Until his doctor pronounced him cured last week, Peter Elster had no idea how close he came to dying.
“He said my chances were probably 50-50,” Elster said. “I didn’t realize I was deathly sick.”
Elster had been in the emergency room in early August after developing excruciating pain in his back. Doctors initially diagnosed a muscle spasm and sent him home.
After 10 days of treating the spasm, the Bend man had lost 20 pounds. He couldn’t eat or sleep. MRI and CT scans revealed a massive abscess in the muscles around his spine and into the abdominal cavity below his right lung. He spent 22 days at St. Charles Bend recovering from two surgeries to remove the abscess and treat the infection.
But what truly threatened his life was the leading killer in U.S. hospitals today, accounting for more deaths than heart attacks, lung cancer or breast cancer.
Elster’s infection had progressed to sepsis, a systemic inflammatory response in which the immune system goes haywire and attacks the body. The condition affects more than 1 million Americans each year, killing 215,000, a mortality rate of more than 20 percent. And the longer the condition goes untreated, the higher the death rate.
It is one of the most vexing challenges in health care and the single most expensive condition for hospitals to treat, accounting for more than $24 billion in spending each year.
Yet research studies and hospital initiatives have shown the rate of sepsis mortality could be cut in half without any new scientific discoveries or medical advances. By focusing the attention of doctors and nurses on the warnings signs of sepsis, catching it early and treating it quickly, hospitals could save thousands of lives and millions of dollars with little more than antibiotics and fluids.
“The fascinating thing about sepsis is we don’t have new drugs or even new technologies to treat it, but we’ve been doing better,” said Dr. Allan Walkey, a medical professor at Boston University School of Medicine. “So how have we been doing better? By improving the processes of care.”
But with great variation in mortality rates among hospitals — including many where more than half of severe sepsis patients die — sepsis care could be the biggest untapped opportunity for improving health care today.
Despite its high mortality rate and heavy financial costs, sepsis remains a mystery to most Americans. A 2013 Harris Interactive poll found that more than half have never heard the term. Few understand the heavy toll it takes in terms of both resources and lives, and even fewer would rank it among the nation’s deadliest diagnoses.
“Not a day goes by that someone does not look at me and say, ‘No, that’s not possible. If it were true I would know about it,’” said Tom Heymann, executive director of the advocacy group Sepsis Alliance. “We felt like we had found a place on earth where no one had been to before. How could this situation exist that so few people know about it, and so little is being done about it proportionately to the effect it’s having? It’s mind-blowing.”
Sepsis represents a continuum more than a specific diagnosis. The initial stage, known as systemic inflammatory response syndrome, or SIRS, starts when an infection triggers an inflammatory reaction throughout the body. It’s diagnosed when patients have some combination of fever, increased heart rate, slowed breathing or abnormal white blood cell counts.
As the condition progresses, the immune response begins to affect different organs, a state known as severe sepsis. And if it worsens still, it can lead to septic shock, when blood vessels expand, causing a dangerous drop in blood pressure that isn’t responding to treatment.
At each stage of the progression, the mortality rate increases. About 10 percent of individuals with infections die, but the risk of death goes up to 20 percent with SIRS, 40 percent with severe sepsis and a staggering 80 percent with septic shock.
But what makes sepsis truly scary is how quickly patients can move along that spectrum.
In December 2011, Dana Mirman of Naples, Fla., noticed a tiny bump on her shoulder. She assumed it was just a bug bite. The next morning, she felt like she had the flu, and by midafternoon, was unable to get out of bed. Within 24 hours of noticing the bump, she was in full septic shock.
“I could have tried to sleep it off,” said Mirman, who now works with Sepsis Alliance. “And it would have been catastrophic.”
Her husband insisted she go to the emergency room, where doctors immediately diagnosed her condition and treated her with antibiotics and fluids. When she didn’t respond to the initial treatments, they quickly transferred her to the intensive care unit, where doctors put in a central line to deliver medications to reverse her dropping blood pressure.
“I’m here today because they responded so appropriately,” she said. “I never waited in the waiting room.”
Her case stands in sharp contrast to the experience of Rory Staunton, a 12-year-old New York City boy who in 2012 scraped his elbow in the school gym and became sick overnight. His parents took him to the emergency room, but doctors sent him home with a diagnosis of gastric flu. Three days later, the previously healthy boy was dead from septic shock with multiple organ failure.
“That was the day our lives changed forever,” his father, Ciaran Staunton, said at a U.S. Senate hearing in September. “It was the day that no parent ever wants to go through. It was the day our beautiful young man died of sepsis, something we had never heard of before.”
Sepsis hasn’t become a household word in part because doctors often don’t use the term with patients and families. In the past, it was commonly called blood poisoning, and doctors may still refer to sepsis as a complication of the initial infection or disease. Patients could be admitted to the hospital with cancer or pneumonia and develop sepsis as a result. If they die, the death is commonly attributed to the underlying disease even though it may have been sepsis that killed them.
