Comparing response to therapy with lactate levels versus mixed-venous oxygen levels showed a statistically significant difference in mortality between groups.²⁵ Further, lactate may be a superior marker for physicians to follow versus mixed-venous oxygenation. Response to therapy in patients undergoing EGDT showed that normalized mixed-venous oxygen saturation did not always predict an improvement in lactate or mortality.²⁴ Trending lactate as a peripheral marker of sepsis is a valuable and valid tool in identifying and treating this disease process.

Initiating Noninvasive Management

Multiple protocols have adapted Dr. Rivers’ EGDT protocol to a noninvasive approach. The below protocol is adapted from the STOP Sepsis Collaborative (http://emcrit.org/wp-content/uploads/non-invasive.pdf).

For initial resuscitation, all patients should have reliable 18-g IV access at multiple sites, be placed on supplemental O2, and have blood drawn and sent for basic labs, ABG/VBG, and lactate levels in addition to blood cultures. Draw cultures from all indwelling vascular access devices. This resuscitation should not be performed with a single IV or one that is positional. The following also apply:

• Antibiotics should be given as soon as possible. Giving broad-spectrum antibiotics within the first hour carries a clear mortality benefit.^25,26,28
• Source control is paramount.²⁹ Note if the patient has a tense belly, necrotic tissue, decubitus ulcers, signs of CNS infection, indwelling devices, or rales or other pulmonary infectious symptoms.²⁷
• Assess volume status using IVC ultrasound to estimate CVP/RA pressure. If greater than 50% collapse, large-volume resuscitation is needed.
• Administer 20-30 mL/kg crystalloid fluid bolus. Large volumes (over 6 L) of normal saline resuscitation will cause hyperchloremia and a resultant nongap acidosis. Consider lactated ringer’s for subsequent boluses if easily available.
• Be aware of laboratory values for evidence of severe sepsis (Table 2). Consider switching these patients to invasive management if they fail to improve their lactate by 10% in the first 2 hours.²⁴

For management of noninvasive resuscitation, the following apply:

• Perform serial ultrasound assessments of the IVC after each bolus of crystalloid.
• Continue giving 500-mL to 1,000-mL boluses of crystaloid every 20 minutes until the IVC no longer shows significant collapse on ultrasound. There still should be respiratory variation of about 30% collapse (Table 3).
• Recheck MAP: If the patient remains hypotensive (MAP under 65 mm Hg) after adequate fluid loading, switch to invasive strategy in order to begin vasopressor therapy.
• Evaluate response to therapy with lactate levels every 2 hours. Assess mental status and urine output.
• If the lactate has decreased by 10%, continue resuscitation and admit to a non-ICU monitored bed.
• If lactate is rising or has decreased to less than 10% and hemoglobin is less than 7 g/dL, transfuse 1 U of PRBC.
• Transfusion may be considered in patients with hemoglobin of 7-10 g/dL. Also, continue additional 1,000-mL crystalloid volume resuscitation or start peripheral dobutamine to improve cardiac output.
• If the above therapies fail to reduce lactate by the second lactate level measured 4 hours into resuscitation, switch to invasive sepsis management.

Nonresponse to Management

For patients refractory to noninvasive sepsis therapy, place a central line and admit to the ICU.

Summary

Noninvasive management of sepsis is a compelling therapy from a number of perspectives: 1) prevention of central line placement, 2) broad applicability to centers with limited resource to follow central venous monitoring, 3) quicker and individualized approach to resuscitation. Time and well-conducted prospective trials will tell the difference between the invasive and noninvasive approaches. Currently, the data suggest that noninvasive sepsis management is safe and effective in appropriate patient populations.