- Email: [email protected]
Fluid Resuscitation in a Model of Uncontrolled Hemorrhage: Too Much Too Early, or Too Little Too Late?
JOURNAL OF SURGICAL RESEARCH ARTICLE NO.
63, 413–418 (1996)
0285
Fluid Resuscitation in a Model of Uncontrolled Hemorrhage: Too Much Too Early, or Too Little Too Late?1 ARI LEPPA¨NIEMI, M.D.,2 RALUAN SOLTERO, M.D., DAVID BURRIS, M.D., EMMANOUIL PIKOULIS, M.D., CHRISTINE WAASDORP, B.A., JENNIFER RATIGAN, B.S., HOWARD HUFNAGEL, B.S., AND DIANA MALCOLM, PH.D. Department of Surgery, Uniformed Services University of the Health Sciences, F. Edward He`bert School of Medicine, Bethesda, Maryland 20814-4799 Submitted for publication November 10, 1995
INTRODUCTION Early fluid resuscitation in hypotensive trauma patients is controversial due to the risk of increasing blood loss and mortality. We determined the effects of infusion rate and time of resuscitation on blood loss and mortality and compared the outcome to nonresuscitated animals in severe, uncontrolled hemorrhagic shock in a rat model. In anesthetized rats, piercing of the infrarenal aorta with a 25-G needle caused a fall of mean arterial pressure to õ20 mm Hg and blood loss of about 20 ml/kg in 90% of the animals. Animals were assigned to the following treatment groups (n Å 6): 60 ml/kg of lactated Ringer’s solution (LR) infused at a rate of 1.5 ml/min and given at 2.5 min (Group I), 5 min (Group II), or 10 min (Group III) postinjury, or LR infused at a rate of 3.0 ml/min and given at 5 min (Group IV) or 10 min (Group V) postinjury. Another group (n Å 9) was not resuscitated. The animals were followed for 3 hr. Total blood loss in Group I (30.5 { 2.6 ml/kg) was significantly (P õ 0.05) higher when compared to nonresuscitated animals (22.1 { 0.8 ml/ kg) or Group III (22.7 { 1.0 ml/kg), and also significantly higher in Group IV (35.8 { 4.1 ml/kg) when compared to nonresuscitated animals or Group V (23.0 { 1.2 ml/kg). The mortality rate was 7/9 in nonresuscitated animals and 5/6 in Group IV; both were significantly higher than in Groups II, III, and V (0 or 1/6) and markedly higher than in Group I (2/6). Conclusions: In this model of uncontrolled hemorrhage, initially uncorrected severe shock resulted in a high mortality rate. The risk of increased blood loss and mortality associated with early fluid resuscitation could be diminished by avoiding too fast of infusion rates early after the injury. q 1996 Academic Press, Inc.
1 Disclaimer: The opinions expressed herein are those of the authors and are not to be construed as reflecting the views of the Uniformed Services University of the Health Sciences, the Department of the Army, or the Department of Defense. 2 To whom reprint requests should be addressed at Second Department of Surgery, Helsinki University Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland.
The benefit of early fluid resuscitation has been questioned after both blunt and penetrating trauma [1–3]. The main reason for postponing volume replacement until after major bleeding sites have been surgically controlled is the concern that aggressive fluid resuscitation may disrupt the natural hemostatic mechanisms by preventing the formation of or disrupting the clots from injured blood vessels, leading to increased or recurrent hemorrhage and decreased survival. Several experimental models have been used to demonstrate the causal relationship between early fluid resuscitation and increased blood loss and mortality when compared to nonresuscitated animals [4–8]. In many of the studies, however, the rate of infusion of the resuscitative fluid has been very high and the time interval between injury and resuscitation very short, which may have contributed significantly to the risk of increased blood loss and mortality. Animal models of uncontrolled hemorrhage simulate the hemodynamic events leading to hemorrhagic shock following trauma by reproducing the primary pathophysiologic event, a blood vessel injury [9]. In a rat, the blood vessel injury has been induced by tail transection at different points ranging from 8 to 75% of the tail length [6, 10 – 14] or by an abdominal vascular injury such as transection of cecal branches of an ileocolic artery and vein [8], a combination of prebleeding of 20 ml/kg and incision of three major branches of the ileocolic artery [4], or a combined hepatic and retrohepatic caval vein injury [15]. In swine models, the vascular injury has been produced by a 4- or 5-mm aortotomy with or without bleeding the animals prior to injury [5, 7]. To avoid the need for bleeding the animal prior to injury due to the relatively small blood loss produced by the ileocolic vessel injuries, and to avoid the massive blood loss and short survival times (median, 10 min in nonresuscitated animals) after the caval injury, we developed a new and simple rat aortic injury model which produces a constant and severe, but not immedi0022-4804/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
413
AID
JSR 4838
/
6n11$$$141
06-10-96 16:05:37
srga
AP: Surg Res
414
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 2, JULY 1, 1996
FIG. 1. Mean arterial pressure (MAP) after sham operation (n Å 6) and after aortic injury in nonresuscitated animals with MAP staying above (NR ú 20, n Å 7) or decreasing below 20 mm Hg (NR õ 20, n Å 9). P õ 0.05 compared to *baseline value within group, # sham-operated animals at the same time point, and $NR ú 20 at the same time point.
ately lethal, hemorrhage in about 90% of the animals. Using this model of uncontrolled hemorrhagic shock, we determined the effects of the infusion rate and the timing of resuscitation on blood loss and mortality.
(LR 1.5/10); IV, LR 3.0 ml/min at 5 min postinjury (LR 3.0/5); V, LR 3.0 ml/min at 10 min postinjury (LR 3.0/10). Initial blood loss occurring immediately after creating the injury and postresuscitation (where appropriate) blood loss were measured by soaking all the blood in the abdominal cavity with preweighed gauze pads, which were then reweighed. A transformation formula of 1 g Å 0.9 ml blood was used. In all animals, the laparotomy incision was closed 30 min after the injury. MAP was recorded, and venous blood samples were taken for measurement of base excess (BE) prior to injury (baseline) at the start of the resuscitation (2.5–10 min postinjury in LR-treated animals; at 10 min in nonresuscitated animals) and at 30 min and 1, 2, and 3 hr postinjury. Except for the first two samples (baseline and preresuscitation), the blood drawn (0.3 ml/sample) was replaced with LR (0.9 ml). At 120 min postinjury an infusion of LR (4 ml/kg/hr) was started in all animals and maintained for the duration of the study. The animals were followed for 3 hr, and the survival time was recorded. This study was approved by the Uniformed Services University of the Health Sciences Animal Use Committee and conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to the principles stated in the Guide for the Care and Use of Laboratory Animals (NIH Publication 86-23, 1985 ed.). Data are presented as means { one standard error. Statistical analysis was performed using two-way analysis of variance followed by multiple comparisons with the Student–Newman–Keuls test to determine the degree of statistical significance within and between the groups for MAP and BE, t test or Mann Whitney rank sum test (if equal variance test failed) for blood loss, and Fischer’s exact test for mortality. A probability value of less than 0.05 was considered significant. NS, not significant.
