- Email: [email protected]
The cardiopulmonary effects of severe blood loss in anesthetized horses
Veterinary Anaesthesia and Analgesia, 2003, 30, 80^86
The cardiopulmonary effects of severe blood loss in anesthetized horses Deborah V Wilson
BVSc, MS, Diplomate ACVA, Yves
Rondenay
DVM
& Phyllis U Shance
LVT
Department of Large Animal Clinical Sciences, Michigan State University, East Lansing, MI, USA
Correspondence: DV Wilson, Department of Large Animal Clinical Sciences, Michigan State University, East Lansing, MI 48864, USA. E-mail: [email protected]
Abstract Objective To characterize the acute cardiopulmonary e¡ects of severe hemorrhage in anesthetized horses. Study design Prospective experimental study. Animals Three geldings and six mares, aged 14.4 2.7 years, weighing 486 41 kg (range: 425^550 kg). Methods Horses were anesthetized using xylazine, guaifenesin, ketamine and halothane or iso£urane. Cardiovascular variables, hematocrit, total solids, capillary re¢ll time (CRT) and color of mucous membranes were measured as blood was collected from the carotid artery into sterile plastic bags. Arterial blood gas analysis was also performed. Results The average amount of blood collected from these horses was (mean SD) 53 4.8 mL kg1 bodyweight (range: 23^32 kg) over 39 4 minutes. Hematocrit decreased from 38 3 to 32 2% after induction of anesthesia and did not change signi¢cantly over the period of blood loss. Total solids decreased signi¢cantly after induction of anesthesia, and over the period of blood loss. Systolic, mean, diastolic and pulse pressures decreased as blood was lost. Heart rate did not change signi¢cantly. Capillary re¢ll time increased from 1.6 0.4 seconds to
Present address: Department of Clinical Sciences, College of Veterinary Medicine, University of Montreal, Montreal, Canada. 80
4.8 1.3 seconds as blood loss increased. Mucous membrane color deteriorated progressively. Arterial PO2 decreased signi¢cantly over the period of blood loss. Conclusions Hematocrit and heart rate do not change signi¢cantly during acute severe hemorrhage in the anesthetized horse. Arterial blood pressure, pulse pressure and PaO2 decrease as blood loss increases. Changes in mucous membrane color and CRT also occur as blood loss increases. Clinical relevance During severe hemorrhage in the inhalant-anesthetized horse, both heart rate and hematocrit remain unchanged. Blood pressure decreases and changes in arterial PO2 correlate most strongly with volume of blood lost. Keywords anesthetized horse, arterial blood pressure, baroreceptor sensitivity, cardiopulmonary, hemorrhage, hypoxia.
Introduction The accurate quanti¢cation of blood loss can be di⁄cult in the anesthetized horse. If the shed blood is not collected, the volume lost can be only poorly approximated. There are little or no data validating clinically measured variables as indicators of volume change during acute hemorrhage in the anesthetized horse. It is generally considered that a conscious 500 kg horse may lose a third of its circulating blood volume (12^16 L) before signi¢cant cardiovascular changes occur (Byars & Divers 1981; Durham 1996).
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
General anesthesia, inhalant anesthesia and the associated normal reduction of cardiac output are likely to reduce this volume signi¢cantly. A study in dogs showed that most commonly used hemodynamic and blood indices are inaccurate when quantifying hemorrhage (Waisman et al. 1993). In that study, the best indicator of volume change was variation in acid^base status re£ected by arterial base de¢cit, bicarbonate and venous pH (Waisman et al. 1993). In people, a decrease in pulse pressure and decreased area under the arterial pulse contour have been shown to accompany blood loss (Skillman et al. 1967). Tachycardia and reduced arterial pressure are reported in response to severe hemorrhage in one group of horses anesthetized with guaifenesin and barbiturates (Schmall et al. 1990). A case report of life-threatening hemorrhage during iso£urane anesthesia in a horse described arterial hypotension, but heart rate was not reported (Trim et al.1997). We know from clinical experience that heart rate in the inhalant-anesthetized horse is notoriously stable in the face of anesthetic-induced hypotension. This phenomenon has been described by Hellyer et al. (1989, 1991) who showed that both halothane and iso£urane anesthesia cause a signi¢cant decrease or abolition of barore£ex sensitivity in the horse. The dynamic function of this re£ex during onset of arterial hypotension accompanying severe hemorrhage has not been evaluated. We hypothesized that heart rate would not change during blood loss in anesthetized horses. Our objective with this study was to measure a number of cardiopulmonary and hematologic variables during acute blood loss in anesthetized horses, and to correlate these variables with volume lost.
