Critical Care

Cardiac Anatomy

An appreciation of the patient’s cardiac anatomy is paramount. When patients present following cardiac surgery, their cardiac anatomy will have been well defined preoperatively with a combination of echocardiography, cardiac catheterization, and magnetic resonance imaging. However, the fallibility of these investigations is reflected by the occasional conflicts with observations made during surgery. In the setting of an emergency admission or a patient who has been lost to follow-up, the anatomy may be less well defined. It is important to gather data from the patient, next of kin, and other institutions that have cared for the patient previously. Echocardiography in complex CHD may be very difficult and requires clinicians with extensive experience in this setting. Key questions to attempt to answer are the presence of abnormal shunts, the nature of the systemic and subpulmonary ventricles, and previous palliative or corrective procedures that have been undertaken.

Detailed anatomic knowledge helps physicians predict the effects of interventions such as increasing systemic vascular tone and increasing heart rate. It is often necessary to compromise certain physiologic targets (eg, the systemic saturations) to achieve other targets (eg, sufficient cardiac output, systemic pressure). The goal of hemodynamic manipulations is to achieve just enough oxygen delivery to end organs such as the kidney or the brain to prevent organ damage. Frequently, this must be accomplished with parameters that are different from those in patients without CHD. Because patients with CHD tend to be younger and have less atherosclerotic disease, they often tolerate moderate hypotension better.

Fluid Therapy

Fluids can be administered to increase systemic preload to the right side of the heart. Initially, fluid expansion will improve right-sided heart function, particularly if it has restrictive physiology, although this will be at the expense of systemic venous hypertension. However, excessive fluid administration may result in ventricular dilation and a reduction in cardiac output. This is particularly true in patients who have had cardiac surgery and whose hearts are not acutely constrained within a pericardial sac. Fluids should be titrated to markers of cardiac output (direct or indirect, such as clinical examination, metabolic status, urine output). Patients who have a Fontan-type circulation are particularly dependent on adequate venous filling to ensure transpulmonary blood flow in the absence of a subpulmonary ventricle. There are few data to support the use of a specific colloid or crystalloid solution. Synthetic colloids are increasingly less favored. A large, multicenter, high-quality randomized trial of hydroxyethyl starch versus saline in critically ill patients demonstrated a higher incidence of renal replacement in patients receiving starch for fluid resuscitation. In another study, the use of synthetic gelatins was temporally associated with increased renal replacement following cardiac surgery. In both studies, the use of colloids was only associated with a small reduction in the total volume of fluids administered to patients. Human albumin solution did not improve mortality in comparison to saline in a large, high-quality, randomized trial performed in a heterogeneous population of critically ill patients. Therefore, the use of albumin is hard to justify in the context of its higher cost. It is possible that selected populations, such as patients with severe sepsis, may benefit from the use of albumin solutions in fluid resuscitation, but this has not been proven. The role of albumin’s wide range of noncolloid effects in these patient groups is actively investigated.

Blood Transfusion

Anemia is common in critically ill patients. Blood is often administered in an attempt to increase oxygen delivery. Moreover, cyanotic patients have increased red cell mass at baseline. This is one part of their adaptation to chronic hypoxemia and is triggered by increased erythropoietin production in the kidney. The increase in red cell numbers is termed correctly as a secondary erythrocytosis, in contrast to a polycythemia, which relates to increases in more than one hematologic lineage. Patients are frequently iron deficient because of consumption by erythropoiesis, inappropriate phlebotomy, gastrointestinal losses, or poor dietary intake.