RESULTS
Model Characterization METHODS Male Sprague–Dawley rats weighing 286–445 g were anesthetized with a combination of pentobarbital (35 mg/kg, ip) and ketamine (60 mg/kg, im). Polyethylene catheters (PE50, Clay Adams, Piscataway, NJ) were introduced into the carotid artery and jugular vein for hemodynamic monitoring and blood sampling, respectively. Through a midline incision, the infrarenal aorta was identified, and a suture was placed loosely behind the aorta to facilitate its later exposure. The animals were placed on a thermostatically controlled heating pad, and the carotid artery catheter was connected to a pressure transducer and computerized physiograph system (Buxco Electronics, Sharon, CT) for continuous blood pressure and heart rate monitoring. Venous blood gas samples were analyzed with a Stat Profile 2 Blood Gas and Electrolyte Analyzer (Nova Biomedical, Walthham, MA), and the resuscitative fluids were administered via the jugular vein with a Harvard Apparatus (South Natick, MA) infusion pump. Following instrumentation, the hemodynamically stable, but still anesthetized, animals were subjected to vascular injury leading to uncontrolled hemorrhagic shock by piercing through the infrarenal aorta with a 25-G needle, creating two standard-sized holes on each side of the aorta. The bowel was repositioned over the bleeding aorta, and the wound was left open, but covered with gauze. The resulting decrease in mean arterial pressure (MAP) was monitored. The shamoperated animals (SHAM, n Å 6) underwent identical procedures except that the aorta was mobilized but not pierced with the needle. If the MAP did not decrease below 20 mm Hg after the aortic injury (which occurs in about 10% of animals in this model), the animals were not resuscitated (NR ú 20, n Å 7) and were used for model characterization. If the MAP decreased below 20 mm Hg, the animals were randomly assigned to one of six groups: no resuscitation (NR õ 20, n Å 9) or resuscitation with 60 ml/kg lactated Ringer’s solution (LR) using five different fluid resuscitation regimens (n Å 6): I, LR 1.5 ml/min at 2.5 min postinjury (LR 1.5/2.5); II, LR 1.5 ml/min at 5 min postinjury (LR 1.5/5); III, LR 1.5 ml/min at 10 min postinjury
AID
JSR 4838
/
6n11$$$142
06-10-96 16:05:37
The baseline values of MAP and BE were not statistically different between the sham-operated and nonresuscitated animals (NR ú 20 and NR õ 20). All shamoperated animals survived 3 hr, and their MAP and BE did not change significantly from baseline values (Figs. 1 and 2). The blood loss caused by the laparotomy and aortic exposure alone was 0.8 { 0.1 ml/kg.
FIG. 2. Base excess (BE) after sham operation (n Å 6) and after aortic injury in nonresuscitated animals with MAP staying above (NR ú 20, n Å 7) or decreasing below 20 mm Hg (NR õ 20, n Å 9). P õ 0.05 compared to *baseline value within group, #sham-operated animals at the same time point, and $NR ú 20 at the same time point.
srga
AP: Surg Res
¨ NIEMI ET AL.: BLOOD LOSS AND FLUID RESUSCITATION LEPPA
415
TABLE 1 Blood Loss, Median Survival Time, and Mortality Following Resuscitation with Lactated Ringer’s Solution (LR) at Different Infusion Rates (1.5 and 3.0 ml/min) and Time Points (2.5, 5, and 10 min Postinjury) (n Å 6) and in Nonresuscitated Animals (NR õ 20, n Å 9) Blood loss (ml/kg) Treatment group
Initial
LR 1.5/2.5 LR 1.5/5 LR 1.5/10 LR 3.0/5 LR 3.0/10 NR õ 20
21.3 21.8 20.1 21.1 20.2 21.0
{ { { { { {
1.2 1.6 1.0 0.9 1.2 0.9
Mortality Total
30.5 27.9 22.7 35.8 23.0 22.1
{ { { { { {
2.6* 3.2 1.0**,*** 4.1* 1.2**,*** 0.8**,***
Median survival time (min)
At 1 hr
At 3 hr
180 180 180 50 180 21
1/6 1/6 1/6 4/6 0/6* 6/9
2/6 1/6*,** 1/6*,** 5/6 0/6*,** 7/9
Note. P õ 0.05 when compared to *NR õ 20, **LR 3.0/5, and ***LR 1.5/2.5.
The aortic injury caused a marked decrease in MAP, reaching, by definition, 26 { 2 mm Hg in the NR ú 20 group and 12 { 2 mm Hg in the NR õ 20 group (P õ 0.05) within 1–1.5 min after injury. At 10 min postinjury (Fig. 1), MAP had increased above 30 mm Hg in 6/7 animals in the NR ú 20 group (mean { SE, 43.7 { 5.4 mm Hg; median, 42.8 mm Hg; range, 28.2–44.1 mm Hg), but only in 2/9 animals in the NR õ 20 group (mean { SE, 21.6 { 3.0 mm Hg; median, 19.3 mm Hg; range, 11.8–41.0 mm Hg), which were the only animals surviving 3 hr in this group. All animals in the NR ú 20 group survived 3 hr. Compared to its own baseline value and to MAP at identical time points in shamtreated animals, MAP in the NR ú 20 group remained significantly lower during the first postinjury hour (Fig. 1), while BE in these animals did not change significantly from that of sham-treated animals or its own baseline values (Fig. 2). If the MAP decreased below 20 mm Hg (NR õ 20), 7/9 animals died, 10 (2 animals), 15, 21 (2), 22, and 80 min postinjury. The mortality rate was significantly higher when compared to animals in the NR ú 20 group (0/7) or sham-operated animals (0/6). Even in surviving animals of the NR õ 20 group, MAP and BE remained significantly lower than the baseline values during the first postinjury hour (Figs. 1 and 2). A base deficit persisted throughout the 3-hr observation period and was significantly greater than in SHAM and NR ú 20 groups during the first postinjury hour (Fig. 2). The initial blood loss caused by the aortic injury was 15.1 { 0.8 ml/kg in the NR ú 20 group and 21.0 { 0.9 ml/kg in the NR õ 20 group (P õ 0.05).
and 3-hr mortality were seen when the resuscitation starting point was changed from 10 to 5 min postinjury (Table 1). With the starting point for resuscitation at 10 min postinjury, doubling the rate of infusion from 1.5 to 3.0 ml/min did not increase blood loss or mortality. At 5 min postinjury, doubling the infusion rate tended to increase blood loss (P Å NS); however, there was a significant increase in 3-hr mortality (P õ 0.05) (Table 1). Compared to nonresuscitated animals (NR õ 20), the total blood loss was significantly higher after resuscitation 2.5 min postinjury (LR 1.5/2.5) or 5 min postinjury at the faster infusion rate (LR 3.0/5), although there was no significant difference in mortality between the three groups. In contrast, the 3-hr mortality rate compared to nonresuscitated animals was significantly lower when resuscitation was started 10 min postinjury regardless of the infusion rate (LR 1.5/10 and LR 3.0/10) or 5 min postinjury at the lower infusion rate (LR 1.5/5). There was no significant difference in total blood loss between the four groups (NR õ 20, LR 1.5/ 5, LR 1.5/10, LR 3.0/10), although the blood loss was somewhat higher in the LR 1.5/5 group (Table 1). The MAP responses in the five groups resuscitated with LR are shown in Fig. 3. There was no significant difference in baseline and preresuscitation MAP and BE (data not shown) values among the LR-treated animals or when compared to the NR õ 20 group. There was no statistically significant difference in MAP (Fig. 3) or BE responses between the groups resuscitated with LR.