“With sepsis, it’s so sudden and it’s so devastating, and it so often ends so badly, that you end up having a population of people who are bereaved or end up having survived, who are very, very traumatized and don’t know what’s happened to them,” Mirman said.
That has left sepsis patients much less organized and outspoken than breast cancer survivors or individuals with HIV, who live at home and interact with others as they are being treated. Sepsis hits patients hard and fast, leaving little time to rally for support or promote public awareness of the condition.
The Stauntons’ story became major news in New York, prompting state officials to adopt Rory’s Regulations, which include new protocols for hospitals to identify and treat sepsis. The Sepsis Alliance is pushing for similar regulations to be adopted nationwide.
Even if patients don’t know much about it, sepsis is garnering more attention at hospitals. According to the federal Agency for Healthcare Research and Quality, the number of sepsis diagnoses increased 32 percent from 492 cases per 100,000 hospitalizations in 2005 to 651 per 100,000 cases in 2010. It’s unclear, however, how much of that increase is due to more actual cases of sepsis and how much to hospitals doing a better job of identifying it when it occurs.
It can be a difficult distinction to parse out because hospitals often differ in how they apply billing codes to sepsis cases. But there is reason to suspect that sepsis cases are becoming more frequent. Age is the biggest risk factor for sepsis, so with an aging population as well as increasing rates of diabetes and kidney disease, sepsis rates can be expected to climb.
“It may not be as dramatic as what the billing codes show, there may be some inflation there, but it does seem to be increasing,” Walkey said.
Mortality rates for sepsis, meanwhile, are dropping, although the magnitude of the decline is also a matter of debate. If hospitals are getting better at finding and coding less severe cases as sepsis, it could artificially lower the death rates. There is also concern that hospitals could be gaming the system, diagnosing marginal cases as septic in order to get higher reimbursements from public and private insurance plans.
But Walkey does see signs of improvement. A study he led last year looked at patients enrolled in the control groups of clinical trials, which assigned a severity score for each patient. He concluded that the mortality rate for patients with severe sepsis had declined steadily from 47 percent in the early 1990s to 29 percent in the mid-2000s.
With no new effective drugs or treatments introduced over that time frame, it could be a sign that hospitals are implementing proven sepsis protocols. A landmark 2001 study published in the New England Journal of Medicine found that early treatment with fluids could help prevent the sudden collapse that often leads to death from sepsis, cutting the mortality rate for severe sepsis or septic shock from 46 percent to 30 percent. Meanwhile, a 2006 study found that for every hour delay in getting an antibiotic after septic shock begins, survival rates drop by 8 percent.
“We’ve always known that people with sepsis need antibiotics and fluids,” Walkey said. “And that’s still basically what we do. But how we give them is a little different.”
Now doctors rely on complex algorithms that trigger interventions on the basis of laboratory test results and patient vital signs. It’s a tricky process, particularly as doctors try to juggle a constant flow of real-time data as well as their own medical judgment. One study found that even with the data in hand, doctors are able to accurately diagnose sepsis at the bedside only 70 to 80 percent of the time.
And evidence suggests that hospitals still have a long way to go to get it right. Researchers from the University of Pennsylvania recently mapped sepsis deaths by county, finding wide variations in outcomes. They identified regional hot spots in the South and Mid-Atlantic states, where sepsis mortality rates were four times the national average. In the Southwest and Mountain states, they found a cool spot cluster of 157 counties that had disproportionately lower death rates.
“The thing that jumps out at you is the variation is so dramatic,” said Dr. David Gaieski, a professor of emergency medicine at the school and lead author of the study. “You sort of think that once you took into account age … you would get similar rates per 100,000 population across the country, and you don’t. So then the question is why?”
Various theories abound from differences in coding to variation in genetic susceptibility to sepsis. Some hospitals may not make the same commitment to training staff and stressing the importance of early treatment for sepsis.
But Gaieski said it may come down to how difficult it is to recognize sepsis in its early stages.
“In the vast majority of cases, it just smolders along and has a real mix of symptoms and signs,” he said. “It’s sort of protean in its presentation.”
That is what makes sepsis such a challenge for hospitals. None of the warning signs of sepsis is present in all sepsis cases. International consensus guidelines list no fewer than 24 possible signs of sepsis. So unless doctors and nurses are vigilant, it can be easily overlooked. Patients could come to the emergency room with normal blood pressure and temperature, yet might be well on their way to severe sepsis.
“They might be asked to sit in triage for two hours, and those are two hours in which their lives could have been saved,” Gaieski said. “One of the hardest things for a lot of health care providers to wrap their heads around is that this person who you just triaged and put in a chair before they can be seen, has a mortality (risk) of 40 to 50 percent.”
And because sepsis can stem from so many different conditions, the treating physicians might not be well-versed in sepsis protocols. A patient with a broken leg, for example, could be treated by an orthopedist who doesn’t see as many sepsis cases as an emergency room physician or critical care specialist.
