- Email: [email protected]
External counterpressure to control postoperative intra-abdominal hemorrhage
External Counterpressure
to Control Postoperative
Intra-Abdominal Hemorrhage James F. Burdick, MD, Boston, Massachusetts Andrew L. Warshaw, MD, Boston, Massachusetts William M. Abbott, MD, Boston, Massachusetts
Crile in 1903 described the use of external counterpressure to produce an intravascular volume redistribution from the lower part of the body in an attempt to ameliorate hypotension secondary to hemorrhage during neurosurgical procedures [I]. More recently, this principle has been applied for temporary preoperative control of bleeding from ruptured abdominal aortic aneurysms [2-51 and massive trauma [6]. External counterpressure has also been employed as primary treatment of hemorrhagic shock due to postpartum coagulopathies [7] and pelvic fractures [8]. This report describes the benefits and problems with the use of external counterpressure by the “G-suit”* in patients with intractable postoperative intra-abdominal hemorrhage. Procedure External counterpressure administered by the G-suit was used in twenty-eight postoperative patients with intra-abdominal bleeding that persisted despite maximal intra- and postoperative hemostatic measures. The indications were either diffuse uncontrollable bleeding, *Curity G-Suit, Kendall Co., Chicago, Illinois. From the General Surgical Services, Massachusetts General Hospital, and the Department of Surgery, Harvard Medical School, Boston, Massachusetts. Reprint requests should be addressed to Andrew L. Warshaw, MD. Massachusetts General Hospital, Boston, Massachusetts 02114. Presented at the Fifty-Fifth Annual Meeting of the New England Surgical Society, Waterville Valley. New Hampshire, September 26-26. 1974.
Volume 129, Aprfll976
apparent during closure of the abdomen, or an excessive blood requirement in the recovery room. The average blood volume transfused during the postoperative period prior to application of the G-suit was 5,000 cc. After the decision to use the G-suit was made, the patient was wrapped carefully in a soft flannel blanket with special attention paid to padding bony prominences and potential pressure points. The G-suit was applied from the costal margin to the lower part of the legs and inflated to 20 to 30 cm of saline using air (oxygen has been found to leak). All patients had indwelling endotracheal tubes for ventilatory assistance and urinary catheters. (Figure 1.) Most had nasogastric or gastrostomy tubes, cardiac monitors, and central venous or pulmonary artery catheters for pressure measurement. After the status of the patient had stabilized, correction of the clotting factor derangement was initiated. Otherwise, the care of the patient was continued using traditional methods. Once control of hemorrhage was achieved, the pressure in the G-suit was reduced in quantities of 5 cm of saline over a period of several hours. Results Although a treatment period of twenty-four hours without evidence of further bleeding was often selected arbitrarily, the twenty-eight patients actually remained in the G-suit for a mean of only twenty-one hours. As shown in Table I, hemorrhage ceased in nineteen of twenty-eight patients while in the G-suit. During the time the Gsuit was inflated, various derangements in clotting
369
Burdick, Warshaw,
and Abbott
had arterial bleeding. In four of these the site of arterial hemorrhage was identified during re-exploration; in the fifth patient, the arterial hemorrhage, demonstrated by angiography to be at an intestinal anastomosis, was successfully treated by selective intra-arterial infusion of Pitressine. Another patient had transection of the right branch of the portal vein within the hepatic hilum. The ability of the G-suit to control capillary hemorrhage but not distinct arterial bleeding is illustrated in the following case report.
Figure. 1. The G-sdt in use.