Resuscitation
DISCUSSION
There was no significant difference in initial blood loss between the NR õ 20 and all LR-treated groups (Table 1). At an infusion rate of 1.5 ml/min, total blood loss was higher when the time interval for starting resuscitation was shortened from 10 to 2.5 min postinjury (P õ 0.05); the mortality rate, however, was low in both groups (1/6 vs 2/6). At a faster infusion rate (3.0 ml/min), significant increases in total blood loss
AID
JSR 4838
/
6n11$$$142
06-10-96 16:05:37
We present a simple and reproducible model of uncontrolled hemorrhagic shock in an attempt to mimic the clinical scenario of severe hemorrhagic shock caused by a major abdominal vascular injury following a stab wound or a low-velocity gunshot wound. The principal disadvantage of this model is that the injury and resuscitation are performed under anesthesia
srga
AP: Surg Res
416
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 2, JULY 1, 1996
FIG. 3. Mean arterial pressure (MAP) in animals subjected to aortic injury and resuscitated with lactated Ringer’s solution (LR) started 2.5–10 min postinjury and given at an infusion rate of 1.5 or 3.0 ml/kg. The 10-min time point represents preresuscitation values in all groups.
which might affect the physiological responses to hypotension and its correction. In addition, the open abdomen and lack of the abdominal compression effect during hemorrhage could increase blood loss, although in this model the overlying bowel seems to have a compressive effect on the bleeding aorta. During model development, different sized (18-, 23-, 25-, and 26-G) needles were tried, and the size producing a severe but not immediately lethal shock in a maximum number of animals was selected. Due to the inability to accurately control the depth and, thus, the size of the puncture, when only one hole was created with the tip of the needle, a through-and-through double injury was chosen for its better reproducibility. When evaluating the benefits of early fluid resuscitation in hypotensive trauma, the severity of shock could be an important marker. In a clinical study evaluating the advantages of paramedic-provided prehospital advanced life support in patients with major penetrating intra-abdominal vascular trauma, survival was significantly increased with prehospital advanced life support in patients with in-field systolic blood pressure of less than 60 mm Hg when compared to patients managed without paramedic services (e.g., scoop and run treatment) [16]. In contrast, in patients with in-field systolic blood pressure higher than 60 mm Hg, both paramedic-provided prehospital treatment and scoop and run management resulted in high (88.5 and 90.7%) survival rates. Experimental studies have confirmed the observation that scoop and run treatment results in low mortality rates after moderate shock and relatively small blood losses (7 – 17 ml/kg), whereas in more severe forms of hemorrhagic shock, early fluid resuscitation has been shown to improve survival in both shortand long-term follow up, especially after moderate ‘‘underresuscitation’’ to MAP of 40 – 60 mm Hg [6, 7, 10 – 14, 17].
AID
JSR 4838
/
6n11$$$143
06-10-96 16:05:37
In our model of hemorrhagic shock induced by an aortic injury, a MAP of 20 mm Hg was an important cutoff point in defining the severity of the shock. Our initial observation during model development was that the lowest MAP reached immediately after injury was a good prognostic sign for early survival. To develop a model where the shock would be severe enough to justify early fluid resuscitation, a cutoff point was chosen, which—in accordance with the selected needle size— produced a maximum number of animals possibly benefiting from fluid resuscitation. A cutoff point of 25 mm Hg would have included a large number of animals in the treatment groups which would have survived without fluid resuscitation. A cutoff point of 15 mm Hg, on the other hand, would have reduced the number of animals suitable for resuscitation without creating additional benefit. In about 90% of the animals, MAP fell initially below 20 mm Hg, resulted in a blood loss of 20 ml/kg corresponding to about 30% of the total blood volume, and lead to a 78% mortality rate within 3 hr in nonresuscitated animals. All but one of the deaths (86% of all deaths) occurred within 22 min postinjury. In contrast, if the initial MAP did not fall below 20 mm Hg, the blood loss averaged 15 ml/kg (about 20% of the total blood volume) and all animals survived through the observation period. Moreover, the BE, which is a reliable measure of the degree and duration of hypoperfusion, oxygen debt, and changes in oxygen delivery in hemorrhagic shock [18–20], did not differ from that of the sham-treated animals. During model development, the animals were initially followed for 24 hr. However, because the survival rates at 24 hr rarely differed from those at 3 hr and because the scope of the study was on the initial postresuscitation phase, the 3-hr follow up period was selected. Whether the severe shock in this model would
srga
AP: Surg Res
¨ NIEMI ET AL.: BLOOD LOSS AND FLUID RESUSCITATION LEPPA
cause delayed mortality within a few days or weeks requires further study. The variables which should be considered when the value of fluid resuscitation is evaluated in experimental models of uncontrolled hemorrhagic shock include the severity of shock, delay from injury to the start of resuscitation, volume and rate of infusion, type of fluid used, end points or goals of resuscitation, and length of follow up. Several studies have demonstrated that early fluid resuscitation in uncontrolled hemorrhagic shock increases blood loss and mortality. The risk of increased blood loss depends on the severity of the injury [10], delay from injury to the start of resuscitation [6], state of hydration and heat load of the animal [13], type and volume of fluid used [4, 6, 8, 12, 21], and whether resuscitation to normal or ‘‘subnormal’’ MAP values is attempted [14, 17]. The rate of infusion could also be significant as demonstrated in a study using the porcine aortotomy model, where 9 ml/kg/min of LR was given 6 min postinjury to swine over a period of 9 min corresponding to 5.6 liters to a 70-kg man over a 9min period, and leading to a mortality rate of 8/8 from exsanguination compared to 0/8 in nonresuscitated animals [5]. In this study, severely hypotensive animals were resuscitated with 60 ml/kg of LR which is equivalent to three times the volume of blood lost. If the resuscitation was started 10 min postinjury, there was no significant blood loss or mortality even with a higher infusion rate. The 10-min delay before starting the resuscitation could have been a factor in allowing a more resilient clot to form. However, if the resuscitation point was moved closer to the time of injury, increased blood loss and mortality were detected. Especially, when the fluid was given at a faster rate (3.0 ml/min) 5 min postinjury, there was a considerable increase in blood loss resulting in a mortality rate of 5/6 (83%) at 3 hr with 4/5 deaths occurring within the first postinjury hour. This study was designed to simulate an urban trauma setting where the response intervals are short. To accommodate for the hypercoagulating, small-sized rat, the delays in starting resuscitation were set between 2.5 and 10 min. As shown from the time of deaths in the NR õ 20 group, increasing the delay to 15 or 30 min postinjury would have resulted in preresuscitation mortality rates of about 33 and 67%, respectively. The significance of fluid resuscitation causing increased blood loss in a clinically relevant situation still seems to be controversial as demonstrated by a recent study which showed no significant difference in the mean estimated blood loss collected during surgery in patients with penetrating torso injuries subjected to immediate (prehospital) or delayed (after surgical control of bleeding) resuscitation [3]. The incidence of major vascular injuries was not stated, but it can be assumed that any major blood loss from vascular injuries would remain intra-abdominal and be collected during the operation. There was, however, a statistically significant difference in mortality between the groups
AID
JSR 4838
/
6n11$$$143
06-10-96 16:05:37
417
(30% vs. 38%), although the causes of death were not listed. It is possible that factors other than increased blood loss were responsible for the excess mortality. The authors concluded: ‘‘It is not the value of fluid resuscitation that is currently debated, but rather the volume, timing, and extent of that resuscitation for certain patients.’’ In this model of uncontrolled hemorrhage, moderate posttraumatic hypotension caused little disturbance in tissue perfusion as measured by BE and had a tendency for rapid spontaneous correction. In contrast, severe hypotension required early fluid resuscitation in order to avoid excess mortality. The potential risk of inducing recurrent hemorrhage from major blood vessels prior to surgical control could be reduced by avoiding too fast of infusion rates given at a very early stage after the injury. It is suggested that optimal timing of fluid resuscitation should be balanced between the urgency to restore perfusion and risk of aggravating blood loss. ACKNOWLEDGMENT This study was supported by a grant administered by the Henry M. Jackson Foundation for the Advancement of Military Medicine from the Baxter Healtcare Corp. (Round Lake, IL).
