If clamps are to be utilized to control thehemorrhage, they should be applied at the beginning of resuscitation, orwhen bleeding cannot be controlled by a tourniquet, or after the bloodpressure is stabilized and it is evident there will be a prolonged delaybefore surgery.

Care must be taken to avoid tourniquet shock when loosening or removingtourniquets that have been in place for a number of hours. We observed,as an example, a patient admitted in profound shock with both legs mangledand with tourniquets high on his thighs. He received 19 pints of bloodbefore a thin pulse became palpable. After receiving 5 more pints of bloodand dextran, he developed a near normal blood pressure. Since his tourniquetshad been in place for 6 hours it was decided to loosen them slowly andreplace them distally in hopes of salvaging the knee joints. While thiswas done without any evidence of external hemorrhage, the patient's bloodpressure dropped again and in spite of 4 more pints of blood given rapidly,the patient died, having received a total of 28 pints of blood.

Tourniquets should normally be applied as low as practicable to controlhemorrhage but for traumatic amputations the tourniquet should be appliedas low as possible on the stump, then it may be left in place indefinitely.

It may be well to mention here the availability of the artery for transfusionin traumatic amputations. Usually such patients have hemorrhaged severelyand are in profound shock. Since the artery is a tough, elastic structureit can usually be easily identified in the mangled remains of the extremity.The exposed artery can be


clamped, a large cannula or needle quickly inserted and the blood administeredrapidly by the route through which it was lost.

In spite of availability of blood and plasma expanders, hemorrhage provedto be a problem in many instances. Artz, Sako and Howard (1) reported89 severely wounded patients who required 15 or more pints of blood orplasma expander. Sixteen of the patients, or 18 percent, died of continuedhemorrhage, most of them with abdominal injury (table l). The patient withintra-abdominal hemorrhage often cannot be resuscitated preoperativelyand has to be taken to surgery for control of hemorrhage.

Table 1. Mortality of Casualties Requiring15 or More Pints of Blood From the 46th Surgical Hospital




Mortality (percent)

Number dying of continued hemorrhage





11 (46 percent)

Abdomen and extremities




3 (10 percent)





1 (3 percent)





1 (14 percent)





16 (18 percent)

In order to control intra-abdominal hemorrhage in these patients whoseshock could not be controlled by replacement of blood, a balloon catheterwas tested as an intra-aortic tamponade in two critically wounded patients.It was arbitrarily decided that the catheter would be used only in moribundpatients with intra-abdominal bleeding in which blood pressure could notbe obtained after administration of 10 pints of blood. The catheter wasutilized by inserting it through the femoral artery to the level of thediaphragm, then inflating the balloon with 20 cc.of sterile saline. Thecatheter was utilized in two moribund patients.

Case 1. The first patient was admitted 1 hour after injury withgrenade wounds of the abdomen, thighs, leg and foot and a compound comminutedfracture of the right tibia and fibula. He had received 1,000 cc. of plasmaprior to admission. On admission, his pressure was 40 systolic with a questionablediastolic level and pulse rate was 120. There was abdominal distentionsuggesting uncontrolled intra-abdominal bleeding. In the receiving wardhe was given 2,000 cc. of whole blood and 500 cc. of dextran, and his bloodpressure was read as 58 systolic with questionable diastolic pressure andpulse 120. He received 2,500 cc. more of blood without improvement. Hewas then taken to surgery in a moribund condition. There, no blood pressure


could be obtained. The catheter was inserted and passed to the levelof the diaphragm, where the balloon was inflated. The abdomen was thenquickly opened and 1,500 to 2,000 cc. of free blood was aspirated. A bleedingexternal iliac vein was controlled and a massive laceration of the rightlobe of the liver was packed and sutured.

After 15 minutes the catheter was slowly deflated, moved down to thebifurcation of the aorta, and reinflated to allow oxygenation of the kidneys,liver and spinal cord. This permitted multiple bleeding points in the boweland mesentery and around the celiac axis to present themselves. With thisbleeding, the blood pressure was again unobtainable and breathing becamelabored. The catheter was then replaced and inflated at the level of thediaphragm. The pressure was obtained at 78/50; then rose to 100/54, andrespiration again improved. While the catheter at the diaphragmatic levelcontrolled the bleeding and maintained a pressure, it also obscured thebleeding points so that they could no longer be found.

After 10 minutes at this level, the balloon was again slowly, partiallydeflated to expose the bleeding points, but the blood pressure was againlost. Re-inflation of the catheter balloon returned the blood pressureto 96/54, but again obscured the multiple bleeding points.

As the catheter was repeatedly deflated to demonstrate the bleedingpoints, the blood pressure continued to fall, and the patient's conditiongradually became worse until he died on the operating table. It was fearedthat prolonged use of the high aortic tamponade might result in liver,renal or spinal cord damage from anoxia.

During surgery, the patient received 13 additional pints of whole bloodand 1,000 cc. of dextran to total 22 pints of blood and 3 pints of dextran.

At autopsy, no damage to the celiac axis was demonstrated, but in additionto the damaged liver and iliac vein, there were multiple injuries to branchesof the splenic artery, the mesenteric arteries, and a severed spermaticartery.

Case II. The catheter was utilized too late in the second patientfor a blood pressure response. Although this patient received 24 pintsof blood, he died almost simultaneously with insertion of the catheter.

The fear of harming a patient who might have been resuscitated withoutuse of the catheter caused us to reserve it for use in absolutely moribundpatients. Possibly these patients would have had a better chance for survivalhad it been used earlier. Although both patients died, the catheter waseffective in temporarily restoring blood pressure in one patient and shouldbe further evaluated experimentally and clinically.



1. Primary gross hemorrhage was the usual immediate problem encounteredin the wounded in Korea but continued ooze also proved to be a problemat times.

2. Secondary hemorrhage played an important role, especially hemorrhagesresulting from pulsating hematomas and in the presence of infection. Thiswas particularly observed during the mass evacuation of the early daysof the war.

3. The control of hemorrage is discussed and the dangers of improperuse of a tourniquet are discussed.

4. The use of an intra-aortic balloon catheter tamponade as a methodof controlling intra-abdominal or intra-thoracic hemorrhage is presented.


1. Report of the Surgical Research Team from Korea-AMSGS,1953.

2. Russell, J. P.: Secondary Hemorrhage in War Wounds.In: Symposium on Military Medicine in the Far East Command, pp. 73-77,Sept. 1951.