factors were corrected. Disseminated intravascular coagulopathy due to sepsis, cirrhosis, or transfusion reactions occurred in three patients and was successfully corrected with heparin and infusion of plasma clotting factors in two of the three. The remainder of the patients whose bleeding stopped during application of the G-suit had decreased platelet counts (eight patients) and/or elevated prothrombin times (fifteen patients), probably secondary to dilution by numerous transfusions of stored bank blood and albumin solution. These abnormalities were corrected by transfusion of platelets and fresh frozen plasma while the suit was inflated. When the suit was subsequently deflated, seventeen of the nineteen patients had no further bleeding. In two of the twenty-eight patients the effect of the G-suit was unclear. Bleeding directly from arteries or transected major veins was not controlled by the G-suit. As shown in Table I, five of the seven patients whose bleeding did not stop while they were in the G-suit
TABLE
I
Results of Treatment Counterpressure
with External
EtiologicFactors Causing Hemorrhage
Result Bleeding stopped Result unclear Bleeding stopped
not
Disseminated Intravascular Coagulopathy
Dilutional Coagulopathy
Arterial Bleeding Point
Other
Total
2
15
0
2
19
0
1
0
1
2
1
0
5
1
7
The patient (WS, MGH 183-85-97), a fifty-two year old man, was found to have severe pancreatitis after hemigastrectomy and vagotomy. On the twelfth postoperative day a massive pancreatic abscess was drained. Ten days later disseminated intravascular coagulation was manifested by thrombocytopenia, hypofibrinogenemia, prolongation of the prothrombin time, elevation of fibrin split products, and massive hemorrhage via the drain sites. The patient was placed in a G-suit and bleeding stopped immediately. He was treated with heparin, fresh frozen plasma, and platelets. The G-suit was gradually deflated and removed after thirty-seven hours. After six days without bleeding, massive intraabdominal hemorrhage recurred in the presence of normal clotting factors. The G-suit was reapplied but was ineffective. At laparotomy the gastroduodenal artery was found to have been eroded and was bleeding freely into the abscess cavity. Ligation of this artery completely stopped the bleeding. Rampant pneumonia caused death eight days after this last operation. Comnent: The first application of the G-suit was successful when bleeding was due to a correctable coagulopathy. The second episode of hemorrhage, however, was due to major arterial bleeding and required direct control by ligation. Various complications have occurred from the use of the G-suit. The G-suit restricts movement of the chest and diaphragm, necessitating mechanical ventilatory assistance, often with high inspiratory pressures. The resulting decrease in vital capacity was associated with significant atelectasis, pulmonary edema, or pneumonia in fourteen patients. One patient had bilateral hemothorax from rupture of a mediastinal hematoma that had dissected upwards from retroperitoneal arterial bleeding points. (Figure 2.) Skin necrosis or blistering occurred in at least five of twenty-eight patients secondary to unremitting local pressure. (Figure 3.) In one case an aortofemoral graft became occluded, and one patient had cardiac arrest just after the G-suit was applied. Of the fifteen patients whose renal function appeared normal (urinary output greater than 20
The American Journal ol Surgery
Intra-AbdominalHemorrhage
cc/hr; normal serum urea nitrogen and creatinine levels) at the time of application of the G-suit only four had a decreased urinary output during external counterpressure; the other eleven had unchanged or increased urinary output. Three of the fifteen subsequently had evidence of renal failure. Comments
Previously, external counterpressure has been employed mainly as an adjunct in short-term control of hemorrhage during transportation or in other preparations for surgery [2-61. In this series the G-suit was used for definitive treatment of ongoing intra-abdominal bleeding after failure of traditional surgical methods. Hemorrhage was arrested successfully in nineteen of twenty-eight patients, most of whom had underlying conditions predisposing them to hemorrhage, including cirrhosis with portal hypertension (fourteen patients), dilutional coagulopathies (sixteen patients), and disseminated intravascular coagulation (three patients). The temporary interruption of massive hemorrhage afforded by external counterpressure allowed first for the restoration of deficient hemostatic mechanisms and then for definitive clotting, so that bleeding usually did not recur after the suit was removed. It is apparent from this experience that it is inappropriate to use the G-suit in an attempt to stop intraperitoneal arterial bleeding despite the fact that intravascular volume redistribution out of the area under pressure may provide a transient benefit [1,9]. Although external counterpressure is usually successful in temporarily limiting leakage from a ruptured abdominal aortic aneurysm, it appears to do so by augmenting the tamponade normally afforded by periaortic and retroperitoneal tissues and perhaps also by decreasing the radius of the vessel, thereby decreasing the tension in the wall and closing the defect [3,10,11]. Venous hemorrhage, whether from a lateral defect or from a transection, may be controlled in a similar way, but completely transected arteries will continue to bleed [12,13]. Continuous external counterpressure for twenty-four hours appears effective in a substantial majority of appropriately selected patients and is associated with relatively few complications when its use is limited to this period of time. Atelectasis, pneumonia, and pulmonary edema were common in this series but can be minimized by positive-pressure assisted ventilation. Although deterioration of liver function as defined by labo-
Volume 129, April 1975
Figure 2. Bilateral massive hemothorax resutting from external counterpressure applied to control retroperitoneal arterial hemorrhage. Cephalad dissection of the retroperitoneal hematoma produced a mediastinal hematoma that ruptured into both pleural spaces.