REFERENCES 1. Smith, J. P., Bodai, B. I., Hill, A. S., and Frey, C. F. Prehospital stabilization of critically injured patients: A failed concept. J. Trauma 25: 65, 1985. 2. Kaweski, S. M., Sise, M. J., and Virgilio, R. W. The effect of prehospital fluids on survival in trauma patients. J. Trauma 30: 1215, 1990. 3. Bickell, W. H., Wall, M. J., Jr., Pepe, P. E., Martin, R. R., Ginger, V. F., Allen, M. K., and Mattox, K. L. Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. N. Engl. J. Med. 331: 1105, 1994. 4. Gross, D., Landau, E. H., Assalia, A., and Krausz, M. M. Is hypertonic saline resuscitation safe in ‘uncontrolled’ hemorrhagic shock? J. Trauma 28: 751, 1988. 5. Bickell, W. H., Bruttig, S. P., Millnamow, G. A., O’Benar, J., and Wade, C. E. The detrimental effects of intravenous crystalloid after aortotomy in swine. Surgery 110: 529, 1991. 6. Krausz, M. M., Bar-Ziv, M., Rabinovici, R., and Gross, D. ‘‘Scoop and run’’ or stabilize hemorrhagic shock with normal saline or small-volume hypertonic saline? J. Trauma 33: 6, 1992. 7. Kowalenko, T., Stern, S., Dronen, S., and Wang, X. Improved outcome with hypotensive resuscitation of uncontrolled hemorrhagic shock in a swine model. J. Trauma 33: 349, 1992. 8. Craig, R. L., and Poole, G. V. Resuscitation in uncontrolled hemorrhage. Am. Surg. 60: 59, 1994. 9. Krausz, M. M. Controversies in shock research: Hypertonic resuscitation—Pros and cons. Shock 3: 69, 1995. 10. Krausz, M. M., Kablan, M., Rabinovici, R., Klin, B., Sherman, Y., and Gross, D. Effect of injured vessel size on bleeding following hypertonic saline infusion in ‘‘uncontrolled’’ hemorrhagic shock in anesthetized rats. Circ. Shock 35: 9, 1991. 11. Krausz, M. M., Landau, E. H., Klin, B., and Gross, D. Hypertonic saline treatment of uncontrolled hemorrhagic shock at different periods from bleeding. Arch. Surg. 127: 93, 1992. 12. Schultz, S. C., Powell, C. C., Burris, D. G., Nguyen, H., Jaffin, J., and Malcolm, D. S. The efficacy of diaspirin cross-linked hemoglobin solution resuscitation in a model of uncontrolled hemorrhage. J. Trauma 37: 408, 1994.
srga
AP: Surg Res
418
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 2, JULY 1, 1996
13. Krausz, M. M., Klemm, O., Amstislavsky, T., and Horovitz, M. The effect of heat load and dehydration on hypertonic saline solution treatment of uncontrolled hemorrhagic shock. J. Trauma 38: 747, 1995. 14. Capone, A. C., Safar, P., Stezoski, W., Tisherman, S., and Peitzman, A. Improved outcome with fluid restriction in treatment of uncontrolled hemorrhagic shock. J. Am. Coll. Surg. 180: 49, 1995. ˚ berg, T., Steen, S., al Othman, K., Norgren, L., and Bengmark, 15. A S. The effect of pneumatic antishock garments in the treatment of lethal combined hepatic and caval injuries in rats. J. Trauma 26: 727, 1986. 16. Aprahamian, C., Thompson, B. M., Towne, J. B., and Darin, J. C. The effect of paramedic system on mortality of major open intra-abdominal vascular trauma. J. Trauma 23: 687, 1983. 17. Stern, S. A., Dronen, S. C., Birrer, P., and Wang, X. Effect of blood pressure on hemorrhage volume and survival in near-
AID
JSR 4838
/
6n11$$$144
06-10-96 16:05:37
18.
19.
20.
21.
fatal hemorrhage model incorporating vascular injury. Ann. Emerg. Med. 22: 155, 1993. Davis, J. W., Shackford, S. R., and Holbrook, T. L. Base deficit as a sensitive indicator of compensated shock and tissue oxygen utilization. Surg. Gynecol. Obstet. 173: 473, 1991. Rutherford, E. J., Morris, J. A., Jr., Reed, G. W., and Hall, K. S. Base deficit stratifies mortality and determines therapy. J. Trauma 33: 417, 1992. Schultz, S. C., Hamilton, I. N., Jr., and Malcolm, D. S. Use of base deficit to compare resuscitation with lactated Ringer’s solution, haemaccel, whole blood, and diaspirin cross-linked hemoglobin following hemorrhage in rats. J. Trauma 35: 619, 1993. Bickell, W. H., Bruttig, S. P., Millnamow, G. A., O’Benar, J., and Wade, C. E. Use of hypertonic saline/Dextran versus lactated Ringer’s solution as a resuscitation fluid after uncontrolled aortic hemorrhage in anesthetized swine. Ann. Emerg. Med. 21: 1077, 1992.
