"Is The Patient Half Full or Half Empty?"

The assessment of adequate intravascular volume in critically ill patients and patients undergoing anaesthesia is vital in ensuring an adequate circulation. We are all fully aware of the consequences of hypovolaemia and hypervolaemia, both of which are associated with adverse outcomes.^1-3 Hypovolaemia is present in 50% of human intensive care patients and is often occult and difficult to detect.⁴ The haemodynamic measurements of filling pressures, urine output and biochemical indicators are misleading and poor indicators of central blood volume.⁴ Some studies have found no correlation between mean arterial blood pressure and heart rate, systemic vascular resistance and cardiac index,⁵ while others have found a weak correlation.⁶ Central venous pressure has been traditionally used as an indicator of fluid load and is a better indicator of fluid load than blood pressure.⁷ Central venous pressure should always be assessed in relation to a fluid challenge, as a single reading is misleading.⁴ Central venous pressure is a poor predictor of fluid responsiveness (ROC 0.61).⁸ Radio-labelled markers (red cells and albumin) and dyes (indocyanine green) indicate the fluid volume in circulation, but they give no idea as to adequacy of fluid volume in relation to intravascular space.⁴

Fluid responsiveness is defined as a positive increase in stroke volume in response to a fluid bolus. A human meta-analysis showed that ventricular pre-load indicators were poor indicators while dynamic parameters were better indicators of fluid responsiveness.⁹ Not all patients will respond to a bolus of fluids and increase in blood pressure and cardiac output will not be seen.¹⁰ Fluid responsive patients are on the preload-dependent portion of the Frank-Starling curve, while nonresponsive patients are on the preload load-independent portion of the curve.¹⁰ Nonresponsive patients potentially benefit more from ionotropic circulatory support.

Frank-Starling's law states that the heart will pump, within physiological limits, the quantity of blood delivered to the right atrium without significant back pressure. This law relates to the delivery of blood (blood flow, ml/kg/min) than to pressure. Simply, this means that the rate of return of blood to the heart determines cardiac output.

Poiseuille's equation tells us that flow is directly proportional to the pressure difference and the radius to the fourth power, and inversely proportional to the length of the tube and the viscosity. A small change in vessel diameter results in a large change in blood flow. Changes in vessel diameter also produce changes in circulating volume space. Changes in arterial diameter produces little change in circulating volume space but do affect left heart cardiac output and hence venous return to the right heart, while changes in venous capacitance has dramatic effects on circulating volume space. These changes introduce the concept of an absolute and relative hypovolaemia. An absolute hypovolaemia is the result of loss of fluid from the circulating space (e.g., haemorrhage or dehydration), while a relative hypovolaemia is the result of changes in venous capacitance (e.g., vasodilatation).

The cardiovascular and respiratory systems are integrated by the fact that they share a common chamber, the chest. A natural increase and decrease in cardiac output is seen during normal spontaneous ventilation;¹¹ during positive pressure ventilation, this relationship changes.¹¹ During the early phase of inspiration, the positive pressure in the alveolus moves blood from the pulmonary circulation to the heart, resulting in an increase in left ventricular preload, a rise in stroke volume and blood pressure.¹² As inspiration continues, the pulmonary vascular bed is depleted of blood and the pulmonary vascular resistance rises, a decrease in left ventricular preload occurs, and stroke volume and blood pressure decrease.¹² The decrease in stroke volume is particularly obvious if the pulmonary vascular bed is hypovolaemic.¹² During expiration, the changes reverse and pulmonary blood flow and hence left ventricular preload and stroke volume return to baseline¹² (Figure 1). These changes in the arterial pulse pressure curve during the ventilatory cycle have been used to assess intravascular volume and fluid responsiveness.

Figure 1. Effects of ventilation of left ventricular preload

A number of parameters from the arterial pulse pressure curve have been defined. These parameters are systolic pressure variance, delta up and delta down and pulse pressure variance.

Systolic pressure variance has been used to assess fluid responsiveness.¹⁰ Systolic pressure variance is defined as the difference between maximum and minimum arterial pressure over 1 ventilatory cycle¹⁰ (Figure 2. Systolic pressure variance). A decrease in peak systolic blood pressure of more than 5 mm Hg during expiration from the end expiratory peak systolic blood pressure indicates fluid responsiveness.¹⁰

Figure 2. Systolic pressure variance

Systolic pressure variance is composed of two components, δ up and δ down. Delta up (δ up) and delta down (δ down) are defined as the change in arterial pressure from baseline.¹⁰ Baseline is defined as the systolic pressure measured during an end expiratory pause.¹⁰ δ up is therefore the maximum systolic pressure above baseline while δ down is the minimum systolic pressure below baseline during a ventilatory cycle¹⁰ (Figure 3. δ up and δ down). δ up shows the greatest change in hypervolaemia. δ down shows the greatest change in hypovolaemia.¹⁰

Figure 3. δ up and δ down

The changes in pleural pressure can result in changes in systolic pressure with the result that systolic pressure variance can result when no changes in stroke volume occur¹⁰ (Figure 4. Pulse pressure variance as a result of changes in airway pressure. The pulse pressure remains constant throughout the ventilatory cycle.). To overcome this potential error, pulse pressure variance has been defined. Pulse pressure is defined as the difference between systolic and diastolic pressure and is directly related to left ventricular stroke volume and inversely related to vascular resistance.¹⁰ Pulse pressure is unaffected by the pleural pressure, as the systolic and diastolic pressure are equally increased by the pleural pressure. Pulse pressure variance is calculated as follows: PP = 100 * ([PP_max – PP_min]/[(PP_max + PP_min)/2]) where PP_max and PP_min are the maximal and minimum pulse pressures over a single respiratory cycle¹⁰ (Figure 5. Pulse pressure variance. PP_min is minimum pulse pressure variance and PP_max is maximum pulse pressure variance.). Pulse pressure variance with a variance greater than 13% indicated fluid responsiveness with sensitivity of 94% and a specificity of 96% in human patients with acute circulatory failure.¹⁰ Pulse pressure variance provides superior results to pulmonary occlusion pressure and central venous pressure.⁴ Pulse pressure variance and left ventricular stroke volume variance have shown to be predictive of fluid load in open chest conditions.¹³

Figure 4. Pulse pressure variance as a result of changes in airway pressure. The pulse pressure remains constant throughout the ventilatory cycle

Figure 5. Pulse pressure variance PPmin is minimum pulse pressure variance and PPmax is maximum pulse pressure variance.