ratory tests was observed in several patients, this was more likely due to hypotension, hemolysis, and hematoma resorption [14] than to deleterious effects of the G-suit. No evidence of acidosis [15] or ischemia of the pancreas or bowel was noted. The observed skin lesions, which are really areas of pressure necrosis or accelerated decubiti, can be kept to a minimum by proper skin care and padding prior to application of the suit. Periods of external counterpressure longer than twenty-four hours greatly increased the incidence of skin necrosis. Experiments in normotensive dogs have indicated that impairment of renal function occurs during external counterpressure [16]. However, many pa-
Figure 3. Bullae after twenty-four fnwrs in the G-suit. Areas of frank skin necrosis may atso occur.
371
Burdick, Warshaw,
and Abbott
tients in this series underwent long periods of external counterpressure with maintenance or improvement in renal function, in agreement with previous findings in hypotensivel dogs [17] and in man [8]. Summary The effectiveness of the G-suit in controlling massive postoperative intra-abdominal hemorrhage was studied in twenty-eight patients whose diffuse bleeding could not be controlled during operation. Most patients had developed deficiencies of platelets and clotting factors due to dilution, and in addition three had documented disseminated intravascular coagulation. After application of the G-suit, hemorrhage stopped in nineteen patients, allowing the replacement of platelets and clotting factors in patients with deficiencies and the administration of heparin to three patients with disseminated intravascular coagulation. After an average of twenty-one hours the G-suit was removed without rebleeding in seventeen patients. The major cause of G-suit failure was an arterial source of intra-abdominal bleeding. Applicatim of the G-suit had no adverse effect on renal function in at least half the patients; urinary output declined in one third. Most patients experienced respiratory impairment and some had ischemic skin lesions. The G-suit is frequently effective in halting postoperative intra-abdominal hemorrhage, allowing correction of acquired coagulopathies with acceptably few complications. Its use does not replace the need for proper surgical hemostasis. References 1. Crile GW: Blood Pressure in Surgery: An Experimental and Clinical Research. Philadelphia, Lippincott. 1903. p 288. 2. Espinosa MH, Updegrove JH: Clinical experience with the G suit. Arch Surg 101: 36, 1970. 3. Gardner WJ, Storer J: The use of the G-suit in control of intra-abdominal bleeding. Surg Gyneco/ Obsfet 123: 792, 1966. 4. Shane RA, Campbell GS: Protective influence of external counter pressure in acute hemorrhagic hypotension in dogs. Am J Surg 110: 355, 1965. 5. Darling RC: Ruptured arteriosclerotic abdominal aortic aneurysms. Am J Surg 119: 397, 1970. 6. Cutler BS, Daggett WM: Application of the “Gsuit” to the control of hemorrhage in massive trauma. Ann Surg 173: 511,197l. 7. Gardner WJ, Taylor HP, Dohn DF: Acute blood loss requiring 56 transfusions. JAMA 167: 985, 1958. 6. McLaughlin AP Ill, McCullough DL, Kerr WS Jr, Darling RC: The use of external counter pressure (G-suit) in the management of traumatic retroperitoneal hemorrhage. J Ural 107: 940.1972. 9. Wangensteen SL, Ludewig RM, Eddy DM: The effect of external counter pressure on the intact circulation. Surg Gy-
372
necol Obstet 127: 253, 1966. 10. Gardner WJ: Circumferential pneumatic compression. JAMA 196: 491, 1966. 11. Gardner WJ: Hemostasis by pneumatic compression. Am Surg 35: 635, 1969. 12. Eddy DM, Wangensteen SL, Ludewig RM: The kinetics of fluid loss from leaks in arteries tested by an experimental ex vivo preparation and external counter pressure. Surgery64: 451, 1968. 13. Wangensteen SL, Ludewig RM, Cox JM, Lynk JN: The effect of external counter pressure. Surgery 64: 922, 1968. 14. Kantrowitz PA, Jones WA, Greenberger NJ, lsselbacher KJ: Severe postoperative hyperbilirubinemia simulating obstructive jaundice. N Engl J Med 276: 59 1, 1967. 15. Wangensteen SL, DeHoll JD, Ludewig RM, Madden JJ: The detrimental effect of the Gsuit in hemorrhagic shock. Ann Surg 170: 167, 1969. 16. Shenasky JH, Gillenwater JY: The renal hemodynamic and functional effects of external counter pressure. Surg Gynecol Obstet 134: 253, 1972. 17. Roth JA, Rutherford RB: Regional blood flow effects of Gsuit application during hemorrhagic shock. Surg Gyneco/ Obstet 133: 637, 1971.