srga
AP: Surg Res
63, 413–418 (1996)
0285
Fluid Resuscitation in a Model of Uncontrolled Hemorrhage: Too Much Too Early, or Too Little Too Late?1 ARI LEPPA¨NIEMI, M.D.,2 RALUAN SOLTERO, M.D., DAVID BURRIS, M.D., EMMANOUIL PIKOULIS, M.D., CHRISTINE WAASDORP, B.A., JENNIFER RATIGAN, B.S., HOWARD HUFNAGEL, B.S., AND DIANA MALCOLM, PH.D. Department of Surgery, Uniformed Services University of the Health Sciences, F. Edward He`bert School of Medicine, Bethesda, Maryland 20814-4799 Submitted for publication November 10, 1995
INTRODUCTION Early fluid resuscitation in hypotensive trauma patients is controversial due to the risk of increasing blood loss and mortality. We determined the effects of infusion rate and time of resuscitation on blood loss and mortality and compared the outcome to nonresuscitated animals in severe, uncontrolled hemorrhagic shock in a rat model. In anesthetized rats, piercing of the infrarenal aorta with a 25-G needle caused a fall of mean arterial pressure to õ20 mm Hg and blood loss of about 20 ml/kg in 90% of the animals. Animals were assigned to the following treatment groups (n Å 6): 60 ml/kg of lactated Ringer’s solution (LR) infused at a rate of 1.5 ml/min and given at 2.5 min (Group I), 5 min (Group II), or 10 min (Group III) postinjury, or LR infused at a rate of 3.0 ml/min and given at 5 min (Group IV) or 10 min (Group V) postinjury. Another group (n Å 9) was not resuscitated. The animals were followed for 3 hr. Total blood loss in Group I (30.5 { 2.6 ml/kg) was significantly (P õ 0.05) higher when compared to nonresuscitated animals (22.1 { 0.8 ml/ kg) or Group III (22.7 { 1.0 ml/kg), and also significantly higher in Group IV (35.8 { 4.1 ml/kg) when compared to nonresuscitated animals or Group V (23.0 { 1.2 ml/kg). The mortality rate was 7/9 in nonresuscitated animals and 5/6 in Group IV; both were significantly higher than in Groups II, III, and V (0 or 1/6) and markedly higher than in Group I (2/6). Conclusions: In this model of uncontrolled hemorrhage, initially uncorrected severe shock resulted in a high mortality rate. The risk of increased blood loss and mortality associated with early fluid resuscitation could be diminished by avoiding too fast of infusion rates early after the injury. q 1996 Academic Press, Inc.
1 Disclaimer: The opinions expressed herein are those of the authors and are not to be construed as reflecting the views of the Uniformed Services University of the Health Sciences, the Department of the Army, or the Department of Defense. 2 To whom reprint requests should be addressed at Second Department of Surgery, Helsinki University Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland.
The benefit of early fluid resuscitation has been questioned after both blunt and penetrating trauma [1–3]. The main reason for postponing volume replacement until after major bleeding sites have been surgically controlled is the concern that aggressive fluid resuscitation may disrupt the natural hemostatic mechanisms by preventing the formation of or disrupting the clots from injured blood vessels, leading to increased or recurrent hemorrhage and decreased survival. Several experimental models have been used to demonstrate the causal relationship between early fluid resuscitation and increased blood loss and mortality when compared to nonresuscitated animals [4–8]. In many of the studies, however, the rate of infusion of the resuscitative fluid has been very high and the time interval between injury and resuscitation very short, which may have contributed significantly to the risk of increased blood loss and mortality. Animal models of uncontrolled hemorrhage simulate the hemodynamic events leading to hemorrhagic shock following trauma by reproducing the primary pathophysiologic event, a blood vessel injury [9]. In a rat, the blood vessel injury has been induced by tail transection at different points ranging from 8 to 75% of the tail length [6, 10 – 14] or by an abdominal vascular injury such as transection of cecal branches of an ileocolic artery and vein [8], a combination of prebleeding of 20 ml/kg and incision of three major branches of the ileocolic artery [4], or a combined hepatic and retrohepatic caval vein injury [15]. In swine models, the vascular injury has been produced by a 4- or 5-mm aortotomy with or without bleeding the animals prior to injury [5, 7]. To avoid the need for bleeding the animal prior to injury due to the relatively small blood loss produced by the ileocolic vessel injuries, and to avoid the massive blood loss and short survival times (median, 10 min in nonresuscitated animals) after the caval injury, we developed a new and simple rat aortic injury model which produces a constant and severe, but not immedi0022-4804/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
413
AID
JSR 4838
/
6n11$$$141
06-10-96 16:05:37
srga
AP: Surg Res
414
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 2, JULY 1, 1996
FIG. 1. Mean arterial pressure (MAP) after sham operation (n Å 6) and after aortic injury in nonresuscitated animals with MAP staying above (NR ú 20, n Å 7) or decreasing below 20 mm Hg (NR õ 20, n Å 9). P õ 0.05 compared to *baseline value within group, # sham-operated animals at the same time point, and $NR ú 20 at the same time point.
ately lethal, hemorrhage in about 90% of the animals. Using this model of uncontrolled hemorrhagic shock, we determined the effects of the infusion rate and the timing of resuscitation on blood loss and mortality.
(LR 1.5/10); IV, LR 3.0 ml/min at 5 min postinjury (LR 3.0/5); V, LR 3.0 ml/min at 10 min postinjury (LR 3.0/10). Initial blood loss occurring immediately after creating the injury and postresuscitation (where appropriate) blood loss were measured by soaking all the blood in the abdominal cavity with preweighed gauze pads, which were then reweighed. A transformation formula of 1 g Å 0.9 ml blood was used. In all animals, the laparotomy incision was closed 30 min after the injury. MAP was recorded, and venous blood samples were taken for measurement of base excess (BE) prior to injury (baseline) at the start of the resuscitation (2.5–10 min postinjury in LR-treated animals; at 10 min in nonresuscitated animals) and at 30 min and 1, 2, and 3 hr postinjury. Except for the first two samples (baseline and preresuscitation), the blood drawn (0.3 ml/sample) was replaced with LR (0.9 ml). At 120 min postinjury an infusion of LR (4 ml/kg/hr) was started in all animals and maintained for the duration of the study. The animals were followed for 3 hr, and the survival time was recorded. This study was approved by the Uniformed Services University of the Health Sciences Animal Use Committee and conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to the principles stated in the Guide for the Care and Use of Laboratory Animals (NIH Publication 86-23, 1985 ed.). Data are presented as means { one standard error. Statistical analysis was performed using two-way analysis of variance followed by multiple comparisons with the Student–Newman–Keuls test to determine the degree of statistical significance within and between the groups for MAP and BE, t test or Mann Whitney rank sum test (if equal variance test failed) for blood loss, and Fischer’s exact test for mortality. A probability value of less than 0.05 was considered significant. NS, not significant.