“A lot hinges on an index of suspicion by the physician, careful assessment of the patient’s vital signs, physical examination, and some basic laboratory tests,” said Dr. Jon Lutz, an infectious disease specialist at Bend Memorial Clinic who treated Elster’s sepsis.
Additionally, reducing sepsis mortality requires hospitals to go against the grain of the current push in health care to avoid excess tests and interventions. Just last week, Centers for Disease Control and Prevention Director Tom Frieden called on hospitals to be more judicious in their use of antibiotics, to help avoid the rise of antibiotic-resistant bacteria.
Many believe those superbugs are exacerbating sepsis cases among younger, healthy individuals, particularly when doctors give more powerful, broad-spectrum antibiotics as a first therapy.
But sepsis protocols call for the early use of broad-spectrum antibiotics that can kill every type of bacteria. It can take two to three days for bacteria cultures to be grown in the lab and identified. If doctors wait that time so they can give a more targeted antibiotic, it might be too late.
Moreover, the differences between a normal immune response to infection or trauma and the start of sepsis can be subtle. A patient infected with the flu could have both a fever and an elevated heart rate, technically meeting the definition for sepsis. Yet, doctors aren’t going to pump every flu patient full of antibiotics and fluids.
“It’s so easy to be deemed ‘septic,’” said Dr. Chris Richards, an emergency physician at St. Charles Bend, using air quotes to underscore the fuzziness of the definition. “It’s kind of a catch-all term.”
Doctors and nurses rely on test results to track patients, trying to determine when they might be moving beyond mere infection to sepsis, and adjusting treatment as their conditions change. But it can be challenging to get it right, and patients can go downhill quickly if it’s not immediately recognized.
“Patients can come in and be right on the fence,” Richards said, “and before we can start therapy on them, they can start deteriorating. Even in treatment, you have to move along the spectrum.”
Those complexities leave hospitals constantly battling to stay on top of sepsis. The four hospitals in Central Oregon have seen more than 600 cases of sepsis in the past year, at an average cost of more than $26,000 per case, according to 2011 data submitted to a state hospital database. St. Charles does not specifically track sepsis mortality rates, and therefore couldn’t report how many of those 600 patients died of sepsis, either in the hospital or after being discharged to home health or nursing home care.
Nationwide, hospitals and quality improvement organizations have made sepsis a point of emphasis, implementing various initiatives to improve recognition and treatment of sepsis, before it progresses.
The Center for Transforming Healthcare, a quality improvement arm of the group that accredits hospitals, is completing a pilot program with five hospitals that had a goal of reducing sepsis mortality by 20 percent. The center plans to release the results soon after their final meeting in April, but organizers called the results “extremely promising.” They plan to roll out a web-based sepsis application for national use.
Meanwhile, nine hospitals in the San Francisco area recently completed a two-year sepsis campaign that reduced sepsis mortality from 28 percent of cases to 17 percent, mainly through education of front-line clinicians and a screening tool.
“It was a paper-based tool, an expanded set of vital signs,” said Julie Kliger, a consultant with The Altos Group, who led the effort. “Every shift, regardless of the reason the patient was in the hospital, and every admission through the emergency room got screened.”
A St. Charles spokesperson was not aware of any specific sepsis initiatives in any of the four Central Oregon hospitals.
Kliger is now working with a technology company that is developing a handheld device that would help clinicians catch sepsis early and cut down on the time lag between when a nurse suspects sepsis, the condition is confirmed, and treatment can begin. Often valuable time is lost when a nurse must track down a physician to order the necessary lab tests to confirm sepsis. They must then contact the lab to send a phlebotomist to get a blood sample, and wait for the results.
“Meanwhile, tick-tock, tick-tock, it’s four hours later sometimes,” Kliger said.
By automating the process, the order for the test can be triggered automatically based on preset protocols.
“So you’re kind of truncating the time element,” she said. “This is the kind of thing that assistive technology will help with.”
Many firms are now trying to develop technological solutions to the sepsis problem. Some hospitals, including Oregon Health & Science University, are using programmable mannequins to help train doctors and nurses on the signs of sepsis.
Lockheed Martin, a defense contractor, recently repurposed a missile control system adept at updating various streams of data in real time, to help clinicians make sense of changing vital signs and test results to identify sepsis faster. Others are working on bedside tests that could more rapidly diagnose sepsis or the type of infection involved.
Until those are perfected and proven, however, success in reducing the impact of sepsis will continue to rely on very old-school approaches, using basic vital signs and good clinical judgment to quickly and accurately diagnose patients.
“People are trying very hard to have technology help us out with rapid integration of information and genetics. That may bail us out,” said Dr. Stephen Opal, an infectious disease specialist at Brown University. “But at the present time, it’s really the astute clinician being aware of the problem. That’s just the nature of the beast.”
—Reporter: 541-617-7814, firstname.lastname@example.org