Discussion Robert Davis (Boston, Mass): I am impressed with the application of current technology to a problem which, until this was suggested, was virtually unsolvable. These results indicate that Doctor Burdick and his colleagues have succeeded in demonstrating a solution to this otherwise unsolvable problem. We have used the G-suit to transport patients with ruptured abdominal aneurysms. Our patients can be divided into two groups. The first group had a G-suit applied. These twelve patients were all transported for more than one hour from outlying hospitals; the mean time was about an hour and fifteen minutes. Another group of ten patients did not have G-suits and the average time from diagnosis until they reached the operating room was less than one hour. If there is any bias here, there was less preoperative time involved in the patients who did not have G-suits. The records of these patients show no bias in terms of survival based on age, diseases, or similar problems. In each group there was one patient with a questionable diagnosis that resulted in delay. There was also one other patient in each group who had such a severe illness, renal disease in one and congestive heart failure in one, that there was a delay in performing surgery. Eight of twelve patients (67 per cent) with G-suits lived; two of ten patients (20 per cent) without G-suits lived. We are impressed with the results in patients with G-suits. If ‘a lack of bias in selecting these patients is accepted, these results were significantly different. R. Clement Darling (Boston, Mass): With the help of Doctor Paul Russell, I first began using the G-suit on our Vascular Service at the Massachusetts General Hospital in October 1967 in the emergency management of patients with ruptured abdominal aortic aneurysms. A preliminary report was given to this Society at its annual meeting in 1969 (Am J Surg 119: 397, 1970). At that
The American Journal of Surgery
Intra-Abdominal
time about one third of all deaths from aortic rupture occurred during transport of the patient to the hospital, in the emergency ward, on the way to the operating room, or during induction of anesthesia, prior to mechanical control of the aorta. These deaths can be minimized or practically eliminated by the use of the G-suit. as has been shown in a personal series of eighty-eight patients with ruptured aneurysms. The survival rate has risen markedly from approximately 40 per cent to about 65 per cent with the use of G-suits. A second but well established use of the G-suit in our institution is in patients with a massively fractured pelvis, in order to delay bleeding prior to angiographic plugging of the offending vessel. Warshaw and his group have now suggested a third use of the G-suit as a means of stopping hemorrhage in postoperative patients. On our service any patient who is returned to the recovery room after aortic reconstruction who is clinically normovolemic and who then requires more than 2,500 cc of whole blood is generally returned for reoperation. This is done not to control active hemorrhage, which is rarely found, but simply to evacuate the massive hematoma in the retroperitoneal area and improve the subsequent period of convalescence. I would like to ask Doctor Burdick in how many of his patients this was necessary.
Thomas S. Risley (Beverly, Mass): I have a suggestion for pumping up the G-suit if no source of compressed air is available. We found that the bicycle pump or the little blood pressure pump that usually comes with it is inadequate. Many people in this room are yachtsmen and know about the Avon inflatable dinghy. That dinghy has a very excellent foot pump that may be
Vokma 129, April la75
Hemorrhage
obtained at almost any marine supply house. We found that it can inflate the G-suit, quickly and safely. Richard Wilson (Boston, Mass): I would like to know the ultimate survival time of the patients reported on by Doctor Rurdick.
James P. Burdick (closing): The success of the Gsuit in controlling blood loss in the preoperative period, as described by Doctor Davis and Doctor Darling, encouraged us to use it postoperatively in these desperately ill patients with multisystem underlying disease and coagulopathies. None of our patients had to be re-explored to evacuate a clot or hematoma. Part of the explanation is that the G-suit tends to decrease the amount of blood that would otherwise have accumulated. Without the G-suit, perhaps they would have continued to bleed to the point of tamponade by their own body structures, resulting in the formation of a much bigger hematoma. Doctor Risley’s suggestion about the method of inflating the suit is ingenious. We were fortunate to have a compressed air source available and used this, in preference to oxygen, to decrease leakage from the suits, but we were able continually to pump the G-suit to the appropriate pressure. Relative to Doctor Wilson’s question about ultimate survival, almost all of these patients eventually died. These patients were critically ill with multisystem failure. We were particularly concerned with the single issue of whether the bleeding could be stopped. The average survival time of the patients was fourteen days, although many survived much longer. The important thing is that no patient died of recurrent hemorrhage.