Materials and methods Horses This study was approved by the All University Committee on Animal Use and Care at Michigan State University. Nine horses were included in this study: three Quarter horses, two Thoroughbred crosses, one Pinto, one Trakehner and two mixed breeds. There were three geldings and six mares in the study, aged 14.4 2.7 years, and weights ranged from 425 to 550 kg (486 41 kg). All horses were determined to be free from evidence of systemic disease by physical examination prior to the study. These horses were scheduled for terminal blood collection as plasma donors. # Association of Veterinary Anaesthetists, 2003, 30, 80^86
Anesthetic protocol The horses had a catheter placed in the right jugular vein and sutured in position following local anesthesia and surgical preparation of the area. The horses were sedated with intravenous xylazine until tractable (0.3 mg kg1 initial dose; Xylazine-100 injectable, The Butler Co., Columbus, OH, USA). Guaifenesin (Guaifenesin injection, Phoenix Scienti¢c Inc., St. Joseph, MO, USA) was administered intravenously until signs of muscle relaxation were observed. Then, a bolus dose of ketamine (2 mg kg1; KetaFlo, Abbott Laboratories, North Chicago, IL, USA) was administered intravenously; additional doses were given as required to facilitate intubation or to deepen anesthesia. The horses were placed in right lateral recumbency. Endotracheal intubation was performed using a cu¡ed tube, and the horses were connected to a semiclosed breathing circuit. The horses were mechanically ventilated using 6^ 10 breaths minute1 and a peak airway pressure <20 cm H2O. Halothane or iso£urane in oxygen (1^ 3% and 3^6 L minute1, respectively) was used to maintain anesthesia. Lactated Ringer’s solution was administered via the catheter in the right jugular vein at a rate of 10 mL kg1 hour1 for the duration of the procedure. A lethal dose of pentobarbital was administered IVat the end of the procedure. Monitoring A catheter was placed in the lateral metatarsal artery for monitoring blood pressure. The transducer was zeroed to the level of the left atrium (PB 240 Puritan Bennett Corp.,Wilmington, MA, USA). Blood was collected from the same catheter for gas analysis. Oral mucous membrane color was assessed using the color choices of: pink, pale pink, white/blanched, gray or blue. Capillary re¢ll time was assessed by standard means. A modi¢ed base^apex electrocardiogram was monitored, and heart rate (HR) was calculated from an R-to-R interval. The change in HR from the baseline value was plotted as a function of the percent change in mean arterial pressure (MAP) from baseline (Hellyer et al.1989,1991). Blood collection The collection procedure was performed by surgically placing a cannula into the carotid artery, as described previously (Eicker & Ainsworth 1984). Blood was collected into sterile 3- and 5-L bags 81
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
containing 4% sodium citrate. The bags were weighed immediately prior to blood collection and remained on the scale during ¢lling. The progression of blood collection was monitored by the weight increase of the bag. The collection line was brie£y clamped (5^10 seconds) each time the collection bags were changed. A catheter was placed in the saphenous vein for sample collection to determine hematocrit and total solids. Sample and data collection Baseline values of the heart and respiratory rates, hematocrit and total solids (microhematocrit centrifugation and refractometry) were determined before placement of the jugular venous catheter. After induction of anesthesia, the heart rate, respiratory rate, and systolic, diastolic and mean arterial pressures were recorded every 5 minutes until the start of the blood collection. During the period of blood collection, the cardiovascular values were recorded and samples for hematocrit and total solids were collected at the end of each kilogram collected. The time was recorded at each of the sample periods. Arterial blood was collected for blood gas analysis before the start of surgery, and at the end of each 5 kg of blood collected. A sample was also collected just before the horse was euthanatized at the end of the blood collection period. All data are presented as mean SEM, unless otherwise stated. Parametric data were analyzed using a one-way repeated measures ANOVA. When there were signi¢cant (p < 0.05) main e¡ects, the post hoc test used was the Tukey’s test. Linear regression was performed where applicable. Statistical analysis was conducted with Sigma StatTM, version 1.03 (Sigma StatTM for Windows, Jandel Scienti¢c, San Rafael, CA, USA).
blood collection commenced. Time from induction of anesthesia until administration of a lethal dose of pentobarbital was100 5.7 minutes. Blood collection The average blood volume collected from these horses was 52.5 1.5 g kg1 bodyweight (range: 46^ 58 g kg1).Total amount collected was 26 0.88 kg1 (range: 23^32 kg1). The average duration of the blood collection period was 39 4 minutes. The collection bag was the same distance below heart level in all horses and the £ow rate was not controlled. The initial rate of blood loss was 1.5 0.2 g kg1 minute1 and this did not vary signi¢cantly for most of the study period. Near the end, when more than 25 kg had been lost, the bleeding rate had decreased to 1 0.2 g kg1 minute1. The horses received a total of 12 2.4 mL kg1 lactated Ringer’s solution over the duration of the anesthetic period. Cardiovascular e¡ects A signi¢cant decrease in mean arterial, systolic, diastolic and pulse pressures occurred as blood was lost (Fig. 1). Systolic blood pressure began to
Results Anesthesia The horses received a total dose of 0.5 0.04 mg kg1 bodyweight of xylazine intravenously prior to induction. Guaifenesin and ketamine were administered intravenously at doses of 79 4.3 and 3.1 0.14 mg kg1 bodyweight, respectively, to induce anesthesia. In eight of the horses, anesthesia was maintained with halothane in oxygen, and iso£urane was used in one horse. These horses were under anesthesia for 61 4.6 minutes whilst the surgical approach was made to the carotid artery, and before 82
Figure 1 Systolic, mean and diastolic arterial blood pressures in nine anesthetized horses during acute development of severe hemorrhagic shock. $Values signi¢cantly di¡erent (p < 0.05) from control values. # Association of Veterinary Anaesthetists, 2003, 30, 80^86
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
Table 1 Hematocrit, plasma total solids, heart rate and mucous membrane color in nine horses during acute, severe hemorrhage
Variable
Control
5 kg
10 kg
15 kg
20 kg
Pre-terminal
Heart rate (beats minute1) Mucous membrane color Hematocrit (vol.%) Total solids (g dL1)
34 1 Pink 32 2 6.0 0.2
32 1 Pink 32 1 5.7 0.2
33 1 Pale pink 33 2 5.6 0.2
34 3 White 31 2 5.4 0.2
32 1 White 32 4 5.2 0.2
38 5 Blue 36 4 5.1 0.2
Values significantly different ( p < 0.01) from control values. Control ¼ under anesthesia and before surgery is started; 5, 10, 15, and 20 kg ¼ after 5, 10, 15, and 20 kg of blood, respectively, had been collected; pre-terminal ¼ sample collected just before the horse was euthanatized. All data are mean SEM except mucous membrane color which is the median.
decrease slowly in six horses as soon as blood loss commenced, eventually decreasing in all horses as blood loss exceeded 10 g kg1 (r ¼ 0.63). The rate of decrease was maximal between 10 and 31 g kg1 of blood loss. Pulse pressure decreased (r ¼ 0.6) in a similar fashion throughout the blood collection period (Fig. 1). Diastolic pressure decreased more slowly than systolic pressure; this decrease was signi¢cant by the time16 kg of blood had been lost. Baseline heart rate under anesthesia was 34 1 beats minute1 and heart rate did not vary signi¢cantly over the blood collection period (Table 1). An increase in heart rate occurred transiently in two horses (from 35 to110 and 28 to 200 beats minute1) at total blood losses of 19 and 25 kg, respectively. These increases did not persist for more than 5 minutes in either horse. Both horses were anesthetized with halothane. Dynamic baroreceptor sensitivity appeared severely obtunded (Fig. 2). Capillary re¢ll time started at 1.5 seconds and increased signi¢cantly (p < 0.001) as blood loss increased (Fig. 3). At the same time, mucous membrane color progressively deteriorated (Table 1).
Figure 2 Change in heart rate versus change in mean arterial pressure (MAP) as a measure of barore£ex sensitivity in nine horses during acute development of severe hemorrhagic shock.
