Jay L. Falk, MD, FCCM
Orlando Regional Healthcare
University of Florida College of Medicine
Orlando, Florida, USA

Early Goal-Directed Therapy in the Emergency Department

References

Recognizing the Signs
Early recognition of patients with severe sepsis requires a high index of suspicion and an understanding of sys¬temic perfusion. Preferential perfusion of the skin in septic patients has led to its description as “warm shock.” Classic signs of shock may be lacking in septic patients. The presence of systemic hypoperfusion, as evidenced by the presence of lactic acidosis in patients with normal blood pressure, has been recognized for decades. In a 1988 paper reviewing fluid resuscitation in shock, Falk et al. stated, “In septic shock, the skin remains warm and the blood pressure near normal despite the presence of tissue hypoxia. Thus, hypotension and other classic clinical signs are insensitive markers of systemic hypoperfusion. More sensitive markers of acute perfusion failure are needed if circulatory shock is to be recognized in its early stages when resuscitation is most likely to be successful.”4

Groundbreaking work by Weil and Afifi introduced the concept of using lactate as a metabolic marker of tissue hypoxia and as a prognostic indi¬cator of the severity of shock and the likelihood of survival.5 As lactate concentration increased from 2.0 to 8.0 mmol/L, the estimated probability of survival decreased from 90% to 10%. Gradually, lactate became a widely accepted meta¬bolic marker of tissue ischemia.6

The advent of commercially available lactate electrodes that enabled rapid, accurate point-of-care testing has made intensive care unit (ICU) and ED use more practical. Studies substantiating the equivalency of peripheral venous and arterial lactate in an ED population remove another potential barrier to lactate use, namely the need for arterial puncture.7 Other studies also have indicated the utility of following serial lactate levels as resuscitation proceeds.8,9 Patients in whom resuscitation has restored perfusion progressively clear their lactate over time, while patients with ongoing tissue ischemia fail to clear lactate and have a higher mortality rate. Recently, emergency medicine investigators substantiated that lactate as a prognostic indicator in septic patients applies to ED patients.10

Resuscitating the Patient
Once a hypoperfused septic patient is recognized, resuscitation must be initiated immediately. No one has ever advocated “delayed resuscitation.” Rather, the lack of recogniition prevented “early” resuscitation. Current knowledge and technology now should allow EPs to accomplish this, leading to a focus on goal-directed therapy. If traditional vital signs are insensitive markers of systemic perfusion, why should they be the target endpoints of resuscitative efforts? Clearly, they should not be.

This also is not a new concept. In 1985, Falk et al. reported that when patients were fully resuscitated to hemodynamic endpoints, those who cleared their lactic acidosis survived, while those who failed to do so died.8 Lactate clearance in survivors was substantiated as a concept by several investigators, leading some to explore using resuscitative endpoints based on systemic oxygen delivery and consumption rather than blood pressure and pulse.9,11-13 In a prospective, randomized clinical trial, Shoemaker et al. introduced the concept of resuscitating to “supranormal” physiologic endpoints based on oxygen delivery.14 They demonstrated a lower mortality rate in surgical patients resuscitated to predetermined supranormal oxygen transport values. Tuchschmidt et al. demonstrated similar findings in septic medical patients.15 These studies used flawed methods and were criticized heavily. A series of subsequent studies failed to prove the value of pushing oxygen delivery beyond normal values and also were flawed.16,17 The practice of resuscitating to supranormal endpoints fell out of favor. At the same time, the notion that the “goal” of resuscitation must be the elimi¬nation of tissue hypoxia became firmly entrenched among intensivists.18

This notion set the stage for a study by Rivers et al., in which the authors identified hypoperfused septic patients utilizing clinical criteria, traditional vital signs and elevated lactate levels (>4 mmol/dL). They prospectively resuscitated them either to traditional hemodynamic endpoints (blood pressure, central venous pressure, urine output) or to an endpoint based on optimizing oxygen delivery, namely normalizing central venous oxygen saturation (ScvO2 >70%).1 The protocol for this goal-directed group utilized a simple algorithm of crystalloid fluid resuscitation to a central venous pressure (CVP) of 8 to 12 mm Hg followed by strategies to maintain mean arterial pressure (MAP) >65 mm Hg (vaso¬pressors) and oxygen delivery sufficient to achieve a ScvO2 of 70% (packed red blood cell transfusion to increase blood oxygen carrying capacity if hematocrit <30%, dobutamine infusion to increase cardiac output if all other conditions were met and ScvO2 <70%).

The standard therapy group (control) was resuscitated with a CVP line in place and received relatively aggressive resuscitation compared to that seen in a community ED set¬ting. Despite this, the early goal-directed therapy (EGDT) group enjoyed a dramatic absolute reduction in mortality of 16%. Remarkably, in the EGDT group, the incidence of sudden cardiovascular collapse was cut in half (10.3% vs 21.0%, p=0.02).