The hemoglobin threshold that should trigger transfusion is unclear. A large study in critically ill patients demonstrated that targeting hemoglobin concentrations to 70 to 90 g/L was as effective and perhaps superior to higher targets. This study excluded patients who had undergone cardiac surgery and most likely patients with cyanotic heart disease because the hemoglobin level had to fall below 90 g/L within 48 hours of admission to the intensive care unit. Nevertheless, more restrictive targets limit the exposure of the detrimental effects of transfusion that may increase morbidity and even mortality. Transfused red cells are immunosuppressive; have poorer rheologic properties, which reduce microvascular flow; and are depleted in 2,3-diphosphoglycerate, which impairs their oxygen-carrying capacity for some days. Targets should be customized to the patient. Typically, a threshold of 70 to 90 g/L is used in currently noncyanotic patients who do not have acute coronary ischemia. Cyanotic patients need higher hemoglobin levels, but the exact levels are hard to estimate. Transfusion is better titrated to markers that suggest oxygen delivery is insufficient (eg, low venous saturations or poor end organ function) despite optimization of arterial oxygen saturations and cardiac output.

Inotropes and Vasoconstrictors

Inotropes are used to increase cardiac contractility in low-output states. Catecholamines such as epinephrine, dobutamine, and dopamine are agonists at β-adrenergic and dopaminergic receptors. Although they may increase the force of contraction by increasing intramyocyte calcium levels, this is often at the expense of an increased heart rate, increased myocardial oxygen consumption, and impaired relaxation of the heart during diastole. Epinephrine, per se, can induce hyperlactatemia, which may complicate the interpretation of systemic acid-base balance. There are potential advantages to phosphodiesterase inhibitors, such as milrinone and enoximone, and the newer calcium sensitizer levosimendan, in patients with significant right ventricular dysfunction.

Milrinone, when compared with dobutamine, causes less tachycardia, more pulmonary and systemic vasodilation, more lusitropy, and causes a lesser increase in myocardial oxygen consumption. Because the morphologic right ventricle is so susceptible to afterload changes, the vasodilating properties are advantageous, even in the setting of needing some vasoconstrictor to maintain systemic pressures.

Levosimendan acts by sensitizing cardiac troponin C for calcium during systole. Because intracellular calcium levels are not elevated, there is a lesser increase in myocardial oxygen consumption and better lusitropy. Experimental data suggest it may also be a pulmonary vasodilator. This appears to be borne out clinically and makes it an attractive agent in patients with right ventricular failure and those with pulmonary hypertension. It has been used successfully in pediatric patients with CHD.

Norepinephrine is an α-adrenergic agonist that is a systemic and pulmonary vasoconstrictor. It is administered in vasodilated states to restore systemic vascular resistance (SVR) and mean arterial pressure to ensure adequate organ perfusion. Although autoregulation maintains perfusion of organs during hypotension, there is a threshold at which this fails, and administration of vasoconstrictors will restore organ perfusion and function such as urine output. Vasodilation is common in critically ill patients because of sepsis, systemic inflammation postoperatively, and the administration of vasodilating drugs such as milrinone and levosimendan.

Arginine vasopressin (up to 0.04 IU/hour) is an alternative systemic vasoconstrictor to norepinephrine. It acts at vasopressinergic (V1) receptors and may be vasodilating at low doses in the pulmonary circulation through a nitric oxide–dependent mechanism. This may manifest clinically as a reduction in the PVR/SVR ratio. It has been used successfully in severe sepsis and safely in patients with pulmonary hypertension. These data provide a rationale for selecting arginine vasopressin above norepinephrine in settings of pulmonary hypertension and systemic vasodilation.

Management of the Pulmonary Vascular Resistance

Management of the PVR is often the cornerstone to the care of patients with complex CHD. In patients with a failing subpulmonary ventricle, reduction in the afterload presented by the pulmonary circulation may improve cardiac output; morphologic right ventricles (the usual scenario) are particularly susceptible to failure with acute increases in the PVR. The balance between pulmonary and systemic blood flow in patients with unrestricted shunting is influenced by the balance between PVR and SVR. Thus, in high PVR–low SVR settings, systemic cardiac output will increase at the expense of decreased pulmonary flow, greater venous admixture, and systemic desaturation. The converse will occur in low PVR–high SVR settings.