Published in the August 2005 issue of Today’s Hospitalist
Talk to an intensivist about treating sepsis, and you’re likely to hear an old adage with a new twist: Time is tissue. In taking a page from their colleagues in cardiology, the critical care community is trying to impart the same sense of urgency that’s used in treating myocardial infarction to sepsis, a condition that strikes 750,000 patients in North America every year. Mounting evidence shows that when it comes to treating sepsis, what physicians do “or don’t do “in the early hours after diagnosis can have long-lasting effects in terms of organ function and mortality.
For more on treating sepsis, read our December 2016 article, Meeting the new sepsis measure.
The statistics on sepsis are daunting. Sepsis without any organ dysfunction raises a patient’s risk of death to about five percent. Severe sepsis, a condition that is probably not as severe as you think, increases your patient’s mortality risk to nearly 20 percent. Once that patient progresses to septic shock, the risk of dying jumps to 46 percent.
But because septic patients can initially appear to be only moderately ill “they might have a slight fever and a slightly elevated white count but a relatively normal blood pressure, for example “a diagnosis is often delayed. Making the diagnosis even more unclear “and delaying it further “is the fact that there is no gold-standard lab test or assay to detect sepsis.
The good news is that there are some strategies that can help you detect sepsis and treat these patients before they progress to severe sepsis or septic shock. But the key, experts say, is to get to those patients in the “golden hours” to minimize any long-term organ dysfunction.
Heightened clinical suspicion
One of the challenges in recognizing sepsis, particularly the more subtle forms of the condition, is that patients typically display signs and symptoms that can be attributed to other common conditions. That’s why experts say that to detect sepsis in its early stages, physicians need a heightened clinical suspicion
“When sepsis is just starting, it can be very subtle,” explains William Janssen, MD, instructor of medicine at the University of Colorado. “You go from a patient who looks relatively normal to someone who is very sick, and the changes in that person might be very subtle. It might start with a fever and tachycardia, which hospitalists see all the time.”
“One of the biggest problems we face on the floor and in the ED is the idea that patients have just a little pneumonia with a little hypotension,” says Terry P. Clemmer, MD, professor of medicine at the University of Utah School of Medicine and director of critical care medicine at LDS hospital. “Physicians don’t realize that this meets the criteria for severe sepsis. They delay getting everything started because they don’t put the right label on the patient.”
While it’s true that many sepsis patients are sent from the ED to the ICU, that’s only part of the picture. Emmanuel Rivers, MD, lead author on a groundbreaking study published in 2001 in the New England Journal of Medicine on the use of early goal-directed therapy in sepsis, says that up to 15 percent of patients admitted to the floors will develop severe sepsis. But because so many physicians miss the telltale signs of the condition, the patients slip through the cracks and end up in the wards.
“If your creatinine goes up and you have a need for oxygen but no pneumonia, and you have a source of infection, then you have severe sepsis,” explains Dr. Rivers, who is associate professor of emergency medicine and surgery at Henry Ford Hospital and Wayne State University in Detroit. “But many times it’s very unappreciated.”
Dr. Clemmer, who helped develop a sepsis bundle produced jointly by the Surviving Sepsis Campaign and the Institute for Healthcare Improvement, says that physicians need a little help to make sure sepsis is caught every time. At LDS Hospital, nurses use a simple checklist to pick up on signs of trouble that may indicate sepsis. When they encounter a red flag, they can order lab tests through the hospital’s system of standing orders. Delays are avoided, and septic patients receive care much more quickly.
“You go down the checklist and ask whether this is just fever or whether it’s falling in the category of severe sepsis,” Dr. Clemmer says. “We need to know if there is an organ dysfunction, so I need the metabolic panel and the lactate level. I need these things that tell me this is more than a little fever and a little dehydration in response to fluid, that this is actually moving into organ failure, which defines severe sepsis.”
Serum lactate levels
While detecting sepsis requires you to recognize a collection of symptoms that by themselves may not point to the condition, testing patients’ serum lactate levels can provide a good, if imperfect, picture about which patients need extra attention.
Dr. Rivers notes that patients who are admitted to the hospital with a lactate level of 4 or higher face an in-hospital mortality of 40 to 50 percent. Because the patient’s blood pressure may be relatively normal in these patients, at least initially, it’s important to look for metabolic evidence of tissue hypotension, and measuring lactate is a relatively effective way to do that.