This study generated some criticism, as do all clinical trials. The patients were not blinded in the ED phase of the study, which could have introduced some bias. The mortality rate in the standard therapy group was higher than that seen in ICU-based clinical trials. Concerns were raised about the use of blood and its ability to transport oxygen in the 48 hours after transfusion and its immunosuppressive potential as well as about the need for “unproven, expensive technology” such as ScvO2 catheters.19-21 Concerns notwithstanding, the Surviving Sepsis Campaign adopted this protocol based on the firm physiologic basis of the study.22 The American College of Emergency Physicians participated in the SSC, and an article in the Annals of Emergency Medicine attempted to create enthusiasm for utilizing this approach in emergency medicine practice.22

Most recently, Trzeciak et al. demonstrated the feasibil¬ity of utilizing the EGDT protocol in clinical practice.24 They implemented the protocol as a collaborative emergen¬cy medicine/critical care quality-improvement initiative, resulting in 91% of patients achieving resuscitative end¬points within the first six hours. Total resource utilization was not different between the EGDT patients and historical controls although, as in the Riverʼs study, use was front-loaded into the ED compared to the ICU. Interestingly, pulmonary artery catheter use was reduced markedly (43% in controls vs 9.1% EGDT group, p=0.001). Although the EGDT group was compared to historical controls, there appeared to be a substantial reduction in mortality, from 43.8% to 18.2%. The authors noted that no additional physician staffing, nursing or other support personnel was required to implement the protocol. Also, other than mak¬ing the ScvO2-CVP catheter available, no additional equip¬ment was required.

So Why the Resistance to EGDT?
Despite these impressive data, a recent survey indicated that EGDT is not being implemented in most EDs through¬out the country.2 Jones and Kline surveyed 30 academic tertiary care hospital EDs to find that standard use of EGDT occurred in only two facilities, or 7% of the sample. Presumably, the percentage of community hospital EDs using the protocol is much lower.

Surveyed departments indicated several barriers, including:
• Too much patient time in the ED (21%)
• Too many ED resources required (43%)
• Too much EP time required (29%)
• Specialty monitoring equipment required (75%)
• Central venous cannulation required (36%)

While these issues present valid points, many other new protocols and treatments have been adopted to pre¬vent sepsis. ST-elevation myocardial infarction alerts are commonplace, and EPs routinely administer thrombolytic agents when immediate access to invasive catheterization laboratories is lacking. EPs have adopted standardized approaches to ongoing ischemic chest pain and routinely administer aspirin, β-blockers, nitroglycerin and anticoagu¬lants. Trauma resuscitation protocols are embraced. Airway management protocols involving difficult airway manage¬ment, rescue devices and failed airway surgical options have been implemented successfully.

Why are EDs not embracing the EGDT protocol simi¬larly? The potential reasons are varied. On the surface, fail¬ure to resuscitate critically ill septic patients fully in the ED does not appear to result in adverse consequences. While the trauma patient, the acute myocardial infarction (AMI) patient or the patient with a compromised airway may experience immediate deterioration or death, the septic patient most often manifests the lack of adequate resuscitation days or even weeks after the ED visit.25 The emergency team doesn’t recognize its contribution to the patient’s demise on a clinical or personal level.
The ED is a hectic place; EPs must prioritize their atten¬tion. Systemic hypoperfusion may seem like a less-pressing issue that can be put aside for the moment. It is easy to think a sick septic older person with reasonable vital signs has been taken care of properly if the site of infection has been identified, antibiotics have been ordered and an admitting physician has been notified. However, the potential hospital mortality for these patients exceeds the hospital mortality of AMI by three to four times.

Putting EGDT Into Action
To ensure widespread implementation of EGDT, EPs must take ownership in the process. EPs ultimately are judged on their success or failure in reducing morbidity and mortality in their most critically ill patients. It has been argued that the reason for the specialty is predicated on the need for a physician to be present constantly to recognize the critically ill and resuscitate them.27 Current concepts of “pay for performance” being floated by government and commercial payors undoubtedly will put more pressure on hospital administrations to force compliance with published guidelines such as those published by the Surviving Sepsis Campaign.

Given national overcrowding of EDs, much focus has been placed on patient throughput, but many patient satisfaction surveys receive feedback only from discharged patients. Hospital administrators must support initiatives to make care of the septic patient a priority despite pressures to meet other competing demands. In academic centers with in-house critical care services, simple communication with a combined commitment from both services is the formula for success.23 Community hospitals with no in-house critical care presence need systems that allow for early consultation with out-of-house consultants. EPs must take responsibility for early recognition and the initiation of goal-directed resuscitation. Intensivists must be willing to respond quickly to take over care. Realization of this collaborative model is long overdue.