Hematocrit Hematocrit decreased signi¢cantly (p < 0.02) between the awake (38 3.2%) and post-induction samples and then did not vary signi¢cantly for the remainder of the study (Table 1). In the two horses that experienced a transient increase in heart rate, an increase in hematocrit either occurred concurrently or 2 minutes after the increase in heart rate (one horse), and persisted for 6 minutes in one and until the end of the study (9 minutes) in the other horse. Total solids decreased signi¢cantly after # Association of Veterinary Anaesthetists, 2003, 30, 80^86
Figure 3 Capillary re¢ll time in nine horses during acute development of severe hemorrhagic shock. $Values signi¢cantly di¡erent (p < 0.05) from control values. 83
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
Figure 4 Arterial oxygen tension in nine horses during acute development of severe hemorrhagic shock. $Values signi¢cantly di¡erent (p < 0.05) from control values.
induction of anesthesia, and also decreased over the study period (Table 1). Arterial blood gas changes Partial pressure of oxygen in arterial blood decreased in an almost linear fashion over the period of blood loss (r ¼ 0.77), the values being signi¢cantly di¡erent from control in the last two samples collected (Fig. 4). Every horse demonstrated tachypnea on top of the breaths delivered by the ventilator by the time the pre-terminal sample was collected. Partial pressure of carbon dioxide did not vary signi¢cantly (Table 2). Base excess, bicarbonate and pH remained constant for most of the study period, but decreased signi¢cantly by the end of the study period (Table 2).
Discussion The results of this study demonstrate that in the anesthetized horse undergoing severe hemorrhage,
several clinically measured variables serve as indicators of volume change. These include arterial blood pressure and pulse pressure which decrease as the volume of blood lost increases. Mucous membrane color and capillary re¢ll time both change to re£ect a gradual reduction in peripheral perfusion. It is worth noting that this change in mucous membrane color occurs in the face of a constant hematocrit and thus relates directly to the change in peripheral perfusion rather than to a change in hemoglobin content of the blood. We did not standardize the rate of blood loss in our study.We aimed to standardize the measurement periods by collecting data when given amounts of blood loss had occurred, rather than at constant time periods. There was a 101-kg (22%) di¡erence between the smallest and largest horse in this study. Since the rate of blood loss was similar in all horses, this means that the smaller horses lost a greater proportion of blood volume more rapidly than the larger. The smallest and largest horses in the study manifested a similar rate of pressure change, so in the horses of our study, no relationship could be made between bodyweight and onset of hypotension. We have reported the blood loss in g kg1 or kg, since that is how it was measured. The speci¢c gravity of human blood is reported as being 1.06^1.09 g L1 (Dittmer 1961), so blood is 6^9% denser than water. It is also likely that the actual speci¢c gravity of the blood collected would have decreased slightly during the collection. Arterial blood pressure decreased over the period of blood loss in a triphasic manner. The fastest decrease in pressures occurred between 10 and 31 g kg1 of blood loss. A reduction in this rate of decrease occurred when the horses had lost more than 31 g kg1. At this point, the mean blood pressures were around 40 mm Hg, and stroke volume
Table 2 Arterial blood gas analyses in nine horses during acute, severe hemorrhage
Variable
Control
5 kg
10 kg
15 kg
20 kg
Pre-terminal
PaCO2 (mm Hg) PaCO2 (kPa) pH Bicarbonate (mmol L1) BE (mmol L1)
50 2.3 6.7 0.3 7.38 0.014 29.5 0.8 3.8 0.6
51 3.9 6.8 0.5 7.37 0.019 28.8 1.1 3.0 0.6
52 3.0 6.9 0.4 7.37 0.019 29.2 0.90 3.1 0.7
49 3.1 6.5 0.4 7.38 0.02 28.8 0.8 3.2 0.6
48 4.1 6.4 0.6 7.36 0.022 26.7 1.1 1.2 0.7
51 3.9 6.8 0.5 7.29 0.027 25 1.6 1.8 1.4
Values significantly different ( p < 0.01) from control values. Control ¼ under anesthesia and before surgery is started; 5, 10, 15, and 20 kg ¼ after 5, 10, 15, and 20 kg of blood, respectively, had been collected; pre-terminal ¼ sample collected just before the horse was euthanatized.
84
# Association of Veterinary Anaesthetists, 2003, 30, 80^86
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
would have been very much reduced. The rate of pressure decrease during hemorrhage would probably be di¡erent were these horses not anesthetized. Anesthetized horses are already prone to hypotension due to the fact that the inhalant anesthetics cause a decrease in cardiac output. This decrease is potentiated by the e¡ects of mechanical ventilation. Mechanical ventilation and anesthesia have been shown to decrease cardiac output by up to 50% from awake values in horses (Mizuno et al. 1994). An interesting and unexpected ¢nding in these horses was a progressive decrease in PaO2 that occurred as the volume of blood lost increased, with no signi¢cant change in PaCO2.We did not measure FIO2 but this is unlikely to have decreased signi¢cantly over the short time of the study. It is also unlikely that di¡usion impairment would have occurred in these horses. Thus, it would seem that one cause of a decreasing PaO2 in a previously stable horse is a signi¢cant reduction in blood volume or in functional cardiac output. It would have been interesting to measure cardiac output and mixed venous PO2 in these horses, they were undoubtedly low. Mixed venous oxygen tension has been shown to be a sensitive measure of volume change during hemorrhage in awake and anesthetized dogs (Scalea et al. 1988; Waisman et al. 1993). This decrease in mixed venous oxygen tension and cardiac output combined with ventilation perfusion mismatch and associated intrapulmonary shunt all contributed to the decrease in PaO2 in these horses. The ‘air hunger’ or terminal tachypnea that accompanies lethal hemorrhage is likely due to this hypoxia. About 53 g kg1 of blood was collected from the horses of this report. Using indicator dilution, the total blood volume of the awake horse has been measured as ranging from 62.4 to 109.6 mL kg1 (Courtice 1943; Julian et al. 1956; Kohn et al. 1978; Carlson et al. 1979; Naylor et al. 1993; Funkquist et al. 2001). Therefore, in the horses in our study, at least 48% and more likely up to 85% of the circulating blood volume was removed. We did not measure blood volume in these horses, and there is a lot of variability in this ¢gure between studies and between individuals. For these reasons, we elected not to present the blood loss from these horses as a per cent of blood volume. The response to acute blood loss in most species is tachycardia (Skillman et al. 1967; Schadt 1989; Syverud et al. 1989; Schmall et al. 1990; Waisman et al.1993;Thrasher & Shi¥ett 2001). The use of inha# Association of Veterinary Anaesthetists, 2003, 30, 80^86
