Sepsis still infected with uncertainty

Experts from CHEST 2016 debate screening, fluids, more

There are so many topics of debate in sepsis care, it's hard to keep track. CHEST 2016, the annual gathering of ICU and pulmonology experts, featured multiple pro/con debates, but they captured only a fraction of the questions about sepsis currently troubling hospital physicians.

How to screen for sepsis, whether to measure central venous oxygen saturation (SCVO2), which and how much fluid to give, and even how to interpret the three major trials of early goal-directed therapy published in 2014 and 2015 were all under discussion at the meeting in Los Angeles in October.

“It has been a very busy four years of critical care data,” said Laura Evans, MD, an associate professor at New York University and director of critical care at Bellevue Hospital Center in New York City.

The data and debate have been so intense that, although Dr. Evans was scheduled to speak at the meeting about new Surviving Sepsis Campaign guidelines, she could only hint at their focus, as the expected release date was pushed back from October 2016 to early 2017. “Achieving consensus can take a little time,” she said.

In the field of sepsis care, that might be an understatement.

The big three

Almost every sepsis speaker at the meeting delved into the data from the ProCESS, ARISE, and ProMISe trials, which were all published in the New England Journal of Medicine and found that early goal-directed therapy (EGDT) didn't reduce mortality of sepsis patients compared to standard care.

There seemed to be a consensus among the expert speakers that the failure of the three trials to show benefit mainly proved that standard care for sepsis had dramatically changed in the decade since EGDT was invented, with interventions like very early antibiotics becoming the norm.

“That wasn't even a concept at the time that [Emanuel Rivers, MD] did the study. We hadn't had any data to show that early antibiotics equates to better survival....Whereas, in these three trials, we had all the way to 100% had received antibiotics within two-and-a-half hours,” said Steven Q. Simpson, MD, FACP, a professor of pulmonary and critical care medicine at the University of Kansas in Kansas City.

Similarly, the new trial patients also received significant quantities of IV fluids soon after sepsis diagnosis, even before their randomization into an intervention or control group. That may partially explain why the research didn't appear to show any benefit to monitoring SCVO2, according to ACP Member Angel Coz, MD, an assistant professor at the University of California, San Francisco, in Fresno.

Dr. Coz took the pro side in a debate over whether SCVO2 is a helpful target in sepsis care. “The new sepsis trials do not negate the utility of normalizing SCVO2 when it's low,” he told the audience, noting that patients in the new trials had average SCVO2 levels close to target at the start, thus the trials weren't really testing the effects of monitoring and treating low SCVO2.

Other recent research has confirmed that patients with low levels have worse outcomes, he noted, citing a study published in November 2014 by Critical Care finding an association between SCVO2 under 70% and 28-day mortality. “It gives us not only prognostic information, but it tells us that we need to act on low SCVO2—try to correct it and change the outcomes,” Dr. Coz said.

His opponent was not so convinced. “If it means treating critically ill patients and making them not critically ill anymore, then absolutely normalizing SCVO2 is a good thing. If it means we should take a low lab value and infuse packed red blood cells or dobutamine to make an abnormal lab value normal again, I would disagree with that,” said Lewis Satterwhite, MD, an assistant professor of pulmonary and critical care at the University of Kansas.

Research has not shown that using transfusions and inotropes to normalize SCVO2 improves mortality, he reported. “The majority of patients with severe sepsis/septic shock will have a normal or high Scvo2 already at baseline, so we're talking about a minority of patients. The question then becomes ‘What is different about them?’” said Dr. Satterwhite. “We should be a little bit more thoughtful when our patients with septic shock do have low SCVO2.”

That thoughtfulness could entail looking for cardiac problems, both experts agreed. Cardiac ultrasounds can help identify appropriate treatment for sepsis patients, noted Dr. Coz. “These are great tools, and if you have the expertise, they should be incorporated in your ICUs.”


The experts, and the guidelines, also agreed on the importance of early fluid resuscitation for sepsis patients. The question up for debate at the conference was whether this fluid should be normal saline or a balanced salt solution.

The case for balanced salt solutions, such as Ringer's lactate, was made by Alexander Niven, MD, FACP, a pulmonology and critical care specialist at Mayo Clinic in Rochester, Minn. “When we use normal saline, what we do is we create a hyperchloremic metabolic acidosis,” he said. “And we think that those elevated levels of chloride may place our patients at increased risk of acute kidney injury, perhaps microcirculatory dysfunction, perhaps more inflammation, and there's at least an association that suggests increased mortality there.”

Experts debated whether sepsis patients should be resuscitated with normal saline or a balanced salt solution Photo by Thinkstock
Experts debated whether sepsis patients should be resuscitated with normal saline or a balanced salt solution. Photo by Thinkstock

His debate opponent, Dr. Evans, conceded his first point but debated the rest. “Are we seeing here a laboratory effect...or are these really clinically relevant and associated with adverse outcomes in our patients? I don't think my patient cares whether his chloride is 115 or 105. My patient cares about the hard end point of whether they do better or worse,” she said.

