Published in the October 2008 issue of Today’s Hospitalist
The news this year has been mixed when it comes to evidence-based medicine for the treatment of sepsis.
In January, updated guidelines for managing severe sepsis and septic shock called for a new sense of urgency in diagnosis, initial antibiotic administration and significant resuscitation of patients with sepsis. But the guidance also muddied the waters on a number of other issues, notably when and how to use steroids and recombinant human activated protein C.
“In many ways, our field is less clear than it was four years ago,” says Mitchell Levy, MD, professor of medicine at Brown University School of Medicine and director of the medical ICU at Rhode Island Hospital in Providence. “The changes in the guidelines reflect that.”
Those changes were backed up by an analysis presented by members of the Surviving Sepsis campaign in late September at the European Society of Intensive Care Medicine. Those data, which looked at 14,000 patients whose treatment was guided by the Surviving Sepsis bundles, found that the guidelines’ recommendations do help.
“We have shown that survival increases” when physicians follow the bundles’ recommended actions, Dr. Levy says, particularly the three listed first in the guidelines: initial diagnosis, resuscitation and antibiotics.
Also borne out in the analysis was the benefit of acting much sooner rather than later with sepsis therapies. “We absolutely think that is the critical time when we can really make a difference,” Dr. Levy says.
The six-hour window
According to Isabelle Kopec, MD, vice president of medical affairs for Advanced ICU Care in St. Louis, an “e-ICU” that serves 10 hospitals in five states, the updated guidelines put a clear emphasis on speed.
“The Surviving Sepsis guidelines are a matter of intervening quickly,” she says. “In trauma, you have the golden hour; in stroke, you have the window to treat with thrombolytics; and in acute MI, ‘time is muscle.’ When it comes to sepsis, it’s almost the same kind of principle.”
Physicians now know that “you can give the same interventions over three days vs. six hours,” says Dr. Kopec, who is also chair of critical care at SSM DePaul Health Center in Bridgeton, Mo. “But the outcome is going to be much better if you institute interventions rapidly.”
The guidelines’ emphasis on adequate fluid resuscitation, meanwhile, is not a change from earlier guidelines issued in 2004. Experts say, however, that the point is extremely important “and one where physicians often err.
A common mistake, experts say, is to resuscitate both too slowly and too late. According to the guidelines, the protocol “should be initiated as soon as hypoperfusion is recognized and should not be delayed pending ICU admission.” During the first six hours, a patient should be resuscitated to these goals: central venous pressure of 8-12 mm Hg; mean arterial pressure of ≥65 mm Hg; urine output of ≥0.5 mL/kg/hr; and central venous oxygen saturation of ≥70% or mixed venous oxygen saturation of ≥65%.
“Although initial resuscitation is listed first in the guidelines,” explains R. Phillip Dellinger, MD, the guidelines’ lead author and head of critical care medicine at Cooper University Hospital in Camden, N.J., “the Surviving Sepsis bundles make that point much more strongly.”
The guidelines do contain some new recommendations, including conservative fluid strategy for patients with established acute lung injury who do not have evidence of tissue hypoperfusion. The updated guidelines also suggest that septic patients receiving mechanical ventilation have the heads of their beds elevated 30 to 45 degrees.
No need for ACTH stim testing
Most of the changes in the new guidelines are fairly subtle, experts say. But a few have sparked a great deal of debate and prompted hospitals to revisit their policies and rewrite their order sets.
This summer, for instance, Dr. Kopec found herself distributing summary charts from the guidelines to hospitalists and intensivists. That’s because she realized that some of those physicians were still ordering adrenocorticotrophic hormone (ACTH) stimulation tests before deciding to prescribe steroids for patients in septic shock. Others had been starting patients on dexamethasone and switching them to hydrocortisone if they “qualified for ongoing steroids” based on the results of an ACTH stim test, Dr. Kopec says.
The updated guidelines, however, specifically recommend against using an ACTH stimulation test “to identify the subset of adults with septic shock who should receive hydrocortisone.” Another significant recommendation: Don’t use dexamethasone if hydrocortisone is available.
These revisions are due to study results published since 1994, explains Dr. Dellinger. “If you elect to give steroids, we recommend that the decision not be based on doing an ACTH stimulation test,” he says, adding that the problem with the test is its sensitivity and specificity. Instead, “base your decision solely on how the blood pressure has responded to adequate fluid resuscitation and vasopressors.”
The guidelines also say that very sick patients in septic shock who require vasopressors to maintain blood pressure are the population most likely to benefit from steroids with the least amount of risk.
“If anybody is going to be deficient in cortisol production,” says the guidelines’ co-author Sean Townsend, MD, associate director of the medical ICU at Rhode Island Hospital, “it is that population.”
Steroids: not for the “moderately septic”
The guidelines say that only patients in septic shock whose hypotension responds poorly to adequate fluid resuscitation and vasopressors should receive hydrocortisone.