While Dr. Rivers’ article in the New England Journal of Medicine helped prove the value of early goal-directed therapy for septic patients, it also demonstrated that measuring lactate levels can help detect the condition. When researchers aggressively treated patients who had normal blood pressure but an elevated lactate level, the group’s mortality rate was 16 percent lower than the control group. Their APACHE scores were also considerably lower.
Recent research from New England Deaconess Hospital reinforced the value of using serum lactate levels to detect sepsis. Researchers found that patients with normal blood pressure but elevated lactate tend to do worse than patients with a normal lactate.
Derek C. Angus, MD, professor and vice chair of critical care medicine at the University of Pittsburgh School of Medicine, says that these two studies are an important part of a growing body of evidence that show why an elevated lactate is a reliable marker for sepsis. “It’s not perfect,” he says of the test, “but it’s probably beneficial to more regularly measure lactate, particularly given the data from the Rivers study.”
When it comes to measuring lactate, an arterial stick is considered the gold standard because it doesn’t produce artificially high readings like venous measures. But Dr. Angus says that if you’re reluctant to do an arterial stick on patients when your suspicion of sepsis is low, there is another approach: Start with a venous stick.
While a venous stick may produce falsely elevated levels of lactate, a low lactate can generally be trusted as a sign that septic shock isn’t likely. If the venous stick produces a high lactate and you strongly suspect sepsis, Dr. Angus explains, you can begin to manage the patient more aggressively.
If the venous stick indicates an elevated lactate and you’re not sure if the problem is sepsis, you can always order an arterial stick to confirm the high lactate level. In many cases, however, Dr. Angus doesn’t think a confirmatory test will be necessary.
“You need to have a much lower threshold for ordering a lactate,” Dr. Angus explains. “One way to make it a lower threshold is to tell yourself that you’re just ordering a venous and add it to your morning labs.”
A narrow window
Once you’re sure the patient has sepsis, you need to act quickly. The New England Journal of Medicine article by Dr. Rivers and the practice guidelines from the Surviving Sepsis Campaign both highlight the importance of beginning treatment in the six-hour window after a patient has been diagnosed as septic.
This is why experts emphasize that simply sending a septic patient to the ICU is not good enough. Dr. Rivers points to recent research indicating that even in a very good community teaching hospital, patients transferred to the ICU often wait hours to see an intensivist.
Additional evidence shows that delays, whether they occur in the ICU or the wards, can be deadly. Richard P. Wenzel, MD, an infectious disease specialist who is professor and chair of the department of medicine at Virginia Commonwealth University in Richmond, co-authored a study a few years ago that found higher mortality rates when patients developed septic shock on the wards, not the ICU.
Researchers found that mortality increased when treatment for sepsis was delayed. “There was a dramatic delay in the time it took to administer a bolus of fluid and pressors that correlated with the poor outcomes,” Dr. Wenzel says, “even though the patients in the ICU had an APACHE score that was 10 points higher.”
What can you do to make sure your patients don’t experience a similar fate? The Surviving Sepsis Campaign guidelines, which were released in early 2004 and sponsored by nearly a dozen international organizations, provide a lengthy list of actions to take within six hours of diagnosing sepsis.
The guidelines urge physicians to focus on early goal-directed resuscitation, followed by blood cultures to determine the proper antibiotic treatment and early administration of broad-spectrum antibiotics. The guidelines also point to therapies like vasopressors and dobutamine.
While you may not wind up managing all these aspects of a septic patient’s care, experts say that hospitalists are in an excellent position to start goal-directed resuscitation in the wards or in the ED.
“The hospitalist can volume-resuscitate someone early on and get the ball rolling,” Dr. Janssen says. “If we wait eight to 12 hours before we get rolling with the protocol, we’ve missed the boat.”
The New England Journal of Medicine article showed exactly why speed is so critical in volume resuscitation of septic patients. When volume resuscitation was started in the first six hours after a diagnosis of sepsis, the study found, patient mortality dropped considerably. When researchers compared different levels of volumes given 24 and 72 hours later, however, they found no significant differences.
When it comes to volume resuscitation of septic patients, speed isn’t the only concern. Intensivists are also urging their colleagues in hospital medicine to embrace higher volumes early on in the resuscitation effort.
