Introduction
Cardiac output is the product of heart rate and stroke volume and is the amount of blood that the heart pumps each minute. In horses, direct blood pressure measurement is technically easy and is considered a standard of care for anesthetized patients, whereas measurement of blood flow is more difficult. This has led to undue interest in blood pressure measurements. However, all organs require blood flow, and blood flow is as important as or more important than blood pressure for tissue health.1 Tissue perfusion, which is synonymous with blood flow within tissues, would be an ideal cardiovascular measurement for monitoring anesthetized horses because it could be used to optimize administration of IV fluids and guide use of inotropes and pressors. Currently, in horses, direct and indirect methods for measurement of CO are technically difficult, compared with direct blood pressure measurement. Thus, development of a simplified method for assessing CO in horses is warranted.
Many methods for measurement of CO have been developed and evaluated in horses including the Fick principle, ultrasonography, and indicator dilution.2–5 The thermodilution technique has been the gold standard for measurement of CO since its introduction in 1970.6, 7 The thermodilution technique requires placement of a Swan-Ganz catheter in the patient and injection of chilled physiologic saline (0.9% NaCl) solution (injectate) of a known temperature and volume through a proximal catheter port directly into the right atrium where it mixes with the atrial blood. The saline solution cools the blood as it passes through the right ventricle and into the pulmonary artery. The temperature of the blood-saline solution admixture is then measured by a thermistor at the tip of the Swan-Ganz catheter, which is located in the pulmonary artery. Computer software is used to create a thermodilution profile from which the CO is derived by means of the Stewart-Hamilton equation, which involves the blood-saline solution admixture temperature and injectate temperature and volume.8 In general, the recommended injectate volume for CO measurement is 1 mL of ice-chilled (< 5 °C) saline solution/15 kg of patient body weight.9 That standard volume maximizes the signal-to-noise ratio in the time-temperature thermodilution curve.9 However, the size and hemodynamic stability of the patient should be considered when the volume and temperature of thermodilution injectate used for CO measurement are selected.10
The recommended thermodilution technique can be modified by administration of injectate at room temperature or in a smaller volume (< 1 mL/15 kg). Potential advantages of administration of injectate at room temperature include immediate amenability and convenience of use of physiologic saline solution off the shelf, stability of injectate temperature throughout injection, and avoidance of dysrhythmias.11, 12 Use of a smaller volume of injectate may be beneficial for patients with compromised cardiac or renal function. In human patients with congestive heart failure or renal failure, smaller volumes of injectate are used to prevent volume overload and the development of pulmonary edema.7,9,10
In large animals, a particular challenge associated with the thermodilution technique for measurement of CO is the ability to infuse an adequate volume of injectate during a given period of time to generate a diagnostic thermodilution curve. A typical Swan-Ganz catheter has a circumference of approximately 7.33 mm (7F) and length of 110 cm. For large horses, injection of an adequate volume of injectate through a typical-sized Swan-Ganz catheter during the constrained period required for generation of a diagnostic thermodilution curve can be difficult.
For horses, studies are lacking regarding the use of injectate at a warmer temperature (> 5 °C) and in smaller volumes (< 1 mL/15 kg) than those currently recommended for measurement of the CO by the thermodilution method. The objective of the prospective study reported here was to compare the within-subject variance for CO in healthy anesthetized adult horses when measured by 4 thermodilution protocols (injection of physiologic saline solution chilled to < 5 °C at volumes of 1 mL/15 kg, 1 mL/25 kg, and 1 mL/35 kg and injection of physiologic saline solution maintained near room temperature [17 °C] at a volume of 1 mL/15 kg). We hypothesized that the within-subject variance for COs derived when the injectate was maintained near room temperature or administered at volumes of 1 mL/25 kg and 1 mL/35 kg would be acceptable13 (bias, < 20%), compared with that when the CO was derived from the currently recommended standard thermodilution protocol14 (injectate temperature < 5 °C and volume 1 mL/15 kg).
Materials and Methods
Animals
All study procedures were reviewed and approved by the University of Pennsylvania Institutional Animal Care and Use Committee (protocol No. 806775). Results of an a priori power analysis and sample size calculation indicated that 8 horses were necessary to detect clinically relevant changes in CO between thermodilution protocols assuming normally distributed data, an SD of 15%, a type I error of 0.05, and a type II error of 0.2.