373
to Control Postoperative
Intra-Abdominal Hemorrhage James F. Burdick, MD, Boston, Massachusetts Andrew L. Warshaw, MD, Boston, Massachusetts William M. Abbott, MD, Boston, Massachusetts
Crile in 1903 described the use of external counterpressure to produce an intravascular volume redistribution from the lower part of the body in an attempt to ameliorate hypotension secondary to hemorrhage during neurosurgical procedures [I]. More recently, this principle has been applied for temporary preoperative control of bleeding from ruptured abdominal aortic aneurysms [2-51 and massive trauma [6]. External counterpressure has also been employed as primary treatment of hemorrhagic shock due to postpartum coagulopathies [7] and pelvic fractures [8]. This report describes the benefits and problems with the use of external counterpressure by the “G-suit”* in patients with intractable postoperative intra-abdominal hemorrhage. Procedure External counterpressure administered by the G-suit was used in twenty-eight postoperative patients with intra-abdominal bleeding that persisted despite maximal intra- and postoperative hemostatic measures. The indications were either diffuse uncontrollable bleeding, *Curity G-Suit, Kendall Co., Chicago, Illinois. From the General Surgical Services, Massachusetts General Hospital, and the Department of Surgery, Harvard Medical School, Boston, Massachusetts. Reprint requests should be addressed to Andrew L. Warshaw, MD. Massachusetts General Hospital, Boston, Massachusetts 02114. Presented at the Fifty-Fifth Annual Meeting of the New England Surgical Society, Waterville Valley. New Hampshire, September 26-26. 1974.
Volume 129, Aprfll976
apparent during closure of the abdomen, or an excessive blood requirement in the recovery room. The average blood volume transfused during the postoperative period prior to application of the G-suit was 5,000 cc. After the decision to use the G-suit was made, the patient was wrapped carefully in a soft flannel blanket with special attention paid to padding bony prominences and potential pressure points. The G-suit was applied from the costal margin to the lower part of the legs and inflated to 20 to 30 cm of saline using air (oxygen has been found to leak). All patients had indwelling endotracheal tubes for ventilatory assistance and urinary catheters. (Figure 1.) Most had nasogastric or gastrostomy tubes, cardiac monitors, and central venous or pulmonary artery catheters for pressure measurement. After the status of the patient had stabilized, correction of the clotting factor derangement was initiated. Otherwise, the care of the patient was continued using traditional methods. Once control of hemorrhage was achieved, the pressure in the G-suit was reduced in quantities of 5 cm of saline over a period of several hours. Results Although a treatment period of twenty-four hours without evidence of further bleeding was often selected arbitrarily, the twenty-eight patients actually remained in the G-suit for a mean of only twenty-one hours. As shown in Table I, hemorrhage ceased in nineteen of twenty-eight patients while in the G-suit. During the time the Gsuit was inflated, various derangements in clotting
369
Burdick, Warshaw,
and Abbott
had arterial bleeding. In four of these the site of arterial hemorrhage was identified during re-exploration; in the fifth patient, the arterial hemorrhage, demonstrated by angiography to be at an intestinal anastomosis, was successfully treated by selective intra-arterial infusion of Pitressine. Another patient had transection of the right branch of the portal vein within the hepatic hilum. The ability of the G-suit to control capillary hemorrhage but not distinct arterial bleeding is illustrated in the following case report.
Figure. 1. The G-sdt in use.
