Critical Care Group – Portex Unit, Institute of Child Health, University College, Paediatric Intensive Care Unit, Great Ormond Street Hospital NHS Trust, London, United Kingdom
Checking for direct PDF access through Ovid
We hold these truths to be self-evident: that myocardial dysfunction is central to sepsis pathophysiology and, in a septic patient requiring multiple vasoactive infusions, a cardiologist will report the ejection fraction to be “good.”In this issue of the Pediatric Critical Care Medicine, Basu and colleagues (1) demonstrate that estimates of ejection fraction and fractional shortening in 15 children receiving inotropic support after fluid resuscitation for sepsis were indistinguishable from controls.At first sight, this is inconsistent with our understanding that myocardial dysfunction is a critical element of the clinical syndrome of sepsis-induced multiorgan failure, and the key cause of early death (2, 3). After all, randomized trials in septic adults and children support the efficacy of structured, early goal-directed resuscitation algorithms, which target the indirect markers of global perfusion: lactate clearance or superior caval vein oxygen saturations (4–6). These algorithms exploit the Frank-Starling relationships between preload, contractility, and stroke volume to increase cardiac output (CO) (and oxygen delivery) with fluid resuscitation, blood transfusions, and catecholamines. The impressive reductions in mortality seen in these studies would be very difficult to understand if myocardial function truly was normal in sepsis.And, of course, myocardial function is not normal in sepsis. Suffredini et al (7) subjected adult volunteers to intravenous endotoxin and observed declines in ejection fraction and left ventricular stroke work. Ventricular end-diastolic volumes rose (the heart got bigger) and hence stroke volume was maintained despite reduced contractility. An increased heart rate in combination with a maintained stroke volume meant that CO was typically supranormal. Simply put: in sepsis, a bigger heart contracts less forcefully, more often. This dilation should probably be considered a beneficial, and perhaps even a necessary, adaptation for survival (8). Similarly, reduced contractility may represent the cardiomyocytes’ attempts at self-preservation (9). The septic heart also relaxes abnormally slowly – so-called “diastolic dysfunction” (7).The confusion arises from how we choose to describe cardiac function. The missing element in this approach of representing cardiac function by ejection fraction, CO, or similar measures is that they provide no adjustment for the degree of “taxation” – the loading conditions under which the heart is attempting to function. Systemic vascular resistance (SVR) is often reduced in sepsis (7, 10, 11). Indeed, the low afterload state of vasomotor paralysis predominates in adult sepsis, facilitating the increase in CO. Werdan et al (12) write: “severe reduction of afterload seen in septic shock may often mask cardiac impairment, enabling a severely diseased heart to pump a seemingly ‘normal’ CO.” Characterizing cardiac function by CO (or ejection fraction), without knowledge of afterload, is analogous to commenting on the speed of cyclist without knowing if he is going uphill or downhill.Numerous echocardiography measures of myocardial performance have been described (Table 1), with varying degrees of dependence on loading conditions. As yet, there is no consensus on the best descriptor.CO (L·min−1) is inversely related to SVR (dynes·cm−5·sec), and directly related to the difference in mean arterial pressure and central venous pressure:CO = (mean arterial pressure −central venous pressure)/SVRRecently, Werdan et al (12) measured CO and SVR in septic adults via pulmonary artery thermodilution and used nonlinear regression from 187 readings to describe the distribution of CO with SVR without reference to mean arterial pressure or central venous pressure. The inverse power relationship was confirmed, taking the form:CO = 394 × SVR−0.