Eight university-owned horses with a mean ± SD body weight of 525 ± 41 kg and age of 8 ± 4 years were enrolled in the study. All horses were concurrently enrolled in an unrelated terminal study and were determined to be healthy on the basis of results of a physical examination conducted prior to study initiation and anesthesia. Horses were individually housed in stalls (3.5 X 3.5 m) for 12 to 24 hours prior to anesthesia induction and were fed hay. Food but not water was withheld for 8 hours before anesthesia induction.
Catheter placement
For each horse prior to anesthesia induction, the hair was clipped from the skin over both jugular veins, and the skin was aseptically prepared for catheter placement. The skin at the catheter insertion sites was infiltrated with 2 mL of a 2% lidocaine solution.a A 12-gauge, 5-cm-long catheter was placed in the left jugular vein. Two 8F catheter introducers were placed in the right jugular vein to facilitate placement of 2 separate standard 7F 110-cm-long Swan-Ganz catheters.b The Swan-Ganz catheters were inserted so that the tip of one was positioned in the pulmonary artery and the tip of the other was positioned in the right atrium. Correct placement of the Swan-Ganz catheters was confirmed by visual inspection of pressure waveforms as described.9
Anesthesia protocol
Each horse was premedicated (sedated) with xylazinea (0.8 mg/kg, IV). Anesthesia was induced with midazolamc (0.05 mg/kg, IV) and ketaminea (2.2 mg/kg, IV). The horse was then intubated with an appropriately sized cuffed endotracheal tube and positioned in dorsal recumbency on a padded large animal surgical table. Anesthesia was maintained with isofluranea in oxygen delivered by a circle rebreathing circuit, with an end-tidal isoflurane concentration of 1.3% targeted. An IV crystalloid solutiond was administered at a rate of 5 mL/kg/h for the entire anesthesia period, and dobutaminee (0.5 or 2.5 µg/kg/min, IV) was infused throughout 4 of the 5 measurement cycles. Following completion of all experimental procedures and measurements, the horse was euthanized with potassium chloride (1.5 mmol/kg, IV) in accordance with the AVMA Guidelines for the Euthanasia of Animals.15 Death was confirmed by the absence of an auscultable heart beat and an ECG tracing consistent with asystole.
Experimental protocol
For each horse, the CO was derived from each of 4 thermodilution protocols (IV injection of physiologic saline solution chilled to < 5 °C at volumes of 1 mL/15 kg of body weight [protocol A; control], 1 mL/25 kg [protocol B], and 1 mL/35 kg [protocol C] and injection of physiologic saline solution maintained at approximately 17 °C at a volume of 1 mL/15 kg [protocol D]). Immediately before measurement of the CO by each experimental protocol, the core body temperature of the horse was measured by the thermistor at the tip of the Swan-Ganz catheter positioned in the pulmonary artery. The saline solution was manually injected through the Swan-Ganz catheter that was positioned in the right atrium at the end of expiration. The temperature of the injectate was measured by means of an inline temperature probe, and the resulting temperature change in the pulmonary artery was analyzed and the CO calculated by a multiparameter patient monitor.f
The CO was measured by each protocol one after another during each of 5 measurement cycles, with a 30-minute interval between measurement cycles. Each protocol was repeated 3 times during each measurement cycle, and the resulting calculated mean CO for each protocol was used for analysis purposes. During each measurement cycle, protocol A was always performed first and protocols B, C, and D were performed in a randomized order as determined by a random number generator.
To increase the variation in CO measurements, the dobutamine infusion rate was increased to and maintained at 2.5 µg/kg/min for the entirely of the fourth measurement cycle. The dobutamine infusion was discontinued and the end-tidal isoflurane concentration was increased from 1.3% to 1.8% as soon as all CO measurments were obtained for the fourth measurement cycle to allow for hemodynamic stabilization during the subsequent 30-minute interval before CO measurements were obtained during the fifth measurement cycle.2,16,17
Statistical analysis
Descriptive data were generated for each of the 4 experimental protocols. Variables of interest included the CO and within-subject variance and width of the range for CO measurements. The data distribution for each variable of interest was assessed for normality by visual examination of histograms and residual plots and the Shapiro-Wilk test. All variables were normally distributed and were summarized as the mean ± SD unless otherwise stipulated. Each variable of interest was compared among the 4 experimental protocols by means of a repeated-measures ANOVA. When necessary, post hoc pairwise comparisons were performed with the Newman-Keuls test.
