The Right Antibiotic Choice for Sepsis : Emergency Medicine News

Journal Logo

ID Rounds

The Right Antibiotic Choice for Sepsis

Isaacs, Lawrence MD

Author Information
Emergency Medicine News: May 2007 - Volume 29 - Issue 5 - p 14
doi: 10.1097/01.EEM.0000271268.52777.05

    After focusing last month on the pathophysiology of sepsis and early goal-directed therapy (EGDT), this month I think it's worth considering traditional methods (antibiotics) and looking at the implementation of EGDT in a community ED. A very aggressive approach is required for a patient who is anywhere in the spectrum of sepsis, from SIRS to septic shock. Managing a septic patient obviously includes antibiotic administration, and EGDT assumes proper and timely antibiotic administration.

    Antibiotic Choices in Sepsis

    Traditional teaching about sepsis has been just to give the patient any antibiotic, but recent studies have shown a mortality benefit to proper antibiotic choices. (Crit Care Med 2003;31:2742; Clin Infect Dis 2004;28:284.) We all know that sometimes the source of infection is obvious, but sometimes it is not. The PROWESS study, which investigated activated protein-C, found the most common source was pneumonia (54%), followed by the abdomen (20%), “other” (16%), and the GU tract (10%).

    The table gives logical antibiotic choices according to the presumed source. The goal is administration of a reasonable antibiotic within one hour of recognition of sepsis. Lactate, a byproduct of anaerobic metabolism, is used as a marker for tissue hypoperfusion and hypoxia. A single lactate level greater than 4 mmol/L on presentation is associated with increased mortality. (Ann Emer Med 2005;45:524.) Lactate levels also can be trended to measure the adequacy of the resuscitation. As a result, obtaining lactate levels is an integral part of following an EGDT protocol.

    The current recommendations for steroids, which have been studied for decades in septic patients, state that patients whose blood pressure is not responsive to an appropriate fluid challenge should receive what some term “physiologic-dosed steroids.” In patients with severe sepsis or septic shock, there is adrenal failure, resulting in decreased synthesis of glucocorticoids. Unlike early studies using high-dose steroids, which resulted in higher mortality (Crit Care Med 1995;23:1430), more recent trials have shown improved outcomes with either a single dose of hydrocortisone (improved MAP) or a long course (>5 days) of low-dose corticosteroids (improvement of hemodynamics resulting in shorter time to removal of vasopressors). (Surg Clin N Am 2006;86[6]:1523.)

    The theory behind the improved outcomes is that there is an increase in peripheral vascular tone with steroid use. The recommendation from the Surviving Sepsis Campaign, a multidisciplinary consensus of guidelines resulting from EGDT (see for the full text) is hydrocortisone 200–300 mg/day divided TID-QID for seven days. A recent review and meta-analysis found that using this regimen reduced 28-day ICU and in-hospital mortality without any increase in adverse events. (BMJ 2004;329:480).

    Steroids for Sepsis

    A more recent, international randomized, placebo-controlled study called CORTICUS (Corticosteroid Therapy of Septic Shock) compared 28-day mortality between hydrocortisone-treated and placebo-controlled patients. Mortality was 33.5 percent in treated and 31 percent in placebo patients, which is considered statistically insignificant. Time to shock reversal was shorter in treated patients, but the difference in rate of reversal was insignificant. It is difficult to say how this will affect the current surviving sepsis guidelines, but it certainly will not add strength to the steroid protocol.

    Activated protein-C, also called drotrecogin alpha (Xigris), was approved as a result of the PROWESS study, a multicenter, randomized double-blinded, placebo-controlled trial whose endpoint was 28-day mortality. The absolute mortality reduction was 6.1 percent (30.8% vs. 24.7%), with the number needed to treat set at 16. In general, the sicker patients — those with higher APACHE II scores (>24), over age 50, with shock, and with more than one organ system dysfunction — showed the greatest benefit.

    There are two concerns about this therapy. The first is bleeding, which occurred in two percent of placebo patients and 3.5 percent of patients treated with activated protein-C. The drug should not be used in patients with thrombocytopenia (<30,000 platelets), prolonged PT/INR, or at high risk of bleeding. This drug has a very short half-life, and should be discontinued one hour before and not restarted for one hour after any percutaneous procedure. For major surgical procedures, it should be discontinued one hour before and 12 hours after. The other concern surrounding this drug is its cost, which is approximately $1700 a day. Although this aspect of the drug is not completely without controversy, it seems that the consensus so far leans toward this drug's cost-effectiveness. (Health Technol Assess 2005;9[11]:1; Am J Med Sci 2004;328:205.)

    All this seems promising, especially in academic centers where there is frequently an abundance of willing and able help with procedures, an atmosphere of learning, and to some degree, less pressure to move patients through as compared with community EDs. Following an EGDT protocol certainly is time-consuming on the physicians and nurses, and there are important additions to the patient's care that need to be in place to follow the guidelines. Many in community hospitals are skeptical about the ability to implement a sepsis protocol based on EGDT and the Surviving Sepsis Campaign. Although there may be resistance at first, it is possible, just like anything worth doing.

    Establishing a sepsis protocol is a good first step. To do this, there needs to be collaboration between the emergency physicians, emergency nurses, and the ICU/intensivists. Sample protocols are readily available, and one can be found at A component of all protocols is getting lactate and random cortisol levels as part of the initial blood draw; lactate levels are usually a different vial from the usual rainbow of vials. The nurses or techs will need to be educated on obtaining those vials as part of a sepsis package of bloods.

    The next hurdle is the physician-dependent procedure, the intubation, and the sometimes more problematic central venous access. Remember, to measure the CVP and Scv02 accurately, the central line needs to be above the diaphragm. Depending on the patient's body habitus and physician's skill, this can be challenging. Having an ED ultrasound and collaboration with the intensives and surgeons can be helpful here. The central lines also need to have the ability to measure Scv02. One catheter, “PreSep” (Edwards Lifesciences), is available and designed for this purpose, although your institution will need to purchase the monitor to get the readings. The nurses will need to learn how to calibrate the catheter and use the monitor (which is not difficult).

    Ideally, much of this monitoring will take place in the ICU, but in these days of ED boarding, hours of ICU care often take place in the ED. One should look at this like the early days of thrombolytics when specific protocols where in place and collaboration between the ED and cardiology became more important. Looking back, that intensely structured care plan seems a little antiquated, but it does show how far we've come in that arena. Maybe 15 years from now we'll be saying the same about sepsis.

    Do You Want to Continue Receiving EMN?

    If this issue contains a business reply card on the front cover, we still need to hear from you! Please return the card today to ensure uninterrupted delivery.

    © 2007 Lippincott Williams & Wilkins, Inc.