RESULTS
Model Characterization METHODS Male Sprague–Dawley rats weighing 286–445 g were anesthetized with a combination of pentobarbital (35 mg/kg, ip) and ketamine (60 mg/kg, im). Polyethylene catheters (PE50, Clay Adams, Piscataway, NJ) were introduced into the carotid artery and jugular vein for hemodynamic monitoring and blood sampling, respectively. Through a midline incision, the infrarenal aorta was identified, and a suture was placed loosely behind the aorta to facilitate its later exposure. The animals were placed on a thermostatically controlled heating pad, and the carotid artery catheter was connected to a pressure transducer and computerized physiograph system (Buxco Electronics, Sharon, CT) for continuous blood pressure and heart rate monitoring. Venous blood gas samples were analyzed with a Stat Profile 2 Blood Gas and Electrolyte Analyzer (Nova Biomedical, Walthham, MA), and the resuscitative fluids were administered via the jugular vein with a Harvard Apparatus (South Natick, MA) infusion pump. Following instrumentation, the hemodynamically stable, but still anesthetized, animals were subjected to vascular injury leading to uncontrolled hemorrhagic shock by piercing through the infrarenal aorta with a 25-G needle, creating two standard-sized holes on each side of the aorta. The bowel was repositioned over the bleeding aorta, and the wound was left open, but covered with gauze. The resulting decrease in mean arterial pressure (MAP) was monitored. The shamoperated animals (SHAM, n Å 6) underwent identical procedures except that the aorta was mobilized but not pierced with the needle. If the MAP did not decrease below 20 mm Hg after the aortic injury (which occurs in about 10% of animals in this model), the animals were not resuscitated (NR ú 20, n Å 7) and were used for model characterization. If the MAP decreased below 20 mm Hg, the animals were randomly assigned to one of six groups: no resuscitation (NR õ 20, n Å 9) or resuscitation with 60 ml/kg lactated Ringer’s solution (LR) using five different fluid resuscitation regimens (n Å 6): I, LR 1.5 ml/min at 2.5 min postinjury (LR 1.5/2.5); II, LR 1.5 ml/min at 5 min postinjury (LR 1.5/5); III, LR 1.5 ml/min at 10 min postinjury
AID
JSR 4838
/
6n11$$$142
06-10-96 16:05:37
The baseline values of MAP and BE were not statistically different between the sham-operated and nonresuscitated animals (NR ú 20 and NR õ 20). All shamoperated animals survived 3 hr, and their MAP and BE did not change significantly from baseline values (Figs. 1 and 2). The blood loss caused by the laparotomy and aortic exposure alone was 0.8 { 0.1 ml/kg.
FIG. 2. Base excess (BE) after sham operation (n Å 6) and after aortic injury in nonresuscitated animals with MAP staying above (NR ú 20, n Å 7) or decreasing below 20 mm Hg (NR õ 20, n Å 9). P õ 0.05 compared to *baseline value within group, #sham-operated animals at the same time point, and $NR ú 20 at the same time point.
srga
AP: Surg Res
¨ NIEMI ET AL.: BLOOD LOSS AND FLUID RESUSCITATION LEPPA
415
TABLE 1 Blood Loss, Median Survival Time, and Mortality Following Resuscitation with Lactated Ringer’s Solution (LR) at Different Infusion Rates (1.5 and 3.0 ml/min) and Time Points (2.5, 5, and 10 min Postinjury) (n Å 6) and in Nonresuscitated Animals (NR õ 20, n Å 9) Blood loss (ml/kg) Treatment group
Initial
LR 1.5/2.5 LR 1.5/5 LR 1.5/10 LR 3.0/5 LR 3.0/10 NR õ 20
21.3 21.8 20.1 21.1 20.2 21.0
{ { { { { {
1.2 1.6 1.0 0.9 1.2 0.9
Mortality Total
30.5 27.9 22.7 35.8 23.0 22.1
{ { { { { {
2.6* 3.2 1.0**,*** 4.1* 1.2**,*** 0.8**,***
Median survival time (min)
At 1 hr
At 3 hr
180 180 180 50 180 21
1/6 1/6 1/6 4/6 0/6* 6/9
2/6 1/6*,** 1/6*,** 5/6 0/6*,** 7/9
Note. P õ 0.05 when compared to *NR õ 20, **LR 3.0/5, and ***LR 1.5/2.5.
The aortic injury caused a marked decrease in MAP, reaching, by definition, 26 { 2 mm Hg in the NR ú 20 group and 12 { 2 mm Hg in the NR õ 20 group (P õ 0.05) within 1–1.5 min after injury. At 10 min postinjury (Fig. 1), MAP had increased above 30 mm Hg in 6/7 animals in the NR ú 20 group (mean { SE, 43.7 { 5.4 mm Hg; median, 42.8 mm Hg; range, 28.2–44.1 mm Hg), but only in 2/9 animals in the NR õ 20 group (mean { SE, 21.6 { 3.0 mm Hg; median, 19.3 mm Hg; range, 11.8–41.0 mm Hg), which were the only animals surviving 3 hr in this group. All animals in the NR ú 20 group survived 3 hr. Compared to its own baseline value and to MAP at identical time points in shamtreated animals, MAP in the NR ú 20 group remained significantly lower during the first postinjury hour (Fig. 1), while BE in these animals did not change significantly from that of sham-treated animals or its own baseline values (Fig. 2). If the MAP decreased below 20 mm Hg (NR õ 20), 7/9 animals died, 10 (2 animals), 15, 21 (2), 22, and 80 min postinjury. The mortality rate was significantly higher when compared to animals in the NR ú 20 group (0/7) or sham-operated animals (0/6). Even in surviving animals of the NR õ 20 group, MAP and BE remained significantly lower than the baseline values during the first postinjury hour (Figs. 1 and 2). A base deficit persisted throughout the 3-hr observation period and was significantly greater than in SHAM and NR ú 20 groups during the first postinjury hour (Fig. 2). The initial blood loss caused by the aortic injury was 15.1 { 0.8 ml/kg in the NR ú 20 group and 21.0 { 0.9 ml/kg in the NR õ 20 group (P õ 0.05).
and 3-hr mortality were seen when the resuscitation starting point was changed from 10 to 5 min postinjury (Table 1). With the starting point for resuscitation at 10 min postinjury, doubling the rate of infusion from 1.5 to 3.0 ml/min did not increase blood loss or mortality. At 5 min postinjury, doubling the infusion rate tended to increase blood loss (P Å NS); however, there was a significant increase in 3-hr mortality (P õ 0.05) (Table 1). Compared to nonresuscitated animals (NR õ 20), the total blood loss was significantly higher after resuscitation 2.5 min postinjury (LR 1.5/2.5) or 5 min postinjury at the faster infusion rate (LR 3.0/5), although there was no significant difference in mortality between the three groups. In contrast, the 3-hr mortality rate compared to nonresuscitated animals was significantly lower when resuscitation was started 10 min postinjury regardless of the infusion rate (LR 1.5/10 and LR 3.0/10) or 5 min postinjury at the lower infusion rate (LR 1.5/5). There was no significant difference in total blood loss between the four groups (NR õ 20, LR 1.5/ 5, LR 1.5/10, LR 3.0/10), although the blood loss was somewhat higher in the LR 1.5/5 group (Table 1). The MAP responses in the five groups resuscitated with LR are shown in Fig. 3. There was no significant difference in baseline and preresuscitation MAP and BE (data not shown) values among the LR-treated animals or when compared to the NR õ 20 group. There was no statistically significant difference in MAP (Fig. 3) or BE responses between the groups resuscitated with LR.