Additionally, the Pearson correlation coefficient (r) was calculated between protocol A (injectate temperature < 5 °C and volume 1 mL/15 kg [control]) and each of the other 3 protocols (B, C, and D). All analyses were performed with commercially available statistical software programs,g,h and values of P < 0.05 were considered significant.
Results
All 4 experimental protocols were well tolerated by the horses, as evidenced by the fact that none of the horses developed arrhythmias while the thermodilution protocols were being performed. The mean ± SD core body temperature measured in the pulmonary artery was 36.8 ± 1.4 °C immediately prior to CO measurments during the first measurement cycle and decreased to 35.2 ± 1.9 °C following completion of the last CO measurement of the fifth measurement cycle. For all 8 study horses, the mean ± SD injectate volume administered was 35 ± 3 mL for protocols A (injectate temperature < 5 °C and volume 1 mL/15 kg [control]) and D (injectate temperature 17 °C and volume 1 mL/15 kg), 21 ± 2 mL for protocol B (injectate temperature < 5 °C and volume 1 mL/25 kg), and 15 ± 1 mL for protocol C (injectate temperature < 5 °C and volume 1 mL/35 kg).
The CO derived from protocol A (control) ranged from 21 to 62 L/min for all 8 horses. Interestingly, although the mean CO did not differ significantly among the 4 experimental protocols, the SD of the mean CO for protocol C (16.7 L/min) was significantly (P = 0.028) greater than that for protocol A (9.1 L/min; Table 1). The within-subject variance of CO measurements for protocols C (6.71 L2/min2; P = 0.033) and D (5.67 L2/min2: P = 0.039) was significantly greater than that for protocol A (1.65 L2/min2). Likewise the mean ± SD within-subject width of the range for CO measurements for protocols C (7.8 ± 3.1 L/min; P = 0.022) and D (7.1 ± 3.6 L/min; P = 0.039) was significantly greater than that for protocol A. There was a strong positive correlation between the CO derived from protocol A and the CO derived from each of the other 3 protocols.
Summary of CO measurements derived from 4 experimental thermodilution protocols in healthy anesthetized adult horses.
| Variable | Experimental protocol | |||
|---|---|---|---|---|
| A | B | C | D | |
| Mean ± SD CO (L/min) | 38.4 ± 9.1 | 41.1 ± 13.6 | 45.8 ± 16.7* | 42.4 ± 12.4 |
| Within-subject variance for CO measurements (L2/min2) | 1.65 | 3.76 | 6.71† | 5.67† |
| Mean ± SD width of range for CO measurements within subjects (L/min) | 3.37 ± 1.1 | 4.9 ± 2.3 | 7.8 ± 3.1† | 7.1 ± 3.6† |
| Correlation with protocol A‡ | — | 0.93 | 0.9 | 0.91 |
The thermodilution protocols involved injection of physiologic saline (0.9% NaCl) solution chilled to < 5 °C at volumes of 1 mL/15 kg (protocol A [control]), 1 mL/25 kg (protocol B), and 1 mL/35 kg (protocol C) and injection of physiologic saline solution maintained at 17 °C (approx room temperature) at a volume of 1 mL/15 kg. The CO was measured by each protocol, one after another, during each of 5 measurement cycles, with a 30-min interval between measurement cycles. Each protocol was repeated 3 times during each measurement cycle, and the resulting calculated mean CO was used for analysis purposes. During each measurement cycle, protocol A was always performed first, and protocols B, C, and D were performed in a randomized order.
The SD of this mean differs significantly (P < 0.05) from the SD of the mean for protocol A.
Value differs significantly (P < 0.05) from the corresponding value for protocol A.
Pearson correlation coefficient.
— = Not applicable.
Altering the dobutamine infusion rate during the fourth measurement cycle and discontinuing dobutamine administration and increasing the isoflurane concentration between the fourth and fifth measurement cycles did not significantly change any of the previously described relationships between protocol A and the other 3 protocols (Supplementary Table S1).