lant rather than injectable anesthetics seems to modify the response of the horse to hemorrhage. Our ¢ndings would suggest that dynamic barore£ex sensitivity is not manifested in the inhalant-anesthetized horse during the development of hypotension. Hellyer et al. (1991) have shown that dynamic barore£ex sensitivity is preserved during an increase in arterial blood pressure under halothane anesthesia, but they did not investigate the e¡ects during a decrease in blood pressure. No apparent di¡erence in responses existed between the horse receiving iso£urane and those receiving halothane. In fact, both inhalants have been shown to abolish barore£ex sensitivity (Hellyer et al.1989). The amount of inhalant delivered to these horses was adjusted to keep the horses at a clinically equivalent anesthetic depth. Although we did not measure end-tidal anesthetic levels, there certainly were some small di¡erences in alveolar anesthetic concentrations between the horses. These may have slightly a¡ected the rate of change of blood pressure but did not change the overall trend. As expected, serial evaluation of hematocrit did not reveal any changes related to the loss of most of the circulating blood volume in these horses. In fact, with the loss of whole blood one would not expect any changes in hematocrit until associated £uid shifts occur.We also attribute this constant hematocrit to increased sympathetic tone and release of sequestered erythrocytes following the release of endogenous catecholamines (Thorten & Schalm 1964). In two horses, a transient large increase in hematocrit occurred concurrent with the development of transient tachycardia. This response was attributed to a large sympathetic response and associated release of erythrocytes from the splenic reservoir (Thorten & Schalm1964). Plasma total solids did decrease signi¢cantly over the study period in these horses, presumably due to some protein loss, £uid shifts and hemodilution that resulted from our administration of a low dose of £uids. We chose to administer lactated Ringer’s solution because this is part of the protocol followed when our plasma donors are anesthetized. We observed nine horses during severe blood loss. We found that heart rate and hematocrit were of no use in monitoring the progression of hemorrhage in these anesthetized horses. Arterial blood pressure, PaO2, pulse pressure, capillary re¢ll time and mucous membrane color were useful markers for assessing the severity of hemorrhage and volume of blood loss. 85
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
Acknowledgements The authors would like to thank Monica Smith and Drs Rachael Carpenter and George Bohart for technical assistance with this project. Also, Dr Joe Hauptman for assistance with statistical analysis.
References Byars TD, Divers TJ (1981) Clinical use of blood transfusions. Calif Vet 35,14^16. Carlson GP, Rumbaugh GE, Harrold DR (1979) Physiologic alterations produced by food and water deprivation during periods of high environmental temperatures. Am J Vet Res 40,982^985. Courtice RC (1943) The blood volume of normal animals. J Physiol102, 290^305. Dittmer D (1961) Blood and Other Body Fluids. Federation of American Societies for Experimental Biology, Washington, DC, USA, p.12. Durham AE (1996) Blood and plasma transfusions in the horse. EqVet Educ 8,8^12. Eicker SW, Ainsworth DM (1984) Equine plasma banking: collection by exsanguination. J Am Vet Med Assoc 185, 772^774. Funkquist P, Sandhagen B, Persson SGB et al. (2001) E¡ects of phlebotomy on haemodynamic characteristics during exercise in standardbred trotters with red cell hypervolaemia. EqVet J 33, 417^424. Hellyer PW, Bednarski RM, HubbellJAE et al. (1989) E¡ects of halothane and iso£urane on barore£ex sensitivity in horses. Am J Vet Res 50, 2127^2134. Hellyer PW, Dodam JR, Light GS (1991) Dynamic barore£ex sensitivity in anesthetized horses, maintained at1.25^1.3 minimal alveolar concentration of halothane. Am J Vet Res 52,1672^1675. Julian LM, Lawrence JH, Berlin NI et al. (1956) Blood volume, body water and body fat of the horse. J Appl Physiol 8, 651^653.
86
Kohn CW, Muir WW, Sams R (1978) Plasma volume and extra-cellular £uid volume in horses at rest and following exercise. Am J Vet Res 39,871^874. MizunoY, Aida H, Hara H et al. (1994) Cardiovascular e¡ects of intermittent positive pressure ventilation in the anesthetized horse. J Vet Med Sci 56,39^44. NaylorJR, BaylyWM, Schott HC II et al. (1993) Equine plasma and blood volumes decrease with dehydration but subsequently increase with exercise. J Appl Physiol 75, 1002^1008. ScaleaTM, Holman M, Fuortes M et al. (1988) Central venous blood oxygen saturation: an early accurate measurement of volume during hemorrhage. J Trauma 28,725^732. Schadt JC (1989) Sympathetic and hemodynamic adjustments to hemorrhage: a possible role for endogenous opioid peptides. Resuscitation18, 219^228. Schmall LM, MuirWW, Robertson JT (1990) Haemodynamic e¡ects of small volume hypertonic saline in experimentally induced haemorrhagic shock. EqVet J 22, 273^277. Skillman JJ, Olson JE, Lyons JH et al. (1967) The hemodynamic e¡ect of acute blood loss in normal man, with observations on the e¡ect of the Valsalva maneuver and breath holding. Ann Surg166,713^738. Syverud SA, Dronen SC, Chudnofsky CR et al. (1989) A continuous hemorrhage model of fatal hemorrhagic shock in swine. Resuscitation17, 287^295. Thorten M, Schalm OW (1964) In£uence of the equine spleen on rapid changes in the concentration of erythrocytes in peripheral blood. Am J Vet Res 25,500^503. Thrasher TN, Shi¥ett C (2001) E¡ect of carotid or aortic baroreceptor denervation on arterial pressure during hemorrhage in conscious dogs. Am J Physiol Regul Integr Comp Physiol 280, R1642^R1649. Trim CM, Eaton SA, Parks AH (1997) Severe nasal hemorrhage in an anesthetized horse. J Am Vet Med Assoc 210, 1324^1327. Waisman Y, Eichacker PQ, Banks SM et al. (1993) Acute hemorrhage in dogs: construction and validation of models to quantify blood loss. J Appl Physiol 74,510^519. Received13 December 2001; accepted 2 May 2002.