Resuscitation
DISCUSSION
There was no significant difference in initial blood loss between the NR õ 20 and all LR-treated groups (Table 1). At an infusion rate of 1.5 ml/min, total blood loss was higher when the time interval for starting resuscitation was shortened from 10 to 2.5 min postinjury (P õ 0.05); the mortality rate, however, was low in both groups (1/6 vs 2/6). At a faster infusion rate (3.0 ml/min), significant increases in total blood loss
AID
JSR 4838
/
6n11$$$142
06-10-96 16:05:37
We present a simple and reproducible model of uncontrolled hemorrhagic shock in an attempt to mimic the clinical scenario of severe hemorrhagic shock caused by a major abdominal vascular injury following a stab wound or a low-velocity gunshot wound. The principal disadvantage of this model is that the injury and resuscitation are performed under anesthesia
srga
AP: Surg Res
416
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 2, JULY 1, 1996
FIG. 3. Mean arterial pressure (MAP) in animals subjected to aortic injury and resuscitated with lactated Ringer’s solution (LR) started 2.5–10 min postinjury and given at an infusion rate of 1.5 or 3.0 ml/kg. The 10-min time point represents preresuscitation values in all groups.
which might affect the physiological responses to hypotension and its correction. In addition, the open abdomen and lack of the abdominal compression effect during hemorrhage could increase blood loss, although in this model the overlying bowel seems to have a compressive effect on the bleeding aorta. During model development, different sized (18-, 23-, 25-, and 26-G) needles were tried, and the size producing a severe but not immediately lethal shock in a maximum number of animals was selected. Due to the inability to accurately control the depth and, thus, the size of the puncture, when only one hole was created with the tip of the needle, a through-and-through double injury was chosen for its better reproducibility. When evaluating the benefits of early fluid resuscitation in hypotensive trauma, the severity of shock could be an important marker. In a clinical study evaluating the advantages of paramedic-provided prehospital advanced life support in patients with major penetrating intra-abdominal vascular trauma, survival was significantly increased with prehospital advanced life support in patients with in-field systolic blood pressure of less than 60 mm Hg when compared to patients managed without paramedic services (e.g., scoop and run treatment) [16]. In contrast, in patients with in-field systolic blood pressure higher than 60 mm Hg, both paramedic-provided prehospital treatment and scoop and run management resulted in high (88.5 and 90.7%) survival rates. Experimental studies have confirmed the observation that scoop and run treatment results in low mortality rates after moderate shock and relatively small blood losses (7 – 17 ml/kg), whereas in more severe forms of hemorrhagic shock, early fluid resuscitation has been shown to improve survival in both shortand long-term follow up, especially after moderate ‘‘underresuscitation’’ to MAP of 40 – 60 mm Hg [6, 7, 10 – 14, 17].
AID
JSR 4838
/
6n11$$$143
06-10-96 16:05:37
In our model of hemorrhagic shock induced by an aortic injury, a MAP of 20 mm Hg was an important cutoff point in defining the severity of the shock. Our initial observation during model development was that the lowest MAP reached immediately after injury was a good prognostic sign for early survival. To develop a model where the shock would be severe enough to justify early fluid resuscitation, a cutoff point was chosen, which—in accordance with the selected needle size— produced a maximum number of animals possibly benefiting from fluid resuscitation. A cutoff point of 25 mm Hg would have included a large number of animals in the treatment groups which would have survived without fluid resuscitation. A cutoff point of 15 mm Hg, on the other hand, would have reduced the number of animals suitable for resuscitation without creating additional benefit. In about 90% of the animals, MAP fell initially below 20 mm Hg, resulted in a blood loss of 20 ml/kg corresponding to about 30% of the total blood volume, and lead to a 78% mortality rate within 3 hr in nonresuscitated animals. All but one of the deaths (86% of all deaths) occurred within 22 min postinjury. In contrast, if the initial MAP did not fall below 20 mm Hg, the blood loss averaged 15 ml/kg (about 20% of the total blood volume) and all animals survived through the observation period. Moreover, the BE, which is a reliable measure of the degree and duration of hypoperfusion, oxygen debt, and changes in oxygen delivery in hemorrhagic shock [18–20], did not differ from that of the sham-treated animals. During model development, the animals were initially followed for 24 hr. However, because the survival rates at 24 hr rarely differed from those at 3 hr and because the scope of the study was on the initial postresuscitation phase, the 3-hr follow up period was selected. Whether the severe shock in this model would
srga
AP: Surg Res
¨ NIEMI ET AL.: BLOOD LOSS AND FLUID RESUSCITATION LEPPA
cause delayed mortality within a few days or weeks requires further study. The variables which should be considered when the value of fluid resuscitation is evaluated in experimental models of uncontrolled hemorrhagic shock include the severity of shock, delay from injury to the start of resuscitation, volume and rate of infusion, type of fluid used, end points or goals of resuscitation, and length of follow up. Several studies have demonstrated that early fluid resuscitation in uncontrolled hemorrhagic shock increases blood loss and mortality. The risk of increased blood loss depends on the severity of the injury [10], delay from injury to the start of resuscitation [6], state of hydration and heat load of the animal [13], type and volume of fluid used [4, 6, 8, 12, 21], and whether resuscitation to normal or ‘‘subnormal’’ MAP values is attempted [14, 17]. The rate of infusion could also be significant as demonstrated in a study using the porcine aortotomy model, where 9 ml/kg/min of LR was given 6 min postinjury to swine over a period of 9 min corresponding to 5.6 liters to a 70-kg man over a 9min period, and leading to a mortality rate of 8/8 from exsanguination compared to 0/8 in nonresuscitated animals [5]. In this study, severely hypotensive animals were resuscitated with 60 ml/kg of LR which is equivalent to three times the volume of blood lost. If the resuscitation was started 10 min postinjury, there was no significant blood loss or mortality even with a higher infusion rate. The 10-min delay before starting the resuscitation could have been a factor in allowing a more resilient clot to form. However, if the resuscitation point was moved closer to the time of injury, increased blood loss and mortality were detected. Especially, when the fluid was given at a faster rate (3.0 ml/min) 5 min postinjury, there was a considerable increase in blood loss resulting in a mortality rate of 5/6 (83%) at 3 hr with 4/5 deaths occurring within the first postinjury hour. This study was designed to simulate an urban trauma setting where the response intervals are short. To accommodate for the hypercoagulating, small-sized rat, the delays in starting resuscitation were set between 2.5 and 10 min. As shown from the time of deaths in the NR õ 20 group, increasing the delay to 15 or 30 min postinjury would have resulted in preresuscitation mortality rates of about 33 and 67%, respectively. The significance of fluid resuscitation causing increased blood loss in a clinically relevant situation still seems to be controversial as demonstrated by a recent study which showed no significant difference in the mean estimated blood loss collected during surgery in patients with penetrating torso injuries subjected to immediate (prehospital) or delayed (after surgical control of bleeding) resuscitation [3]. The incidence of major vascular injuries was not stated, but it can be assumed that any major blood loss from vascular injuries would remain intra-abdominal and be collected during the operation. There was, however, a statistically significant difference in mortality between the groups
AID
JSR 4838
/
6n11$$$143
06-10-96 16:05:37
417
(30% vs. 38%), although the causes of death were not listed. It is possible that factors other than increased blood loss were responsible for the excess mortality. The authors concluded: ‘‘It is not the value of fluid resuscitation that is currently debated, but rather the volume, timing, and extent of that resuscitation for certain patients.’’ In this model of uncontrolled hemorrhage, moderate posttraumatic hypotension caused little disturbance in tissue perfusion as measured by BE and had a tendency for rapid spontaneous correction. In contrast, severe hypotension required early fluid resuscitation in order to avoid excess mortality. The potential risk of inducing recurrent hemorrhage from major blood vessels prior to surgical control could be reduced by avoiding too fast of infusion rates given at a very early stage after the injury. It is suggested that optimal timing of fluid resuscitation should be balanced between the urgency to restore perfusion and risk of aggravating blood loss. ACKNOWLEDGMENT This study was supported by a grant administered by the Henry M. Jackson Foundation for the Advancement of Military Medicine from the Baxter Healtcare Corp. (Round Lake, IL).
