Letters to the Editor

Abstract

Hemodynamic and perfusion end points for volemic resuscitation in sepsis. Shock 34(Suppl 1):34-39, 2010 To the Editor: We would like to congratulate Drs da Silva Ramos and Pontes Azevedo on their elegant review about end points of volume resuscitation in sepsis (1). Except for lactate, base deficit, and venous oximetry, the authors focused on macrocirculatory variables. We wish to comment on potential caveats connected with the use of macrocirculatory variables as resuscitation end points and propose alternative targets for volume replacement in sepsis. So far, fluid resuscitation has been guided by responses of macrocirculatory variables to fluid loading. Hence, a positive fluid response is commonly defined as an increase of 10% or greater in arterial blood pressure and/or cardiac output. Physiologically, this may hold true, given that the heart acts as a preload-dependent servo pump (2). However, in sepsis, when cardiac dysfunction and disturbed cardiovascular regulation are highly prevalent, this concept may bear potential pitfalls. Although patients with cardiac dysfunction may fail to respond to fluid loading because of impaired cardiac reserve, peripheral hypovolemia and tissue hypoperfusion may persist, requiring further fluid loading. On the other hand, a patient with intact heart function may respond to fluid loading even if tissue perfusion has been restored. Given that tissue hypoperfusion is one of the key pathogenetic features of organ dysfunction sepsis and septic shock (3), hemodynamic interventions-fluid resuscitation included-should target tissue perfusion. Doing so, macrocirculatory parameters, such as those discussed by the authors, exhibit a notoriously weak association with tissue perfusion in critically ill patients (4). Accordingly, a clinical study reported that changes in MAP and cardiac index failed to predict microcirculatory changes in response to early fluid resuscitation in severe sepsis. The predictive value of both variables to foresee improvements in microcirculatory flow was only marginally better than flipping a coin (area under the receiver operating characteristic curve for MAP and cardiac index: 0.51 and 0.62, respectively). Interestingly, only changes in lactate levels were related to microcirculatory but not macrocirculatory changes (5). If tissue hypoperfusion is the problem, why not consider tissue perfusion as the end point for fluid resuscitation? As summarized in a recent review (6), several (experimental) devices may help the clinician to quantify tissue perfusion. From a clinical point of view, however, assessment of skin perfusion (e.g., by examination of skin temperature and capillary refill time) may be used as a sensitive, easy, cheap, and readily available means to evaluate peripheral volemia and tissue perfusion and guide volume loading at the bedside. Considering that the skin hosts the highest density of vasoconstrictor hormone receptors (7), it is the first of all organs where vasoconstriction occurs and peripheral hypovolemia can be noted. With few exceptions (e.g., serious hypothermia/hyperthermia, arterial occlusive disease), cold and mottled skin reflects relevant vasoconstriction and can serve as an early indicator of reduced systemic blood flow due to hypovolemia or reduced cardiac pump function. If fluid resuscitation with or without inotropic support has reestablished peripheral perfusion and signs of tissue hypoperfusion persist (e.g., oliguria or elevated lactate) (8), other hemodynamic interventions, such as vasopressor infusion, are indicated. Mats Dahlqvist Walter R. Hasibeder Martin W. Dünser