Discussion
The present study was conducted to investigate the effect of altering the temperature and volume of the injectate used during thermodilution protocols on the CO measurements derived from those protocols in healthy adult horses. The effects of reducing injectate volume and increasing injectate temperature during thermodilution protocols on the CO of human subjects have been described.10 To our knowledge, the present study was the first to investigate those effects in horses. The use of thermodilution methods for measuring CO in horses is particularly challenging owing the large volume of injectate required for infusion during a prescribed time to generate a diagnostic thermodilution curve.4
Results of the present study did not support our hypothesis that the within-subject variance for CO measurements would be acceptable when the injectate volume was decreased or temperature was increased from the injectate volume (1 mL/15 kg) and temperature (< 5 °C) currently recommended for measurement of CO by thermodilution. The CO measurements derived from each of the 4 thermodilution protocols evaluated in the present study were within the expected range3 for healthy anesthetized horses, and the mean CO did not differ significantly among the 4 protocols. However, the within-subject variance of CO measurements (ie, precision) for protocols C (injectate temperature < 5 °C and injectate volume 1 mL/35 kg) and D (injectate temperature 17 °C and volume 1 mL/15 kg) was significantly greater than that for protocol A (injectate temperature < 5 °C and injectate volume 1 mL/15 kg [control]). This suggested that, when the CO was derived from protocols C and D, the measurement would need to be repeated multiple times to achieve the same accuracy and precision as that for the CO derived from the control protocol. It is possible that the lower volume of injectate administered for protocol C, compared with the injectate volume administered for the control protocol, might have resulted in faster mixing of the injectate with the blood and a smaller change in blood temperature.18 When the thermistor registers only a small change in blood temperature, the signal-to-noise ratio decreases, which increases measurement variability.4 A small temperature change before and after injectate administration resulting in a decrease in the signal-to-noise ratio might also explain the high within-subject variance in CO measurement when a warmer-than-recommended injectate was administered for protocol D. Given that obtaining multiple CO measurements is not efficient or practical in a clinical setting, protocols C and D appear to be unacceptable for clinical use in horses.
Errors in thermodilution-derived CO measurements can occur as a consequence of injectate handling before infusion and by heat transfer while the injectate flows through the catheter.19 Increased injectate temperatures have been associated with overestimation of CO in human subjects.10 It is likely that the same type of errors occur in horses, and that may explain the increased variability and decreased accuracy of CO measurements. Although not assessed in the present study, the use of an insulated syringe or insulated gloves20 or an additional inline thermistor at the injection site could help mitigate this type of heat-transfer error.
Although use of injectate at a warmer temperature or in a smaller volume than generally recommended for thermodilution can lead to highly variable CO measurements, those modifications may be justified in patients with cardiac or renal disease.10 When a modified thermodilution protocol is used, a minimum of 3 CO measurements should be obtained and averaged to account for the increased variability of the measurements.4 Further investigation of the effects of modified thermodilution techniques on CO measurements is necessary to develop standardized correction factors to account for the increase in variability of those measurments.
Limitations of the present study included the fact that the individual who administered the injectate was aware of (ie, was not blinded to) the volume of injectate used and an autoinjector was not used to standardize the rate at which the injectate was infused. Use of an autoinjector would have facilitated administration of the injectate at an uniform rate over a set period of time, thereby decreasing the variability among the triplicate measurements obtained for each protocol during each measurement cycle.21 The thermodilution injectate required for adult horses is allometrically greater than that required for human patients; therefore, use of a calibrated syringe might be beneficial for accurate physical administration of the injectate. The use of calibrated syringes in preference to plastic disposal syringes has been proposed in human studies because calibrated syringes can administer a constant volume of injectate over a set period of time.10, 19
Another limitation of the present study was the size of the Swan-Ganz catheters used for delivery of the injectate. We used 7F 110-cm-long Swan-Ganz catheters in this study. A 7.5F 110-cm-long Swan-Ganz catheter is currently the largest such catheter available, so our options for increasing catheter diameter to facilitate ease of and minimize the time required for injectate infusion were limited. Use of a catheter with a larger diameter should allow for more efficient delivery of the injectate and thereby reduce the potential for mixing errors and result in more accurate CO measurements, especially when an injectate with a higher temperature or smaller volume than currently recommended is used for thermodilution.4
Results of the present study suggested that, for healthy anesthetized adult horses, measurements of CO derived from a thermodilution protocol that used chilled (< 5 °C) injectate at a volume of 1 mL/25 kg of body weight (protocol B) had acceptable accuracy and precision when compared with CO measurements derived from the currently recommended thermodilution protocol (injectate temperature < 5 °C and volume 1 mL/15 kg). However, use of lower volumes of injectate (ie, 1 mL/35 kg; protocol C) is not recommended, nor is use of injectate at near room temperature even at the recommended volume, because it leads to clinically unacceptable variability in the CO measurements. Further investigation is necessary to determine whether the use of lower-than-recommended injectate volumes in thermodilution protocols results in accurate and clinically acceptable CO measurements in horses that are hemodynamically unstable.