factors were corrected. Disseminated intravascular coagulopathy due to sepsis, cirrhosis, or transfusion reactions occurred in three patients and was successfully corrected with heparin and infusion of plasma clotting factors in two of the three. The remainder of the patients whose bleeding stopped during application of the G-suit had decreased platelet counts (eight patients) and/or elevated prothrombin times (fifteen patients), probably secondary to dilution by numerous transfusions of stored bank blood and albumin solution. These abnormalities were corrected by transfusion of platelets and fresh frozen plasma while the suit was inflated. When the suit was subsequently deflated, seventeen of the nineteen patients had no further bleeding. In two of the twenty-eight patients the effect of the G-suit was unclear. Bleeding directly from arteries or transected major veins was not controlled by the G-suit. As shown in Table I, five of the seven patients whose bleeding did not stop while they were in the G-suit
TABLE
I
Results of Treatment Counterpressure
with External
EtiologicFactors Causing Hemorrhage
Result Bleeding stopped Result unclear Bleeding stopped
not
Disseminated Intravascular Coagulopathy
Dilutional Coagulopathy
Arterial Bleeding Point
Other
Total
2
15
0
2
19
0
1
0
1
2
1
0
5
1
7
The patient (WS, MGH 183-85-97), a fifty-two year old man, was found to have severe pancreatitis after hemigastrectomy and vagotomy. On the twelfth postoperative day a massive pancreatic abscess was drained. Ten days later disseminated intravascular coagulation was manifested by thrombocytopenia, hypofibrinogenemia, prolongation of the prothrombin time, elevation of fibrin split products, and massive hemorrhage via the drain sites. The patient was placed in a G-suit and bleeding stopped immediately. He was treated with heparin, fresh frozen plasma, and platelets. The G-suit was gradually deflated and removed after thirty-seven hours. After six days without bleeding, massive intraabdominal hemorrhage recurred in the presence of normal clotting factors. The G-suit was reapplied but was ineffective. At laparotomy the gastroduodenal artery was found to have been eroded and was bleeding freely into the abscess cavity. Ligation of this artery completely stopped the bleeding. Rampant pneumonia caused death eight days after this last operation. Comnent: The first application of the G-suit was successful when bleeding was due to a correctable coagulopathy. The second episode of hemorrhage, however, was due to major arterial bleeding and required direct control by ligation. Various complications have occurred from the use of the G-suit. The G-suit restricts movement of the chest and diaphragm, necessitating mechanical ventilatory assistance, often with high inspiratory pressures. The resulting decrease in vital capacity was associated with significant atelectasis, pulmonary edema, or pneumonia in fourteen patients. One patient had bilateral hemothorax from rupture of a mediastinal hematoma that had dissected upwards from retroperitoneal arterial bleeding points. (Figure 2.) Skin necrosis or blistering occurred in at least five of twenty-eight patients secondary to unremitting local pressure. (Figure 3.) In one case an aortofemoral graft became occluded, and one patient had cardiac arrest just after the G-suit was applied. Of the fifteen patients whose renal function appeared normal (urinary output greater than 20
The American Journal ol Surgery
Intra-AbdominalHemorrhage
cc/hr; normal serum urea nitrogen and creatinine levels) at the time of application of the G-suit only four had a decreased urinary output during external counterpressure; the other eleven had unchanged or increased urinary output. Three of the fifteen subsequently had evidence of renal failure. Comments
Previously, external counterpressure has been employed mainly as an adjunct in short-term control of hemorrhage during transportation or in other preparations for surgery [2-61. In this series the G-suit was used for definitive treatment of ongoing intra-abdominal bleeding after failure of traditional surgical methods. Hemorrhage was arrested successfully in nineteen of twenty-eight patients, most of whom had underlying conditions predisposing them to hemorrhage, including cirrhosis with portal hypertension (fourteen patients), dilutional coagulopathies (sixteen patients), and disseminated intravascular coagulation (three patients). The temporary interruption of massive hemorrhage afforded by external counterpressure allowed first for the restoration of deficient hemostatic mechanisms and then for definitive clotting, so that bleeding usually did not recur after the suit was removed. It is apparent from this experience that it is inappropriate to use the G-suit in an attempt to stop intraperitoneal arterial bleeding despite the fact that intravascular volume redistribution out of the area under pressure may provide a transient benefit [1,9]. Although external counterpressure is usually successful in temporarily limiting leakage from a ruptured abdominal aortic aneurysm, it appears to do so by augmenting the tamponade normally afforded by periaortic and retroperitoneal tissues and perhaps also by decreasing the radius of the vessel, thereby decreasing the tension in the wall and closing the defect [3,10,11]. Venous hemorrhage, whether from a lateral defect or from a transection, may be controlled in a similar way, but completely transected arteries will continue to bleed [12,13]. Continuous external counterpressure for twenty-four hours appears effective in a substantial majority of appropriately selected patients and is associated with relatively few complications when its use is limited to this period of time. Atelectasis, pneumonia, and pulmonary edema were common in this series but can be minimized by positive-pressure assisted ventilation. Although deterioration of liver function as defined by labo-
Volume 129, April 1975
Figure 2. Bilateral massive hemothorax resutting from external counterpressure applied to control retroperitoneal arterial hemorrhage. Cephalad dissection of the retroperitoneal hematoma produced a mediastinal hematoma that ruptured into both pleural spaces.
ratory tests was observed in several patients, this was more likely due to hypotension, hemolysis, and hematoma resorption [14] than to deleterious effects of the G-suit. No evidence of acidosis [15] or ischemia of the pancreas or bowel was noted. The observed skin lesions, which are really areas of pressure necrosis or accelerated decubiti, can be kept to a minimum by proper skin care and padding prior to application of the suit. Periods of external counterpressure longer than twenty-four hours greatly increased the incidence of skin necrosis. Experiments in normotensive dogs have indicated that impairment of renal function occurs during external counterpressure [16]. However, many pa-
Figure 3. Bullae after twenty-four fnwrs in the G-suit. Areas of frank skin necrosis may atso occur.
