Journal Logo

Special Report

Special Report

Sepsis: What Works, What Doesn't?

Shaw, Gina

doi: 10.1097/01.EEM.0000669356.83740.76

    The human and financial costs of sepsis are on the rise, according to new estimates based on the largest-ever study of sepsis in Medicare beneficiaries conducted by researchers at the Biomedical Advanced Research and Development Authority (BARDA) and the Centers for Medicare and Medicaid Services (CMS). (Crit Care Med. 2020;48[3]:276, 289, 302;;;

    Approximately 1.7 million Americans develop sepsis each year and nearly 270,000 die from it, with direct and indirect costs of care reaching $62 billion in 2019, the researchers found. In light of the COVID-19 pandemic, these numbers can be expected to rise significantly this year.

    Analyzing data for almost 9.6 million sepsis patients admitted to acute care hospitals from January 2012 to December 2018, the investigators found that the number of fee-for-service Medicare beneficiaries hospitalized with sepsis increased by about 40 percent, while the increase in the Medicare population overall was 22 percent. Three-year mortality rates for sepsis were startlingly high: 75 percent for those who experienced septic shock and 60 percent for those with nonsevere sepsis, compared with 40 percent for patients hospitalized for causes unrelated to sepsis.

    Recognizing Sepsis

    Sometimes the life-threatening infection can be obvious but not always. “I'm looking for signs of infection. Do I think they have an infection driving their process?” said Sara Gray, MD, an emergency and critical care physician at St. Michael's Hospital in Toronto, Canada, and an author of the Canadian Association of Emergency Physicians Sepsis Guidelines, during a 2019 EM Cases Course podcast on sepsis and septic shock.

    “Next, do I think they have organ dysfunction? The guidelines tell us to use the qSOFA [quick Sequential Organ Failure Assessment] score. Practically speaking, I look for organ dysfunction. Have they doubled their creatinine or their bili? Are they altered? Are they hypotensive? Are their platelets half of what they ought to be? Any of those count as signs of organ dysfunction, and if I've recognized that, then I know they're potentially sick. That's when I start looking at that life-threatening piece, and that's where the scores come in.”

    The latest screening tool being used to identify patients with life-threatening sepsis, Dr. Gray noted, is the National Early Warning Score (NEWS), which was more accurate in detecting all sepsis endpoints than qSOFA or its predecessor, Systemic Inflammatory Response Syndrome (SIRS), in one large retrospective study published involving 130,595 adult visits to the ED. (Am J Emerg Med. 2019;37[8]:1490.)

    NEWS incorporates simple measures: respiratory rate, supplemental oxygen requirement, temperature, systolic blood pressure, heart rate, and level of awareness. “It started out as a score to help us find people at high risk of dying on that hospital admission, not specifically for sepsis,” said Dr. Gray. “But we're now seeing it used in sepsis to find that life-threatening group who might otherwise not look that sick. There is good data that when you're looking at somebody with an infection and you're trying to decide are they sick or not sick, the NEWS score is one thing that can help you make that decision appropriately.”

    The Bundle of Care

    The most current version of the Surviving Sepsis Campaign's Hour-1 Bundle of Care for adult sepsis patients includes the following elements:

    • Measure lactate level (remeasure lactate if initial lactate is elevated [>2 mmol/L]).
    • Obtain blood cultures before administering antibiotics.
    • Administer broad-spectrum antibiotics.
    • Begin rapid administration of 30 mL/kg crystalloid for hypotension or lactate level of ≥4 mmol/L.
    • Apply vasopressors if hypotensive during or after fluid resuscitation to maintain MAP of ≥65 mm Hg.

    “The hour-1 bundle encourages clinicians to act as quickly as possible to obtain blood cultures, administer broad-spectrum antibiotics, start appropriate fluid resuscitation, measure lactate, and begin vasopressors if clinically indicated,” says the campaign's website.

    “There is a lot of debate around a number of questions regarding the management of sepsis, but there are also a few things we absolutely know for sure,” said Jessica Whittle, MD, PhD, the emergency medicine research director at the University of Tennessee Health Sciences Center in Chattanooga, who served on the ACEP Sepsis Technical Expert Panel and CMS's Sepsis National Hospital Inpatient Quality Measure (SEP-1) Multistakeholder Work Group. “One is that the early administration of antibiotics saves lives. A couple of large studies have validated this, and they have actually been able to quantify it. A delay in antibiotic administration increases mortality at a rate of somewhere around 7.5-10 percent per hour.”

