Despite the success of early goal-directed therapy as a protocol, considerable discussion has focused on which of components are effective and which are not necessary for sepsis and septic shock.
The concept of increasing tissue oxygen delivery (the basis for the systemic central venous oxygen saturation goal [ScVO2] of greater than 70%) has not been shown to have any benefit in several studies. (New Engl J Med 1995;333:1025; 1994;330:1717.) Transfusion requirements to keep hematocrit above 30% in patients not meeting the ScVO2 goal have also been dispelled by transfusion trials in critically ill patients, such as the TRICC Trial (New Engl J Med 1999;340:409) and a recent trial involving patients with significant gastrointestinal bleeding. (New Engl J Med 2013;368:11.) A more conservative hemoglobin goal of 7–9 mg/dl in these studies led to either no difference in or in reduced mortality more than the goal of 10–12 mg/dl.
EGDT has been an integral component of the emergency care of patients with severe sepsis or septic shock for more than a decade. (N Engl J Med 2001;345:1368.) Developed as a protocol from the best available evidence at the time, it has been successfully implemented by the vast majority of medical institutions globally. Close to 40 clinical studies done over the past decade have shown a uniform reduction in mortality from about 50 percent to under 30 percent when EGDT is adopted. (Minerva Anestesiol 2012;78:712.) Its components of closely monitoring and optimizing preload via aggressive fluid resuscitation, tissue perfusion via vasoactive medication, and tissue oxygen delivery with inotropic agents and blood transfusion have been adopted by many critical care guidelines, including the Surviving Sepsis Campaign. (Crit Care Med 2013;41:580.)
A work by Jones et al. showed that serial normal lactates or a lactate clearance of more than 10 percent is just as effective as optimizing central venous pressure range and ScVO2. (JAMA 2010;303:739.) Paul Marik's seminal systematic review also largely debunked the utility of CVP and other static measurements in predicting fluid responsiveness. (Chest 2008;134:172.) Instead, many experts recommend dynamic measurements to predict fluid responsiveness, such as inferior vena cava diameter variability, pulse pressure variability, or left ventricular outflow tract velocity variability associated with respiration or passive leg raise maneuver.
A Randomized Trial of Protocol-Based Care for Early Septic Shock
The ProCESS Investigators
N Engl J Med
2014 Mar 18 [Epub ahead of print]
This multicenter randomized clinical trial recruited more than 1500 patients presenting with severe sepsis or septic shock in 29 emergency departments across the country. Patients were given early antibiotics and a 30 ml/kg bolus of intravenous saline and randomized to one of three treatment arms: EGDT; protocol-based care to keep mean arterial pressure above 65 mm Hg with the use of IVF and/or vasoactive medications; or usual care as dictated by the treating physician. Mortality did not differ across treatment arms after 60 days — EGDT: 21%; protocol-based: 18.2%; and usual care: 18.9% (p=0.83).
The rate of central venous catheter usage, as expected, in the EGDT group was much higher (93%) as compared with roughly 55 percent in the other two groups that did not rely on CVP measurement. Additionally, 14 percent of subjects in the EGDT group required pRBC transfusion while only eight percent required transfusion in the other two groups. Interestingly, in the original EGDT trial, more than 60 percent of subjects in the treatment arm received a blood transfusion. The significantly reduced incidence of blood transfusions in the ProCESS trial suggests that core practice changes since the original EGDT study such as aggressive IVF and vasopressor therapy may have reduced the need for blood transfusions in septic patients overall.
The amount of intravenous saline in the ProCESS study did not significantly differ among treatment arms, with the usual care group receiving an average 4.5 L, the protocol-based group receiving 5.0 L, and the EGDT group receiving 5.5 L. This is in contrast to the original EGDT group, in which the average amount of IVF administered for the control arm was 3.5 L, suggesting that patients in the original study were under-resuscitated.
The ProCESS study did not report what interventions if any were made in the EGDT group based on CVP or ScVO2 monitoring, so it is unknown if lower readings in select patients resulted in more aggressive fluid resuscitation and a subsequent effect on outcomes. Detailed analysis of this subgroup and these data may suggest a subset of patients that might still benefit from this more invasive type of monitoring.
The results of the ProCESS study are likely to have a major effect on managing sepsis. Given no mortality difference among the study groups, less consideration should be placed on more invasive interventions like central line insertion, blood transfusion, and the role of dobutamine. Instead, even greater emphasis should be on the core strategies of early recognition of sepsis, aggressive fluid management, and early antibiotics. It is this emphasis on early resuscitation of sepsis patients in the ED that is the legacy of EGDT and for which we have Dr. Rivers to thank.
Access the links in EMN by reading this on our website or in our free iPad app, both available at www.EM-News.com. Comments? Write to us at firstname.lastname@example.org.