Dr. Janssen, for example, says that hospitalists may be surprised by just how much fluid septic patients may require. He notes that an average patient with sepsis may need four to six liters of fluid during resuscitation. That amount that can spike up to nine liters over a 12-hour period in severe cases.
“Surgeons are very good at giving that amount of fluid,” Dr. Janssen explains, “but hospitalists see a lot of patients with heart failure and other diseases, so that volume might be surprising.”
Dr. Angus from the University of Pittsburgh acknowledges that giving so much fluid to patients in such a short amount of time may make some hospitalists nervous, particularly with older, medically complex patients. His personal bias, however, is that a lack of fluids is probably worse than too much fluid.
“The intensivist’s perspective is that the worst that can happen with too much fluid is that you have to give patients a diuretic or intubate them,” Dr. Angus says. “But if you don’t give enough fluids and organ dysfunction results, you may well be giving them a death sentence.”
He adds that in the long run, physicians are probably giving roughly the same total amount of fluids, but they’re front-loading the amounts. “It might feel like you’re giving more fluids than you intended,” he explains, “but over the next two or three days it will even itself out.
“We need physicians to be a bit bolder now and hang a little more fluid than they would have done,” he adds. “The chances are that you’re going to do it anyway. If you do it sooner, you’ll protect the patient against a prolonged episode of hypoperfusion.”
Volume resuscitation may be just the first of many steps in treating septic patients, but experts say that it plays to hospitalists’ strengths. Because of their constant presence on the wards, most hospitalists can regularly monitor patients and make sure that precious time isn’t slipping away.
Dr. Angus urges hospitalists to take advantage of their presence by making sure that fluids are administered soon after being ordered. “It’s not enough to write an order for a liter of normal saline and then walk away from the bedside,” he says.
Dr. Janssen says the same is true of the early use of antibiotics. “Ordering antibiotics is the first step, but making sure that the patients have received them is equally important,” he explains. “As the ordering physician, the hospitalist should keep an eye on whether patients actually receive their antibiotics or not. It’s not enough to just write that order.”
But perhaps most important is raising the awareness of everyone on the inpatient team, from physicians to nurses to residents. And while Dr. Angus sees some movement in the right direction, he compares the current thinking about sepsis to the push to establish coronary care units 30 years ago.
“We’re in the early days,” he says. “It took about 15 years from the time we knew that thrombolytics made a difference before they were widely used for myocardial infarction.
Still, he hopes that physicians won’t take quite so long to embrace new evidence about sepsis. “Back then,” he says, “they didn’t know that the diffusion of new evidence to the bedside took so long, so they had an excuse. We don’t have that excuse today.”
Edward Doyle is Editor of Today’s Hospitalist.
A surprising look at the mortality risk of sepsis
You have a patient with an Escherichia coli bloodstream infection. What are the odds that the patient will die of the underlying disease “or the infection itself?
Richard P. Wenzel, MD, an infectious disease specialist who is professor and chair of the department of medicine at Virginia Commonwealth University in Richmond, likes to ask that question to illustrate just how deadly sepsis can be.
While most physicians guess that the mortality from sepsis is about 5 percent, Dr. Wenzel says the number is closer to 20 percent. Talk about attributable mortality “the number of patients who will die from the infection, and not another underlying disease “and the situation is even more alarming. Dr. Wenzel says that roughly half of the patients with sepsis who die succumb to infection, not the underlying disease.
He says most physicians don’t have an appreciation for the mortality rate of sepsis because they see patients for only a brief period of time. They don’t typically follow patients for 30 days, for example, or they don’t have both a clinical and a population-based sense of illnesses like sepsis.
“If you’re into the field, you understand that there’s a fairly high attributable mortality and that probably half of your mortality is attributable to the bloodstream infection,” he explains. “But if that’s not your training, you may think they had this horrible disease and they were going to die anyway. That’s not true.”
Sepsis resources on the Web
* Practice guidelines from the Surviving Sepsis Campaign, which were published in March 2004, are online. (Click on the button for publications on the left side of the page, then click on the link for guidelines and practice parameters on the right side of the page.)
* More information about the Surviving Sepsis Campaign is online.
* A sepsis bundle jointly developed by the Surviving Sepsis Campaign and the Institute for Healthcare Improvement, along with other resources, is online.
* The study on early goal-directed therapy by Emmanuel Rivers, MD, appeared in the New England Journal of Medicine on Nov. 8, 2001, and is available online.