    Dr. Gray noted that ACEP has not endorsed the SSC antibiotic guideline, however, and that there has been significant pushback about whether this is safe or feasible. “My personal practice has been to have the antibiotics in within an hour of making the diagnosis,” she said. “The best evidence for this is in septic shock, less so in just sepsis.”

    And blood cultures need to be done first to ensure a reliable diagnosis. “If the sepsis is not bacterial, giving early antibiotics has ruined the potential workup findings and culture results for someone who may or may not have been bacterial,” said Murtaza Akhter, MD, an assistant professor of emergency medicine at the University of Arizona College of Medicine. Dr. Akhter and colleagues found that blood cultures taken prior to antibiotic therapy were positive for one or more microbial pathogens in 31.4 percent of patients, compared with 19.4 percent of patients for cultures taken after empirical antibiotic treatment had already started. (Ann Intern Med. 2019 Sep 17; online ahead of print.)

    Fluid Factors

    Early resuscitation of fluid-responsive patients is also critical, Dr. Whittle said. “However, the optimal total volume and total duration of fluid resuscitation remains unclear. Most people would agree that somewhere in neighborhood of a 20 cc/kg IV fluid bolus for most patients with true sepsis would be reasonable. It's continuing high rates of hydration that is potentially a problem.”

    The current SSC recommendations, she noted, recommend that at least 30 mL/kg of IV crystalloid fluid be given within the first three hours. “That seems unreasonable for some patients, particularly because in the earliest recommendations, there was no accommodation for wildly unusual BMIs and no allowance for administration based on ideal body weight,” she said. “That has since been modified, which is really important because for some patients you could otherwise end up recommending unbelievable amounts of fluid based on body habitus.”

    “I want to be able to give the cookbook and say you can give everybody two liters, but that's not true,” Dr. Gray said. “We want to give not too little, not too much, just the right amount for that individual to perfuse their end organs. We're shooting for a MAP of greater than 65 in perfusing them.”

    Dr. Whittle noted that stop orders for fluids must be incorporated. “Sometimes we begin to hydrate these patients and don't include that stop so that they are persistently hydrated with liters more fluid than we realize, which is where people can really get into a dangerous situation. We can't create an order set where you initiate all these orders for the patient and just walk away for 12 hours. We need to monitor their fluid status in real time and give all the fluid they need and no more.”

    What should be your fluid of choice? Dr. Gray said she typically prefers a balanced crystalloid such as Ringer's lactate, and she does not adhere to the once-recommended practice of giving two liters of normal saline as a fluid challenge to sepsis patients before switching to Ringer's to avoid the acidosis associated with normal saline.

    “It's very common these days to still do that, but that is not my practice. I use Ringer's for almost everybody, up front, all the time,” she said. “If you have a septic patient who's acidotic to start and then you give them a bunch of saline, which makes them more acidotic, we are increasing their risk of death. The other thing I dislike is the fact that chloride load is a renal vasoconstrictor and reduces your renal perfusion. So the organ you're trying to save is not being saved effectively when you choose a fluid with that much chloride in it.”

    A significant mortality difference was seen in the sepsis subgroup of the SMART trial between the Ringer's lactate group and the normal saline group: 26.3 percent v. 31.2 percent. (Am J Respir Crit CareMed. 2019;200[12]:1487; “There is mortality data here in that subgroup,” said Dr. Gray. “The main group where I avoid Ringer's is traumatic brain injury, where I want to keep the sodium high.”

    Pressors: Which, When, How Much?

    Norepinephrine is the established pressor of choice for sepsis, Dr. Whittle said, noting that high-quality evidence from multiple randomized, controlled trials shows that it is superior to other vasopressors for patients with septic shock. “But there is some question as to what your second-line choice should be,” she said. “The guidelines call for vasopressin, but, for example, if you believe the patient needs increased cardiac output, dobutamine would be another option.

    Begin to titrate the dosing of norepinephrine at the low end of your hospital's dosing range, Dr. Gray said. “Start low, but this is something you can dial up every minute or two based on their blood pressure. Do a cuff pressure, if they're still low, turn it up again. This is not something where you walk away and come back after an hour. This is something where they'll be very responsive.”