REFERENCES 1. Smith, J. P., Bodai, B. I., Hill, A. S., and Frey, C. F. Prehospital stabilization of critically injured patients: A failed concept. J. Trauma 25: 65, 1985. 2. Kaweski, S. M., Sise, M. J., and Virgilio, R. W. The effect of prehospital fluids on survival in trauma patients. J. Trauma 30: 1215, 1990. 3. Bickell, W. H., Wall, M. J., Jr., Pepe, P. E., Martin, R. R., Ginger, V. F., Allen, M. K., and Mattox, K. L. Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. N. Engl. J. Med. 331: 1105, 1994. 4. Gross, D., Landau, E. H., Assalia, A., and Krausz, M. M. Is hypertonic saline resuscitation safe in ‘uncontrolled’ hemorrhagic shock? J. Trauma 28: 751, 1988. 5. Bickell, W. H., Bruttig, S. P., Millnamow, G. A., O’Benar, J., and Wade, C. E. The detrimental effects of intravenous crystalloid after aortotomy in swine. Surgery 110: 529, 1991. 6. Krausz, M. M., Bar-Ziv, M., Rabinovici, R., and Gross, D. ‘‘Scoop and run’’ or stabilize hemorrhagic shock with normal saline or small-volume hypertonic saline? J. Trauma 33: 6, 1992. 7. Kowalenko, T., Stern, S., Dronen, S., and Wang, X. Improved outcome with hypotensive resuscitation of uncontrolled hemorrhagic shock in a swine model. J. Trauma 33: 349, 1992. 8. Craig, R. L., and Poole, G. V. Resuscitation in uncontrolled hemorrhage. Am. Surg. 60: 59, 1994. 9. Krausz, M. M. Controversies in shock research: Hypertonic resuscitation—Pros and cons. Shock 3: 69, 1995. 10. Krausz, M. M., Kablan, M., Rabinovici, R., Klin, B., Sherman, Y., and Gross, D. Effect of injured vessel size on bleeding following hypertonic saline infusion in ‘‘uncontrolled’’ hemorrhagic shock in anesthetized rats. Circ. Shock 35: 9, 1991. 11. Krausz, M. M., Landau, E. H., Klin, B., and Gross, D. Hypertonic saline treatment of uncontrolled hemorrhagic shock at different periods from bleeding. Arch. Surg. 127: 93, 1992. 12. Schultz, S. C., Powell, C. C., Burris, D. G., Nguyen, H., Jaffin, J., and Malcolm, D. S. The efficacy of diaspirin cross-linked hemoglobin solution resuscitation in a model of uncontrolled hemorrhage. J. Trauma 37: 408, 1994.
srga
AP: Surg Res
418
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 2, JULY 1, 1996
13. Krausz, M. M., Klemm, O., Amstislavsky, T., and Horovitz, M. The effect of heat load and dehydration on hypertonic saline solution treatment of uncontrolled hemorrhagic shock. J. Trauma 38: 747, 1995. 14. Capone, A. C., Safar, P., Stezoski, W., Tisherman, S., and Peitzman, A. Improved outcome with fluid restriction in treatment of uncontrolled hemorrhagic shock. J. Am. Coll. Surg. 180: 49, 1995. ˚ berg, T., Steen, S., al Othman, K., Norgren, L., and Bengmark, 15. A S. The effect of pneumatic antishock garments in the treatment of lethal combined hepatic and caval injuries in rats. J. Trauma 26: 727, 1986. 16. Aprahamian, C., Thompson, B. M., Towne, J. B., and Darin, J. C. The effect of paramedic system on mortality of major open intra-abdominal vascular trauma. J. Trauma 23: 687, 1983. 17. Stern, S. A., Dronen, S. C., Birrer, P., and Wang, X. Effect of blood pressure on hemorrhage volume and survival in near-
AID
JSR 4838
/
6n11$$$144
06-10-96 16:05:37
18.
19.
20.
21.
fatal hemorrhage model incorporating vascular injury. Ann. Emerg. Med. 22: 155, 1993. Davis, J. W., Shackford, S. R., and Holbrook, T. L. Base deficit as a sensitive indicator of compensated shock and tissue oxygen utilization. Surg. Gynecol. Obstet. 173: 473, 1991. Rutherford, E. J., Morris, J. A., Jr., Reed, G. W., and Hall, K. S. Base deficit stratifies mortality and determines therapy. J. Trauma 33: 417, 1992. Schultz, S. C., Hamilton, I. N., Jr., and Malcolm, D. S. Use of base deficit to compare resuscitation with lactated Ringer’s solution, haemaccel, whole blood, and diaspirin cross-linked hemoglobin following hemorrhage in rats. J. Trauma 35: 619, 1993. Bickell, W. H., Bruttig, S. P., Millnamow, G. A., O’Benar, J., and Wade, C. E. Use of hypertonic saline/Dextran versus lactated Ringer’s solution as a resuscitation fluid after uncontrolled aortic hemorrhage in anesthetized swine. Ann. Emerg. Med. 21: 1077, 1992.
srga
AP: Surg Res