371
Burdick, Warshaw,
and Abbott
tients in this series underwent long periods of external counterpressure with maintenance or improvement in renal function, in agreement with previous findings in hypotensivel dogs [17] and in man [8]. Summary The effectiveness of the G-suit in controlling massive postoperative intra-abdominal hemorrhage was studied in twenty-eight patients whose diffuse bleeding could not be controlled during operation. Most patients had developed deficiencies of platelets and clotting factors due to dilution, and in addition three had documented disseminated intravascular coagulation. After application of the G-suit, hemorrhage stopped in nineteen patients, allowing the replacement of platelets and clotting factors in patients with deficiencies and the administration of heparin to three patients with disseminated intravascular coagulation. After an average of twenty-one hours the G-suit was removed without rebleeding in seventeen patients. The major cause of G-suit failure was an arterial source of intra-abdominal bleeding. Applicatim of the G-suit had no adverse effect on renal function in at least half the patients; urinary output declined in one third. Most patients experienced respiratory impairment and some had ischemic skin lesions. The G-suit is frequently effective in halting postoperative intra-abdominal hemorrhage, allowing correction of acquired coagulopathies with acceptably few complications. Its use does not replace the need for proper surgical hemostasis. References 1. Crile GW: Blood Pressure in Surgery: An Experimental and Clinical Research. Philadelphia, Lippincott. 1903. p 288. 2. Espinosa MH, Updegrove JH: Clinical experience with the G suit. Arch Surg 101: 36, 1970. 3. Gardner WJ, Storer J: The use of the G-suit in control of intra-abdominal bleeding. Surg Gyneco/ Obsfet 123: 792, 1966. 4. Shane RA, Campbell GS: Protective influence of external counter pressure in acute hemorrhagic hypotension in dogs. Am J Surg 110: 355, 1965. 5. Darling RC: Ruptured arteriosclerotic abdominal aortic aneurysms. Am J Surg 119: 397, 1970. 6. Cutler BS, Daggett WM: Application of the “Gsuit” to the control of hemorrhage in massive trauma. Ann Surg 173: 511,197l. 7. Gardner WJ, Taylor HP, Dohn DF: Acute blood loss requiring 56 transfusions. JAMA 167: 985, 1958. 6. McLaughlin AP Ill, McCullough DL, Kerr WS Jr, Darling RC: The use of external counter pressure (G-suit) in the management of traumatic retroperitoneal hemorrhage. J Ural 107: 940.1972. 9. Wangensteen SL, Ludewig RM, Eddy DM: The effect of external counter pressure on the intact circulation. Surg Gy-
372
necol Obstet 127: 253, 1966. 10. Gardner WJ: Circumferential pneumatic compression. JAMA 196: 491, 1966. 11. Gardner WJ: Hemostasis by pneumatic compression. Am Surg 35: 635, 1969. 12. Eddy DM, Wangensteen SL, Ludewig RM: The kinetics of fluid loss from leaks in arteries tested by an experimental ex vivo preparation and external counter pressure. Surgery64: 451, 1968. 13. Wangensteen SL, Ludewig RM, Cox JM, Lynk JN: The effect of external counter pressure. Surgery 64: 922, 1968. 14. Kantrowitz PA, Jones WA, Greenberger NJ, lsselbacher KJ: Severe postoperative hyperbilirubinemia simulating obstructive jaundice. N Engl J Med 276: 59 1, 1967. 15. Wangensteen SL, DeHoll JD, Ludewig RM, Madden JJ: The detrimental effect of the Gsuit in hemorrhagic shock. Ann Surg 170: 167, 1969. 16. Shenasky JH, Gillenwater JY: The renal hemodynamic and functional effects of external counter pressure. Surg Gynecol Obstet 134: 253, 1972. 17. Roth JA, Rutherford RB: Regional blood flow effects of Gsuit application during hemorrhagic shock. Surg Gyneco/ Obstet 133: 637, 1971.