    Dr. Gray recommends adding the second pressor if needed when you're at a moderate dose of norepinephrine for your institution. “You want to add it before you've maxed out your levo,” she said.

    Central lines are no longer thought to be mandatory for all sepsis patients, Dr. Whittle said. “We now have solid data saying that high-quality peripheral lines are safe for administration of low-dose vasopressors for a limited time. Now, if your patient is going to need pressors longer than that, or if they're going to need more ports than you can get in a peripheral line because you're administering numerous meds, sometimes you will need a central line. Over the last few years, my practice has evolved to where I initiate pressors more quickly and allow less hypotension, preferably almost none, but I initiate far fewer central lines.”

    Dr. Akhter added that the ProCESS trial showed that a central line isn't needed to monitor everyone's central venous pressure. “But that doesn't mean it's a meaningless metric,” he said. “What would be nice would be a noninvasive way of measuring CVP. In the ED, some will still use the passive leg raise maneuver—some data suggest that's a decent way of monitoring fluid status. We are always looking for other ways to see whether a patient has enough intravascular volume.”

    What About Vitamin C?

    What about the promise of a cocktail of intravenous vitamin C, hydrocortisone, and thiamine, heralded as a potentially life-saving therapy in sepsis and septic shock? A protocol developed by Paul Marik, MD, a professor of medicine and the chief of pulmonary and critical care medicine at Eastern Virginia Medical School, calls for 1.5 g IV vitamin C every six hours, 50 mg IV hydrocortisone every six hours, and 200 mg IV thiamine every 12 hours. A retrospective study of the intervention published in Chest in 2017 involving treatment and control groups of 47 patients each found that hospital mortality was 8.5 percent (4/47) in the treatment group, compared with 40.4 percent (19/47) in the control group (p<0.001). (Chest. 2017;151[6]:1229.)

    The VITAMINS trial called those findings into question in early February, however. That trial randomized 216 patients with septic shock at 10 ICUs in Australia, New Zealand, and Brazil to receive the Marik protocol or IV hydrocortisone alone until shock resolved or 10 days had passed. After seven days, there was no difference in the primary outcome, time alive and off vasopressors (122.1 hours with vitamin C, 124.6 hours among controls). The trial did, however, find a significant reduction in organ failure with vitamin C. (JAMA. 2020;323[5]:423;

    Dr. Marik questioned the validity of the VITAMINS findings, however. “Despite the fact that VITAMINS had an IRB-approved delayed consent provision, it appears to have taken an inordinate time from hospital admission until the first dose of vitamin C was administered. VITAMINS therefore does not replicate real-word experience,” he said. “In addition, it is important to recognize that sepsis is not a disease but a syndrome that includes a heterogenous group of patients with various diagnoses.

    “It would appear that there was marked heterogeneity of patients in the VITAMINS study; this would dilute out any clinical benefit,” he said. “Furthermore, the approach to fluid resuscitation is a critical factor in the management of septic shock as fluid overload will diminish the clinical benefit of this therapy. No fluid data was presented in the VITAMINS publication.

    “There is no question in my mind that this intervention saves lives” Dr. Marik said. “This has been the experience of hundreds of clinicians around the world who have used this intervention early in the course of severe sepsis and septic shock when combined with high-quality supportive care.”

    Dr. Whittle said vitamin C's potential in sepsis remains an area of interest. “We are still unfortunately in the same place that we have been; people who really believe in vitamin C are going to find parts of the trials that they feel validate their arguments, and people who see no benefits will certainly have plenty of ammunition for their side of the argument,” she said. “At this point, people still are digging their heels in to their own opinions and interpreting this data through that lens. I believe we still don't have an answer, but there is not overwhelming, reproducible evidence that we should change our standard of care. These findings do suggest that we should try to get more rigorous and standardized ways of measuring vitamin C and thiamine deficiencies in septic patients, so that we can begin to ask this question in a more meaningful way.”

    Ms. Shawis a freelance writer with more than 20 years of experience writing about health and medicine. She is also the author of Having Children After Cancer, the only guide for cancer survivors hoping to build their families after a cancer diagnosis. You can find her work

    Read “A Dream of Nuanced Sepsis Care” by Graham Walker, MD, on p. 18.

    Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.