The guidelines clearly state that more unstable patients should be targeted for steroids, and that the drugs need to be started as soon as possible.
Dr. Townsend notes that recent research stresses the need to add steroids to the treatment regimen sooner rather than later. That means administering the drugs as soon as possible after resuscitation, IV antibiotics and vasopressors are on board but not working adequately.
One unintended consequence of the 2004 guidelines, Dr. Townsend explains, “was probably the over-administration of steroids.” He speculates that hospitalists have gotten used to looking at a “moderately septic” patient, “meaning they are mentating OK, they have a little bit of renal failure and their blood pressure is low but hanging in there,” and wondering if they should give steroids or not.
That’s when physicians used to rely on ACTH stim test results to see if steroids were warranted. “But those patients really should never qualify for steroids because they are not vasopressor-dependant,” Dr. Townsend points out. “We had all sorts of people coming up to the ICU who had been given steroids but who were not on pressors, ever.”
The updated guidelines also stress, however, the risks of being too conservative with steroids. It’s important to have a good protocol in place to screen patients for sepsis and to track deterioration so you can act quickly.
“We say, ‘If someone is not doing well in septic shock, give them steroids,’ ” says Dr. Levy. “We don’t say, ‘Give steroids to them just as they are about to die.’ "
This is not, he warns, “a case of wait-and-see. If at the end of 24 hours, you are looking at a patient who is still sick but stable, that’s a person you should be nervous about. Our data suggest that someone who is not getting better at the end of 24 hours is as bad as getting worse. You need to treat them with steroids. You need to look at their glucose levels. You need to do more.”
Unstable patients and drotrecogin alfa
The guidelines suggest that many unstable patients should also receive drotrecogin alfa (Xigris). That represents a change from the 2004 guidelines, which recommended the therapy outright.
The new guidelines clearly recommend against ordering the drug for patients “with severe sepsis and low risk of death.” Instead, the drug should be reserved for the sickest of patients, those with multiple organ failure and judged clinically to be of “high risk of death,” or who have an APACHE II score of greater than 25.
“Our suggestion is totally in line with FDA recommendations and European regulatory labeling,” Dr. Dellinger says. (That guidance may be upgraded to a recommendation, depending on results from a confirmatory trial that the FDA has asked the drug’s manufacturer, Eli Lilly, to provide, he adds.) Primary concerns are with the drug’s increased bleeding risk.
Through her work with Advanced ICU Care’s e-ICU, Dr. Kopec says she thinks doctors tend to “underuse” drotrecogin alfa in extremely sick septic patients. While some physicians are very uncomfortable using the drug, she points out, her ICU has seen “some dramatic improvements.”
At the same time, her group has also seen “quite a few” patients in whom they’ve had to stop using the drug because of bleeding. “You have to be judicious, and the data do show that if you take too long to make up your mind and you institute it later, the outcome is going to be worse.” Once again, Dr. Kopec points out, timing is everything.
Deborah Gesensway is a freelance health care writer based in Sierra Madre, Calif.
A new way to grade evidence
WHEN UPDATED GUIDELINES were issued earlier this year on managing severe sepsis, perhaps the biggest change was the way the guidelines presented recommendations. The authors created a new grading system to rank recommendations, based both on the quality of the evidence and on accepted medical opinion, even when that opinion is backed with little in the way of evidence.
“One of the criticisms of the previous guidelines was that generally accepted therapy, such as antibiotics, received a lower evidence-based medicine grade because randomized trials had not “and could not “be done,” explains R. Phillip Dellinger, MD, the revised guidelines’ lead author and head of critical care medicine at Cooper University Hospital in Camden, N. J. “The new grading system handles that issue.”
The updated guidelines use a two-part system: All points are given either a 1 (strong recommendation, worded as “We recommend “¦”) or a 2 (weak recommendation, worded as “We suggest “¦”). In addition, each point is assigned an A-D grade summarizing the quality of the evidence behind the recommendation.
An “A” indicates that evidence comes from randomized controlled trials, while a “D” indicates only case studies or expert opinion.
“For example, there is not strong evidence that says that giving antibiotics is a top-notch practice of medicine and that it saves lives compared to not giving antibiotics,” points out co-author Sean Townsend, MD, associate director of the medical ICU at Rhode Island Hospital in Providence. “But nobody on the committee would ever feel comfortable saying that.” The recommendation that physicians should start IV antibiotics even before obtaining appropriate cultures is 1D.
For activated protein C, on the other hand, the new guidelines give it a grade of 2B&C. “We still suggest you give it, but some read the evidence as being in conflict,” Dr. Townsend says. “This is an area where the jury is still out on the drug. Is it useful to give to only the sickest patients?” The problem, he adds, is that “the available studies used different enrollment criteria, so now we’re trying to compare those trials in a post-hoc manner.”