# Association of Veterinary Anaesthetists, 2003, 30, 80^86
The cardiopulmonary effects of severe blood loss in anesthetized horses Deborah V Wilson
BVSc, MS, Diplomate ACVA, Yves
Rondenay
DVM
& Phyllis U Shance
LVT
Department of Large Animal Clinical Sciences, Michigan State University, East Lansing, MI, USA
Correspondence: DV Wilson, Department of Large Animal Clinical Sciences, Michigan State University, East Lansing, MI 48864, USA. E-mail: [email protected]
Abstract Objective To characterize the acute cardiopulmonary e¡ects of severe hemorrhage in anesthetized horses. Study design Prospective experimental study. Animals Three geldings and six mares, aged 14.4 2.7 years, weighing 486 41 kg (range: 425^550 kg). Methods Horses were anesthetized using xylazine, guaifenesin, ketamine and halothane or iso£urane. Cardiovascular variables, hematocrit, total solids, capillary re¢ll time (CRT) and color of mucous membranes were measured as blood was collected from the carotid artery into sterile plastic bags. Arterial blood gas analysis was also performed. Results The average amount of blood collected from these horses was (mean SD) 53 4.8 mL kg1 bodyweight (range: 23^32 kg) over 39 4 minutes. Hematocrit decreased from 38 3 to 32 2% after induction of anesthesia and did not change signi¢cantly over the period of blood loss. Total solids decreased signi¢cantly after induction of anesthesia, and over the period of blood loss. Systolic, mean, diastolic and pulse pressures decreased as blood was lost. Heart rate did not change signi¢cantly. Capillary re¢ll time increased from 1.6 0.4 seconds to
Present address: Department of Clinical Sciences, College of Veterinary Medicine, University of Montreal, Montreal, Canada. 80
4.8 1.3 seconds as blood loss increased. Mucous membrane color deteriorated progressively. Arterial PO2 decreased signi¢cantly over the period of blood loss. Conclusions Hematocrit and heart rate do not change signi¢cantly during acute severe hemorrhage in the anesthetized horse. Arterial blood pressure, pulse pressure and PaO2 decrease as blood loss increases. Changes in mucous membrane color and CRT also occur as blood loss increases. Clinical relevance During severe hemorrhage in the inhalant-anesthetized horse, both heart rate and hematocrit remain unchanged. Blood pressure decreases and changes in arterial PO2 correlate most strongly with volume of blood lost. Keywords anesthetized horse, arterial blood pressure, baroreceptor sensitivity, cardiopulmonary, hemorrhage, hypoxia.
Introduction The accurate quanti¢cation of blood loss can be di⁄cult in the anesthetized horse. If the shed blood is not collected, the volume lost can be only poorly approximated. There are little or no data validating clinically measured variables as indicators of volume change during acute hemorrhage in the anesthetized horse. It is generally considered that a conscious 500 kg horse may lose a third of its circulating blood volume (12^16 L) before signi¢cant cardiovascular changes occur (Byars & Divers 1981; Durham 1996).
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
General anesthesia, inhalant anesthesia and the associated normal reduction of cardiac output are likely to reduce this volume signi¢cantly. A study in dogs showed that most commonly used hemodynamic and blood indices are inaccurate when quantifying hemorrhage (Waisman et al. 1993). In that study, the best indicator of volume change was variation in acid^base status re£ected by arterial base de¢cit, bicarbonate and venous pH (Waisman et al. 1993). In people, a decrease in pulse pressure and decreased area under the arterial pulse contour have been shown to accompany blood loss (Skillman et al. 1967). Tachycardia and reduced arterial pressure are reported in response to severe hemorrhage in one group of horses anesthetized with guaifenesin and barbiturates (Schmall et al. 1990). A case report of life-threatening hemorrhage during iso£urane anesthesia in a horse described arterial hypotension, but heart rate was not reported (Trim et al.1997). We know from clinical experience that heart rate in the inhalant-anesthetized horse is notoriously stable in the face of anesthetic-induced hypotension. This phenomenon has been described by Hellyer et al. (1989, 1991) who showed that both halothane and iso£urane anesthesia cause a signi¢cant decrease or abolition of barore£ex sensitivity in the horse. The dynamic function of this re£ex during onset of arterial hypotension accompanying severe hemorrhage has not been evaluated. We hypothesized that heart rate would not change during blood loss in anesthetized horses. Our objective with this study was to measure a number of cardiopulmonary and hematologic variables during acute blood loss in anesthetized horses, and to correlate these variables with volume lost.
Materials and methods Horses This study was approved by the All University Committee on Animal Use and Care at Michigan State University. Nine horses were included in this study: three Quarter horses, two Thoroughbred crosses, one Pinto, one Trakehner and two mixed breeds. There were three geldings and six mares in the study, aged 14.4 2.7 years, and weights ranged from 425 to 550 kg (486 41 kg). All horses were determined to be free from evidence of systemic disease by physical examination prior to the study. These horses were scheduled for terminal blood collection as plasma donors. # Association of Veterinary Anaesthetists, 2003, 30, 80^86
Anesthetic protocol The horses had a catheter placed in the right jugular vein and sutured in position following local anesthesia and surgical preparation of the area. The horses were sedated with intravenous xylazine until tractable (0.3 mg kg1 initial dose; Xylazine-100 injectable, The Butler Co., Columbus, OH, USA). Guaifenesin (Guaifenesin injection, Phoenix Scienti¢c Inc., St. Joseph, MO, USA) was administered intravenously until signs of muscle relaxation were observed. Then, a bolus dose of ketamine (2 mg kg1; KetaFlo, Abbott Laboratories, North Chicago, IL, USA) was administered intravenously; additional doses were given as required to facilitate intubation or to deepen anesthesia. The horses were placed in right lateral recumbency. Endotracheal intubation was performed using a cu¡ed tube, and the horses were connected to a semiclosed breathing circuit. The horses were mechanically ventilated using 6^ 10 breaths minute1 and a peak airway pressure <20 cm H2O. Halothane or iso£urane in oxygen (1^ 3% and 3^6 L minute1, respectively) was used to maintain anesthesia. Lactated Ringer’s solution was administered via the catheter in the right jugular vein at a rate of 10 mL kg1 hour1 for the duration of the