A new technique based on changes in systolic pressure associated with different tidal volumes has been shown to be a good predictor of fluid responsiveness.⁸ This technique has been termed the respiratory systolic variance test.⁸ Three different pressure limited breaths were delivered at 10, 20 and 30 cmH₂O.⁸ This value is then normalised to the respiratory pressure delivered. In fluid responsive patients, a decrease in systolic pressure is seen with successive breath⁸ (Figure 6. Respiratory systolic pressure variance.).

Figure 6. Respiratory systolic pressure variance

Most studies to date have only validated these techniques in closed chested patients.¹³ Recently, work has been undertaken in man during thoracotomy on stroke volume variation and pulse pressure variation.¹³ Under these conditions, these parameters were found to be reliable indicators of fluid responsiveness (r = 0.74).¹³ Central venous pressure (r = 0.20) was a poor predictor.¹³ The following has been proposed as cut of values for fluid responsive human patients.⁸ These values would need to be validated in veterinary patients.

Pulse pressure changes have been reported in a horse associated with postanaesthetic myopathy.¹⁴ Unfortunately, there is not much veterinary clinical work on pulse pressures; there are numerous experimental studies in dogs validating the technique. In man, these techniques are reliable and have been shown to have good sensitivity and specificity. Pulse pressure variance (ROC 0.95, sensitivity 86%, specificity 89%), stroke volume variation (ROC 0.87, sensitivity 81%, specificity 82%) and respiratory systolic variation (ROC 0.96, sensitivity 93%, specificity 89%) are good predictors of fluid responsiveness.⁸ These techniques need to be validated in a veterinary clinical setting. Evidence in man already shows that maintaining adequate perfusion and cardiac output can affect outcome. Fluids administered to man undergoing major bowel surgery adjusted to stroke volume reduced hospital length of stay and complications.¹⁵ Fluids were administered in boluses until CVP rose by more than 3 mm Hg and stroke volume did not increase by 10%.¹⁵ We need to consider these techniques as valuable adjuncts in the management of our patients.

References

1.  Lobetti RG, Jacobson LS. Renal involvement in dogs with babesiosis. J S Afr Vet Assoc. 2001;72(1):23–28.

2.  Matejovic M, Krouzecky A, Rokyta R, Novak I. Fluid challenge in patients at risk for fluid loading induced pulmonary edema. Acta Anaesthesiol Scand. 2004;48:69–73.

3.  Lobetti RG. Cardiac involvement in canine babesiosis. J S Afr Vet Assoc. 2005;76(1):4–8.

4.  James MFM. Relative hypovolaemia. South Afr J Anaesth Analg. 2005;11(2):68–72.

5.  Haskins SC, Pascoe PJ, Ilkiw JE, Fudge M, Hopper K, Aldrich J. The effect of moderate hypovolemia on cardiopulmonary function in dogs. J Vet Emerg Crit Care. 2005;15(2):100–109.

6.  Giguére S, Knowles HA, Valverde A, Bucki E, Young L. Accuracy of indirect measurement of blood pressure in neonatal foals. J Vet Intern Med. 2005;19:571–576.

7.  Boag AK, Hughes D. Assessment and treatment of perfusion abnormalities in the emergency patient. Vet Clin North Am Small Anim Pract. 2005;35(2):319–342.

8.  Preisman S, Kogan S, Berkenstadt H, Perel A. Predicting fluid responsiveness in patients undergoing cardiac surgery: functional haemodynamic parameters including the respiratory systolic pressure variation test and static preload indicators. Brit J Anaesth. 2005;95(6):746–755.

9.  Michard F, Teboul J-L. Predicting fluid responsiveness in ICU patients: a critical analysis of the evidence. Chest. 2002;121:2000–2008.

10. Michard F, Teboul J-L. Using heart-lung interactions to assess fluid responsiveness during mechanical ventilation. Crit Care. 2000;4:282–289.

11. Steingrub JS, Tidswell M, Higgins TL. Hemodynamic consequences of heart-lung interactions. J Intensive Care Med. 2003;18:92–99.

12. Joubert IA. Assessing intravascular fluid status. South Afr J Crit Care. 2004;20(1):14–18.

13. Reuter DA, Goepfert MSG, Goresch T, Schmoeckel M, Kilger E, Goets AE. Assessing fluid responsiveness during open chest conditions. Brit J Anaesth. 2005;94(3):318–323.

14. Duke T, Filzek U, Read MR, Read EK, Ferguson JG. Clinical observations surrounding an increase incidence of postanesthetic myopathy in halothane-anesthetized horses. Vet Anaesth Analg. 2006;33:122–127.

15. Wakeling HG, McFall MR, Jenkins CS, Woods WGA, Miles WFA, Barclay GR, et al. Intraoperative oesophageal Doppler guided fluid management shortens postoperative hospital stay after major bowel surgery. Brit J Anaesth. 2005;95(5):634–642.