While Dr. Niven cited studies finding associations between saline and kidney injuries, Dr. Evans noted the absence of research directly comparing balanced and unbalanced crystalloids in sepsis patients. “I'm willing to change my mind if somebody shows me good data, but at this point in time, I think it's a very reasonable choice to use either,” said Dr. Evans. “I don't want you to delay any fluid resuscitation worrying about what bag you're bringing to the bedside.”

In a separate lecture, Dr. Coz agreed that the evidence for choosing between crystalloids is limited, but he noted other areas of certainty. “What is clear is we should not use starches. Albumin—I think there is a lot of controversy whether it's really useful or not. We know for a fact that albumin is more expensive, and we know so far in the literature, there is no survival benefit,” he said.

Other hot topics

Dr. Coz also weighed in on how to screen for sepsis, which has been a hot-button issue since the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) last year introduced the quick Sequential (Sepsis-related) Organ Failure Assessment (qSOFA) as an alternative to the systemic inflammatory response syndrome (SIRS) criteria.

“We can all agree that SIRS are very sensitive, sometimes extremely sensitive. qSOFA is proposed as a new screening tool to recognize sepsis outside of the ICU. However, we have to recognize a lot of times it will not be as sensitive as we would like,” said Dr. Coz.

The two scores were compared to each other, the Modified Early Warning Score (MEWS), and the National Early Warning Score (NEWS) in a study published online in September 2016 by the American Journal of Respiratory and Critical Care Medicine. The MEWS and NEWS showed better discrimination than qSOFA in that comparison, Dr. Coz reported.

“These tools are just as specific, if not better, but more sensitive. So I think there's some food for thought here as far as how we should screen,” he said. “This is something that we should tailor to our hospital and adapt to what tools are available.”

Alert systems that are based on highly sensitive tools can be modified to work efficiently, noted Dr. Simpson. “The problem with your EMR and its best practice alert is that it doesn't know who actually has SIRS present and is a septic patient or just has SIRS present with no infection. How we dealt with that in our institution is that there is a best practice alert, but the first question it asks the nurse is, ‘Do you have a reason to believe that the patient has an infection?’” If the answer is no, then the screening process can stop right there, he explained.

Good sepsis tools have recently become a hot commodity, noted Mitchell Levy, MD, a professor of medicine and chief of critical care, pulmonary, and sleep medicine at Brown University in Providence, R.I.

He described the growth of the Surviving Sepsis Campaign from a voluntary quality improvement project of interested physicians and nurses to the basis of CMS's SEP-1 core measures. “If you live in the United States, you have to adhere to them,” he said. “We may disagree about whether this is good or bad news, but really the campaign led to this, which is mandatory public reporting.”

The experts offered advice on complying with the new measures, which were slated to start being collected in either October 2016 or January 2017 and could eventually be tied to hospital payments, according to Dr. Levy.

Dr. Coz focused on the SEP-1 requirement to reassess patients after initial therapy. “One way, that is probably the easiest, to accomplish the six-hour bundle measure is repeat a focused physical exam, and if you find the skin is looking [mottled] or the capillary refill is delayed, that is going to tell you that something needs to be done,” he said. “I think no one can argue whether that's something we should do, even if the core measure didn't say we need to reassess our patients before the six-hour mark.”

Dr. Evans offered some reassurance that hospitals would not be expected to achieve perfect or near-perfect compliance with the SEP-1 bundles as they might with some other measures. “Even in these very, very high-reliability organizations, they're not approaching 100% on this core measure,” she said.

The future

Given uncertainties on so many fronts, the markers of ideal sepsis care will also probably continue to change. Dr. Evans couldn't reveal any details of the upcoming guidelines, but she did review some studies that the writers of them considered.

The most recent iteration of the guidelines, published in 2012, recommended seven to 10 days of antibiotic therapy. “This is an interesting discussion point, because we're trying to figure out antibiotic stewardship and duration of therapy becomes a big question,” said Dr. Evans.

A recent trial of patients with intraabdominal infections that compared fixed-duration antibiotic therapy of about four days with a longer course found similar outcomes in each group and was considered by the guideline writers. “That's suggesting that perhaps there is some leeway to reconsider duration of antibiotic therapy,” said Dr. Evans. The study appeared in the May 21, 2015, New England Journal of Medicine.

The experts also looked at new data on the positioning of ventilated patients, specifically a study that tried 16 hours of prone positioning for patients with acute respiratory distress syndrome and found better outcomes than those of patients kept in supine position. “What they saw here was a very, very large impact on mortality,” said Dr. Evans. That trial was published in the June 6, 2013, New England Journal of Medicine.

Looking even farther into the future of sepsis care than the upcoming Surviving Sepsis Campaign guidelines, multiple speakers mentioned the debate over steroids for sepsis patients, noting that recent research has had mixed results. “Steroids come and go. And I think they're about to come back again,” said Dr. Levy. “Stay tuned.”