procedure. A lethal dose of pentobarbital was administered IVat the end of the procedure. Monitoring A catheter was placed in the lateral metatarsal artery for monitoring blood pressure. The transducer was zeroed to the level of the left atrium (PB 240 Puritan Bennett Corp.,Wilmington, MA, USA). Blood was collected from the same catheter for gas analysis. Oral mucous membrane color was assessed using the color choices of: pink, pale pink, white/blanched, gray or blue. Capillary re¢ll time was assessed by standard means. A modi¢ed base^apex electrocardiogram was monitored, and heart rate (HR) was calculated from an R-to-R interval. The change in HR from the baseline value was plotted as a function of the percent change in mean arterial pressure (MAP) from baseline (Hellyer et al.1989,1991). Blood collection The collection procedure was performed by surgically placing a cannula into the carotid artery, as described previously (Eicker & Ainsworth 1984). Blood was collected into sterile 3- and 5-L bags 81
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
containing 4% sodium citrate. The bags were weighed immediately prior to blood collection and remained on the scale during ¢lling. The progression of blood collection was monitored by the weight increase of the bag. The collection line was brie£y clamped (5^10 seconds) each time the collection bags were changed. A catheter was placed in the saphenous vein for sample collection to determine hematocrit and total solids. Sample and data collection Baseline values of the heart and respiratory rates, hematocrit and total solids (microhematocrit centrifugation and refractometry) were determined before placement of the jugular venous catheter. After induction of anesthesia, the heart rate, respiratory rate, and systolic, diastolic and mean arterial pressures were recorded every 5 minutes until the start of the blood collection. During the period of blood collection, the cardiovascular values were recorded and samples for hematocrit and total solids were collected at the end of each kilogram collected. The time was recorded at each of the sample periods. Arterial blood was collected for blood gas analysis before the start of surgery, and at the end of each 5 kg of blood collected. A sample was also collected just before the horse was euthanatized at the end of the blood collection period. All data are presented as mean SEM, unless otherwise stated. Parametric data were analyzed using a one-way repeated measures ANOVA. When there were signi¢cant (p < 0.05) main e¡ects, the post hoc test used was the Tukey’s test. Linear regression was performed where applicable. Statistical analysis was conducted with Sigma StatTM, version 1.03 (Sigma StatTM for Windows, Jandel Scienti¢c, San Rafael, CA, USA).
blood collection commenced. Time from induction of anesthesia until administration of a lethal dose of pentobarbital was100 5.7 minutes. Blood collection The average blood volume collected from these horses was 52.5 1.5 g kg1 bodyweight (range: 46^ 58 g kg1).Total amount collected was 26 0.88 kg1 (range: 23^32 kg1). The average duration of the blood collection period was 39 4 minutes. The collection bag was the same distance below heart level in all horses and the £ow rate was not controlled. The initial rate of blood loss was 1.5 0.2 g kg1 minute1 and this did not vary signi¢cantly for most of the study period. Near the end, when more than 25 kg had been lost, the bleeding rate had decreased to 1 0.2 g kg1 minute1. The horses received a total of 12 2.4 mL kg1 lactated Ringer’s solution over the duration of the anesthetic period. Cardiovascular e¡ects A signi¢cant decrease in mean arterial, systolic, diastolic and pulse pressures occurred as blood was lost (Fig. 1). Systolic blood pressure began to
Results Anesthesia The horses received a total dose of 0.5 0.04 mg kg1 bodyweight of xylazine intravenously prior to induction. Guaifenesin and ketamine were administered intravenously at doses of 79 4.3 and 3.1 0.14 mg kg1 bodyweight, respectively, to induce anesthesia. In eight of the horses, anesthesia was maintained with halothane in oxygen, and iso£urane was used in one horse. These horses were under anesthesia for 61 4.6 minutes whilst the surgical approach was made to the carotid artery, and before 82
Figure 1 Systolic, mean and diastolic arterial blood pressures in nine anesthetized horses during acute development of severe hemorrhagic shock. $Values signi¢cantly di¡erent (p < 0.05) from control values. # Association of Veterinary Anaesthetists, 2003, 30, 80^86
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
Table 1 Hematocrit, plasma total solids, heart rate and mucous membrane color in nine horses during acute, severe hemorrhage
Variable
Control
5 kg
10 kg
15 kg
20 kg
Pre-terminal
Heart rate (beats minute1) Mucous membrane color Hematocrit (vol.%) Total solids (g dL1)
34 1 Pink 32 2 6.0 0.2
32 1 Pink 32 1 5.7 0.2
33 1 Pale pink 33 2 5.6 0.2
34 3 White 31 2 5.4 0.2
32 1 White 32 4 5.2 0.2
38 5 Blue 36 4 5.1 0.2
Values significantly different ( p < 0.01) from control values. Control ¼ under anesthesia and before surgery is started; 5, 10, 15, and 20 kg ¼ after 5, 10, 15, and 20 kg of blood, respectively, had been collected; pre-terminal ¼ sample collected just before the horse was euthanatized. All data are mean SEM except mucous membrane color which is the median.
decrease slowly in six horses as soon as blood loss commenced, eventually decreasing in all horses as blood loss exceeded 10 g kg1 (r ¼ 0.63). The rate of decrease was maximal between 10 and 31 g kg1 of blood loss. Pulse pressure decreased (r ¼ 0.6) in a similar fashion throughout the blood collection period (Fig. 1). Diastolic pressure decreased more slowly than systolic pressure; this decrease was signi¢cant by the time16 kg of blood had been lost. Baseline heart rate under anesthesia was 34 1 beats minute1 and heart rate did not vary signi¢cantly over the blood collection period (Table 1). An increase in heart rate occurred transiently in two horses (from 35 to110 and 28 to 200 beats minute1) at total blood losses of 19 and 25 kg, respectively. These increases did not persist for more than 5 minutes in either horse. Both horses were anesthetized with halothane. Dynamic baroreceptor sensitivity appeared severely obtunded (Fig. 2). Capillary re¢ll time started at 1.5 seconds and increased signi¢cantly (p < 0.001) as blood loss increased (Fig. 3). At the same time, mucous membrane color progressively deteriorated (Table 1).
Figure 2 Change in heart rate versus change in mean arterial pressure (MAP) as a measure of barore£ex sensitivity in nine horses during acute development of severe hemorrhagic shock.