Discussion Robert Davis (Boston, Mass): I am impressed with the application of current technology to a problem which, until this was suggested, was virtually unsolvable. These results indicate that Doctor Burdick and his colleagues have succeeded in demonstrating a solution to this otherwise unsolvable problem. We have used the G-suit to transport patients with ruptured abdominal aneurysms. Our patients can be divided into two groups. The first group had a G-suit applied. These twelve patients were all transported for more than one hour from outlying hospitals; the mean time was about an hour and fifteen minutes. Another group of ten patients did not have G-suits and the average time from diagnosis until they reached the operating room was less than one hour. If there is any bias here, there was less preoperative time involved in the patients who did not have G-suits. The records of these patients show no bias in terms of survival based on age, diseases, or similar problems. In each group there was one patient with a questionable diagnosis that resulted in delay. There was also one other patient in each group who had such a severe illness, renal disease in one and congestive heart failure in one, that there was a delay in performing surgery. Eight of twelve patients (67 per cent) with G-suits lived; two of ten patients (20 per cent) without G-suits lived. We are impressed with the results in patients with G-suits. If ‘a lack of bias in selecting these patients is accepted, these results were significantly different. R. Clement Darling (Boston, Mass): With the help of Doctor Paul Russell, I first began using the G-suit on our Vascular Service at the Massachusetts General Hospital in October 1967 in the emergency management of patients with ruptured abdominal aortic aneurysms. A preliminary report was given to this Society at its annual meeting in 1969 (Am J Surg 119: 397, 1970). At that
The American Journal of Surgery
Intra-Abdominal
time about one third of all deaths from aortic rupture occurred during transport of the patient to the hospital, in the emergency ward, on the way to the operating room, or during induction of anesthesia, prior to mechanical control of the aorta. These deaths can be minimized or practically eliminated by the use of the G-suit. as has been shown in a personal series of eighty-eight patients with ruptured aneurysms. The survival rate has risen markedly from approximately 40 per cent to about 65 per cent with the use of G-suits. A second but well established use of the G-suit in our institution is in patients with a massively fractured pelvis, in order to delay bleeding prior to angiographic plugging of the offending vessel. Warshaw and his group have now suggested a third use of the G-suit as a means of stopping hemorrhage in postoperative patients. On our service any patient who is returned to the recovery room after aortic reconstruction who is clinically normovolemic and who then requires more than 2,500 cc of whole blood is generally returned for reoperation. This is done not to control active hemorrhage, which is rarely found, but simply to evacuate the massive hematoma in the retroperitoneal area and improve the subsequent period of convalescence. I would like to ask Doctor Burdick in how many of his patients this was necessary.
Thomas S. Risley (Beverly, Mass): I have a suggestion for pumping up the G-suit if no source of compressed air is available. We found that the bicycle pump or the little blood pressure pump that usually comes with it is inadequate. Many people in this room are yachtsmen and know about the Avon inflatable dinghy. That dinghy has a very excellent foot pump that may be
Vokma 129, April la75
Hemorrhage
obtained at almost any marine supply house. We found that it can inflate the G-suit, quickly and safely. Richard Wilson (Boston, Mass): I would like to know the ultimate survival time of the patients reported on by Doctor Rurdick.
James P. Burdick (closing): The success of the Gsuit in controlling blood loss in the preoperative period, as described by Doctor Davis and Doctor Darling, encouraged us to use it postoperatively in these desperately ill patients with multisystem underlying disease and coagulopathies. None of our patients had to be re-explored to evacuate a clot or hematoma. Part of the explanation is that the G-suit tends to decrease the amount of blood that would otherwise have accumulated. Without the G-suit, perhaps they would have continued to bleed to the point of tamponade by their own body structures, resulting in the formation of a much bigger hematoma. Doctor Risley’s suggestion about the method of inflating the suit is ingenious. We were fortunate to have a compressed air source available and used this, in preference to oxygen, to decrease leakage from the suits, but we were able continually to pump the G-suit to the appropriate pressure. Relative to Doctor Wilson’s question about ultimate survival, almost all of these patients eventually died. These patients were critically ill with multisystem failure. We were particularly concerned with the single issue of whether the bleeding could be stopped. The average survival time of the patients was fourteen days, although many survived much longer. The important thing is that no patient died of recurrent hemorrhage.
373