Supplementary Materials
Supplementary materials are available online at: avmajournals.avma.org/doi/suppl/10.2460/ajvr.82.10.818.
Acknowledgments
Supported by a USDA National Institutes of Food and Agriculture formula fund grant (No. 1021791).
The authors declare that there were no conflicts of interest.
Abbreviations
| CO | Cardiac output |
Footnotes
Henry Schein, Dublin, Ohio.
Criticath, Merit Medical Systems Inc, South Jordan, Utah.
Akorn Medical, Vernon Hills, Ill.
Normosol, Abbott Laboratories, Chicago, Ill.
Dobutamine, Hospira Inc, Lake Forest, Ill.
Cardiocap/5, Datex-Ohmeda Inc, Madison, Wis.
SAS, version 9.3, SAS Institute Inc, Cary, NC.
Prism, version 7, GraphPad Software Inc, San Diego, Calif.
References
- 2. ↑
Blissitt KJ, Young LE, Jones RS, et al. Measurement of the cardiac output in standing horses by Doppler echocardiography and thermodilution. Equine Vet J 1997;29:18–25.
- 3. ↑
Corley KTT, Donaldson LL, Durando MM, et al. Cardiac output technologies with special reference to the horse. J Vet Intern Med 2003;17:262–272.
- 4. ↑
Muir WW, Skarda RT, Milne DW. Estimation of cardiac output in the horse by thermodilution techniques. Am J Vet Res 1976;37:697–700.
- 6. ↑
Jansen JR. The thermodilution method for the clinical assessment of cardiac output. Intensive Care Med 1995;21:691–697.
- 7. ↑
Ganz W, Donoso R, Marcus HS, et al. A new technique for measurement of cardiac output by thermodilution in man. Am J Cardiol 1971;27:392–396.
- 8. ↑
Swan HJ, Ganz W, Forrester J, et al. Catheterization of the heart in man with use of a flow-directed balloon-tipped catheter. N Engl J Med 1970;283:447–451.
- 9. ↑
Buchbinder N, Ganz W. Hemodynamic monitoring: invasive techniques. Anesthesiology 1976;45:146–155.
- 10. ↑
Nishikawa T, Dohi S. Errors in the measurement of cardiac output by thermodilution. Can J Anaesth 1993;40:142–153.
- 12. ↑
Binkley PF, Murray KD, Watson KM, et al. Dobutamine increases cardiac output of the total artificial heart. Implications for vascular contribution of inotropic agents to augmented ventricular function. Circulation 1991;84:1210–1215.
- 13. ↑
Critchley LA. Bias and precision statistics: should we still adhere to the 30% benchmark for cardiac output monitor validation studies? (lett) Anesthesiology 2011;114:1245.
- 14. ↑
Todd MM. Atrial fibrillation induced by right atrial injection of cold fluids during thermal dilution cardiac output determination: a case report. Anesthesiology 1983;59:253–255.
- 15. ↑
AVMA. AVMA guidelines for the euthanasia of animals: 2020 edition. Available at: www.avma.org/sites/default/files/2020–01/2020-Euthanasia-Final-1–17–20.pdf. Accessed Apr 20, 2021.
- 16. ↑
Runciman WB, Ilsley AH, Roberts JG. Thermodilution cardiac output—a systematic error. Anaesth Intensive Care 1981;9:135–139.
- 17. ↑
Robie NW, Goldberg LI. Comparative systemic and regional effects of dopamine and dobutamine. Am Heart J 1975;90:340–345.
- 18. ↑
Pearl RG, Rosenthal MH, Nielson L, et al. Effect of injectate volume and temperature on thermodilution cardiac output determination. Anesthesiology 1986;64:798–801.
- 19. ↑
Jarvis KA, Woliner MJ, Steffey EP. Accuracy of the thermodilution method in estimating high flow—an in vitro study. J Vet Anaesth 1992;19:41–45.
- 20. ↑
Fegler G. Measurement of cardiac output in anaesthetized animals by a thermodilution method. Q J Exp Physiol Cogn Med Sci 1954;39:153–164.
- 21. ↑
Lépiz ML, Keegan RD, Bayly WM, et al. Comparison of Fick and thermodilution cardiac output determinations in standing horses. Res Vet Sci 2008;85:307–314.