Hematocrit Hematocrit decreased signi¢cantly (p < 0.02) between the awake (38 3.2%) and post-induction samples and then did not vary signi¢cantly for the remainder of the study (Table 1). In the two horses that experienced a transient increase in heart rate, an increase in hematocrit either occurred concurrently or 2 minutes after the increase in heart rate (one horse), and persisted for 6 minutes in one and until the end of the study (9 minutes) in the other horse. Total solids decreased signi¢cantly after # Association of Veterinary Anaesthetists, 2003, 30, 80^86
Figure 3 Capillary re¢ll time in nine horses during acute development of severe hemorrhagic shock. $Values signi¢cantly di¡erent (p < 0.05) from control values. 83
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
Figure 4 Arterial oxygen tension in nine horses during acute development of severe hemorrhagic shock. $Values signi¢cantly di¡erent (p < 0.05) from control values.
induction of anesthesia, and also decreased over the study period (Table 1). Arterial blood gas changes Partial pressure of oxygen in arterial blood decreased in an almost linear fashion over the period of blood loss (r ¼ 0.77), the values being signi¢cantly di¡erent from control in the last two samples collected (Fig. 4). Every horse demonstrated tachypnea on top of the breaths delivered by the ventilator by the time the pre-terminal sample was collected. Partial pressure of carbon dioxide did not vary signi¢cantly (Table 2). Base excess, bicarbonate and pH remained constant for most of the study period, but decreased signi¢cantly by the end of the study period (Table 2).
Discussion The results of this study demonstrate that in the anesthetized horse undergoing severe hemorrhage,
several clinically measured variables serve as indicators of volume change. These include arterial blood pressure and pulse pressure which decrease as the volume of blood lost increases. Mucous membrane color and capillary re¢ll time both change to re£ect a gradual reduction in peripheral perfusion. It is worth noting that this change in mucous membrane color occurs in the face of a constant hematocrit and thus relates directly to the change in peripheral perfusion rather than to a change in hemoglobin content of the blood. We did not standardize the rate of blood loss in our study.We aimed to standardize the measurement periods by collecting data when given amounts of blood loss had occurred, rather than at constant time periods. There was a 101-kg (22%) di¡erence between the smallest and largest horse in this study. Since the rate of blood loss was similar in all horses, this means that the smaller horses lost a greater proportion of blood volume more rapidly than the larger. The smallest and largest horses in the study manifested a similar rate of pressure change, so in the horses of our study, no relationship could be made between bodyweight and onset of hypotension. We have reported the blood loss in g kg1 or kg, since that is how it was measured. The speci¢c gravity of human blood is reported as being 1.06^1.09 g L1 (Dittmer 1961), so blood is 6^9% denser than water. It is also likely that the actual speci¢c gravity of the blood collected would have decreased slightly during the collection. Arterial blood pressure decreased over the period of blood loss in a triphasic manner. The fastest decrease in pressures occurred between 10 and 31 g kg1 of blood loss. A reduction in this rate of decrease occurred when the horses had lost more than 31 g kg1. At this point, the mean blood pressures were around 40 mm Hg, and stroke volume
Table 2 Arterial blood gas analyses in nine horses during acute, severe hemorrhage
Variable
Control
5 kg
10 kg
15 kg
20 kg
Pre-terminal
PaCO2 (mm Hg) PaCO2 (kPa) pH Bicarbonate (mmol L1) BE (mmol L1)
50 2.3 6.7 0.3 7.38 0.014 29.5 0.8 3.8 0.6
51 3.9 6.8 0.5 7.37 0.019 28.8 1.1 3.0 0.6
52 3.0 6.9 0.4 7.37 0.019 29.2 0.90 3.1 0.7
49 3.1 6.5 0.4 7.38 0.02 28.8 0.8 3.2 0.6
48 4.1 6.4 0.6 7.36 0.022 26.7 1.1 1.2 0.7
51 3.9 6.8 0.5 7.29 0.027 25 1.6 1.8 1.4
Values significantly different ( p < 0.01) from control values. Control ¼ under anesthesia and before surgery is started; 5, 10, 15, and 20 kg ¼ after 5, 10, 15, and 20 kg of blood, respectively, had been collected; pre-terminal ¼ sample collected just before the horse was euthanatized.
84
# Association of Veterinary Anaesthetists, 2003, 30, 80^86
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
would have been very much reduced. The rate of pressure decrease during hemorrhage would probably be di¡erent were these horses not anesthetized. Anesthetized horses are already prone to hypotension due to the fact that the inhalant anesthetics cause a decrease in cardiac output. This decrease is potentiated by the e¡ects of mechanical ventilation. Mechanical ventilation and anesthesia have been shown to decrease cardiac output by up to 50% from awake values in horses (Mizuno et al. 1994). An interesting and unexpected ¢nding in these horses was a progressive decrease in PaO2 that occurred as the volume of blood lost increased, with no signi¢cant change in PaCO2.We did not measure FIO2 but this is unlikely to have decreased signi¢cantly over the short time of the study. It is also unlikely that di¡usion impairment would have occurred in these horses. Thus, it would seem that one cause of a decreasing PaO2 in a previously stable horse is a signi¢cant reduction in blood volume or in functional cardiac output. It would have been interesting to measure cardiac output and mixed venous PO2 in these horses, they were undoubtedly low. Mixed venous oxygen tension has been shown to be a sensitive measure of volume change during hemorrhage in awake and anesthetized dogs (Scalea et al. 1988; Waisman et al. 1993). This decrease in mixed venous oxygen tension and cardiac output combined with ventilation perfusion mismatch and associated intrapulmonary shunt all contributed to the decrease in PaO2 in these horses. The ‘air hunger’ or terminal tachypnea that accompanies lethal hemorrhage is likely due to this hypoxia. About 53 g kg1 of blood was collected from the horses of this report. Using indicator dilution, the total blood volume of the awake horse has been measured as ranging from 62.4 to 109.6 mL kg1 (Courtice 1943; Julian et al. 1956; Kohn et al. 1978; Carlson et al. 1979; Naylor et al. 1993; Funkquist et al. 2001). Therefore, in the horses in our study, at least 48% and more likely up to 85% of the circulating blood volume was removed. We did not measure blood volume in these horses, and there is a lot of variability in this ¢gure between studies and between individuals. For these reasons, we elected not to present the blood loss from these horses as a per cent of blood volume. The response to acute blood loss in most species is tachycardia (Skillman et al. 1967; Schadt 1989; Syverud et al. 1989; Schmall et al. 1990; Waisman et al.1993;Thrasher & Shi¥ett 2001). The use of inha# Association of Veterinary Anaesthetists, 2003, 30, 80^86
lant rather than injectable anesthetics seems to modify the response of the horse to hemorrhage. Our ¢ndings would suggest that dynamic barore£ex sensitivity is not manifested in the inhalant-anesthetized horse during the development of hypotension. Hellyer et al. (1991) have shown that dynamic barore£ex sensitivity is preserved during an increase in arterial blood pressure under halothane anesthesia, but they did not investigate the e¡ects during a decrease in blood pressure. No apparent di¡erence in responses existed between the horse receiving iso£urane and those receiving halothane. In fact, both inhalants have been shown to abolish barore£ex sensitivity (Hellyer et al.1989). The amount of inhalant delivered to these horses was adjusted to keep the horses at a clinically equivalent anesthetic depth. Although we did not measure end-tidal anesthetic levels, there certainly were some small di¡erences in alveolar anesthetic concentrations between the horses. These may have slightly a¡ected the rate of change of blood pressure but did not change the overall trend. As expected, serial evaluation of hematocrit did not reveal any changes related to the loss of most of the circulating blood volume in these horses. In fact, with the loss of whole blood one would not expect any changes in hematocrit until associated £uid shifts occur.We also attribute this constant hematocrit to increased sympathetic tone and release of sequestered erythrocytes following the release of endogenous catecholamines (Thorten & Schalm 1964). In two horses, a transient large increase in hematocrit occurred concurrent with the development of transient tachycardia. This response was attributed to a large sympathetic response and associated release of erythrocytes from the splenic reservoir (Thorten & Schalm1964). Plasma total solids did decrease signi¢cantly over the study period in these horses, presumably due to some protein loss, £uid shifts and hemodilution that resulted from our administration of a low dose of £uids. We chose to administer lactated Ringer’s solution because this is part of the protocol followed when our plasma donors are anesthetized. We observed nine horses during severe blood loss. We found that heart rate and hematocrit were of no use in monitoring the progression of hemorrhage in these anesthetized horses. Arterial blood pressure, PaO2, pulse pressure, capillary re¢ll time and mucous membrane color were useful markers for assessing the severity of hemorrhage and volume of blood loss. 85
Cardiopulmonary e¡ects of severe blood loss in anesthetized horses DV Wilson et al.
Acknowledgements The authors would like to thank Monica Smith and Drs Rachael Carpenter and George Bohart for technical assistance with this project. Also, Dr Joe Hauptman for assistance with statistical analysis.
References Byars TD, Divers TJ (1981) Clinical use of blood transfusions. Calif Vet 35,14^16. Carlson GP, Rumbaugh GE, Harrold DR (1979) Physiologic alterations produced by food and water deprivation during periods of high environmental temperatures. Am J Vet Res 40,982^985. Courtice RC (1943) The blood volume of normal animals. J Physiol102, 290^305. Dittmer D (1961) Blood and Other Body Fluids. Federation of American Societies for Experimental Biology, Washington, DC, USA, p.12. Durham AE (1996) Blood and plasma transfusions in the horse. EqVet Educ 8,8^12. Eicker SW, Ainsworth DM (1984) Equine plasma banking: collection by exsanguination. J Am Vet Med Assoc 185, 772^774. Funkquist P, Sandhagen B, Persson SGB et al. (2001) E¡ects of phlebotomy on haemodynamic characteristics during exercise in standardbred trotters with red cell hypervolaemia. EqVet J 33, 417^424. Hellyer PW, Bednarski RM, HubbellJAE et al. (1989) E¡ects of halothane and iso£urane on barore£ex sensitivity in horses. Am J Vet Res 50, 2127^2134. Hellyer PW, Dodam JR, Light GS (1991) Dynamic barore£ex sensitivity in anesthetized horses, maintained at1.25^1.3 minimal alveolar concentration of halothane. Am J Vet Res 52,1672^1675. Julian LM, Lawrence JH, Berlin NI et al. (1956) Blood volume, body water and body fat of the horse. J Appl Physiol 8, 651^653.
86
Kohn CW, Muir WW, Sams R (1978) Plasma volume and extra-cellular £uid volume in horses at rest and following exercise. Am J Vet Res 39,871^874. MizunoY, Aida H, Hara H et al. (1994) Cardiovascular e¡ects of intermittent positive pressure ventilation in the anesthetized horse. J Vet Med Sci 56,39^44. NaylorJR, BaylyWM, Schott HC II et al. (1993) Equine plasma and blood volumes decrease with dehydration but subsequently increase with exercise. J Appl Physiol 75, 1002^1008. ScaleaTM, Holman M, Fuortes M et al. (1988) Central venous blood oxygen saturation: an early accurate measurement of volume during hemorrhage. J Trauma 28,725^732. Schadt JC (1989) Sympathetic and hemodynamic adjustments to hemorrhage: a possible role for endogenous opioid peptides. Resuscitation18, 219^228. Schmall LM, MuirWW, Robertson JT (1990) Haemodynamic e¡ects of small volume hypertonic saline in experimentally induced haemorrhagic shock. EqVet J 22, 273^277. Skillman JJ, Olson JE, Lyons JH et al. (1967) The hemodynamic e¡ect of acute blood loss in normal man, with observations on the e¡ect of the Valsalva maneuver and breath holding. Ann Surg166,713^738. Syverud SA, Dronen SC, Chudnofsky CR et al. (1989) A continuous hemorrhage model of fatal hemorrhagic shock in swine. Resuscitation17, 287^295. Thorten M, Schalm OW (1964) In£uence of the equine spleen on rapid changes in the concentration of erythrocytes in peripheral blood. Am J Vet Res 25,500^503. Thrasher TN, Shi¥ett C (2001) E¡ect of carotid or aortic baroreceptor denervation on arterial pressure during hemorrhage in conscious dogs. Am J Physiol Regul Integr Comp Physiol 280, R1642^R1649. Trim CM, Eaton SA, Parks AH (1997) Severe nasal hemorrhage in an anesthetized horse. J Am Vet Med Assoc 210, 1324^1327. Waisman Y, Eichacker PQ, Banks SM et al. (1993) Acute hemorrhage in dogs: construction and validation of models to quantify blood loss. J Appl Physiol 74,510^519. Received13 December 2001; accepted 2 May 2002.
# Association of Veterinary Anaesthetists, 2003, 30, 80^86