Some hospitals are addressing sepsis by quickly identifying high-risk patients in the emergency department and treating them to stop progression.
This article appears in the June 2014 issue of HealthLeaders magazine.
Sepsis, which can lead to severe organ failure and shock, is a leading if not the biggest cause of hospital mortality. This insidious, under-recognized consequence of infection sickens as many as 1.1 million people in the United States each year. The mortality rate for sepsis is estimated to be between 28% and 50%.
It also is an expensive disease to treat in a hospital, with healthcare costs estimated at $17 billion a year in the United States, in part because half of sepsis patients require an intensive care unit stay. Many who survive its later stages endure impaired physical and mental function, and thus a deteriorated quality of life that costs the healthcare system even more.
But now, better protocols and greater penalties against hospitals with higher mortality rates have prompted bold and successful efforts to stop sepsis in older people at higher risk, as well as young children and adults. These days, some hospitals are leading the way to quickly identify high-risk sepsis patients in the emergency department and treat them to stop progression, and get it all done within the first three or six hours.
Some 16 states now have a sepsis law or regulation governing protocols or reporting, and a 17th, Massachusetts, is considering one in its state Senate.
At the federal level, the Centers for Medicare & Medicaid Services last year signaled its intention to add a "severe sepsis and septic shock management bundle" to its EHR Hospital Inpatient Quality Reporting program, which could take effect as early as next year. If that happens, it will serve as another financial incentive to push sepsis prevention, for hospitals already are penalized for higher 30-day mortality rates, driven all the higher by so many sepsis patient deaths.
So it stands to reason that acute care providers, especially big ones like 17-hospital North Shore-LIJ Health System in New York City, Northern California's 21-hospital Kaiser Permanente, and the 42-hospital Carolinas HealthCare System in North Carolina, South Carolina, and Georgia, would be rushing to find better ways to deal with sepsis.
They want to recognize sepsis when patients first appear with symptoms in the emergency department, and thwart it before it causes organ dysfunction and shock.
That's what prompted NS-LIJ President and CEO Michael Dowling and his team to make sepsis a top priority, and get mortality rates way down. It's been a five-year journey but, Dowling says, they've made tremendous progress.
In January 2008, the mortality rate for sepsis, severe sepsis, and septic shock was about 32% at NS-LIJ. After systemwide guidelines were initiated in the summer of 2009, the rate dropped to 23%. The system held an international sepsis conference in 2010, and after a focus on early identification and early antibiotic administration in the ED was launched in November 2011, mortality dropped to 20%. By July 2013, the rate was down to 13%, and NS-LIJ infection experts expect further reductions this year.
"This whole effort emanated from our discussions over the years about making sure we have the lowest mortality we can possibly have, and to try to get to zero," Dowling says. "We recognized that sepsis was the biggest driver."
Back then, Dowling and sepsis specialists from other hospitals recall, sepsis was a condition that was dealt with only after patients had became so sick they needed to be in the ICU. Sepsis turned to severe sepsis with organ dysfunction, and often to septic shock.
Now the entire process of sepsis detection and treatment has been moved up to within 60 minutes after the patients arrive for administration of antibiotics. Bundles of care are administered within goal points of three or six hours, depending on the patient's symptoms. Protocols involve measuring lactate levels, obtaining blood cultures prior to antibiotic administration and then administering appropriate antibiotics, treating low blood pressure, administering fluids, and several other strategies including the use of a catheter or PICC line to measure blood oxygenation in the heart.
"We now believe very, very strongly that you have to begin dealing with sepsis much earlier in the process," within moments that a potentially septic patient arrives in the ED, Dowling says.
That's not easy because early sepsis symptoms—fever, chills, rapid breathing, rapid heart or respiratory rate, rash, confusion and disorientation—can resemble a number of other conditions, from influenza to pulmonary embolism or even side-effects of certain drugs.
Martin Doerfler, MD, NS-LIJ's senior vice president of clinical strategy and development, says that the Long Island system's success is now "something that's a source of institutional pride, not just at Mr. Dowling's level, but at the very front lines. People at the reception desk, the first nurse to see the patient, the house staff in the ED—they're all trying to keep moving these numbers, pay attention to the month-to-month reports on how well we're providing antibiotics within a specified period of time."
John D'Angelo, MD, NS-LIJ's senior vice president of emergency medicine, says the key is speedy action—even before test results come back—when a potential septic patient presents in the ED.
"What's challenging is that we define folks that have sepsis infections with these subtle vital sign changes, or little bumps in their white counts. But the reality in emergency medicine is that for every one person who has these changes that are sepsis, we have a dozen people with the flu or a cold or strep or other things. Every college kid in the ED with a strep throat has a pulse over 90 when they have a fever."
So the NS-LIJ team has devised a "code sepsis," which is recognition of a series of vital sign abnormalities for patients who are being tested for sepsis. "And for those folks, the team descends on those people even though we don't have lactate levels or white blood count levels back. We're not going to wait an hour and a half for that. We'll get the IVs and fluids in, make our best assessment on what the infection might be, and start antibiotics. We don't wait a long period of time when we'd lose ground on fighting this infection."
It's complicated, D'Angelo says, "because people present to the ED with a plethora of complaints; that's often an extension of their current diseases, like cellulitis, pneumonia, urinary tract, or even viral infections. But now, it's gotten to a point where there is a systemic response that is initially very subtle."
NS-LIJ is not the only healthcare system that has been leading the way since 2008.
Kaiser Permanente's 21 Northern California hospitals implemented a comprehensive sepsis program to care for its patients.
Kaiser's medical director for quality and safety, Alan Whippy, MD, says that in 2008, teams reviewed the last 50 deaths in each of its hospitals at the time "to see whether there were opportunities to plan better, to recognize problems and prevent them."
They realized that a huge number of hospital deaths were due to sepsis, "and at the time, we were not doing a fabulous job at identifying people with sepsis early. We said, 'We want to find it early; we don't want to find it on day 2, after the patient has become seriously ill.' "
By working with Kaiser's emergency department teams, the hospitals tripled the number of patients they diagnosed with sepsis, from 35.7 per 1,000 admissions in July 2009 to 119.4 in May 2011 "and greatly increased the aggressiveness of treatment in the early hours," Whippy says. They also increased the percentage of patients who had blood cultures and serum lactate levels drawn from 27% to 97%.
Their three rules are "find it early; identify patients at risk for deterioration; and treat sepsis aggressively and reliably at all levels of risk."
Whippy says that it's hard to tell for sure how much Kaiser's sepsis mortality has dropped to date, because so many more people who are less sick, with a lower risk of death, are being captured in the data, and because so many sepsis patients may not die in a hospital but within 30 days after discharge. But Whippy says the Northern California system has reported a 40% decline in risk-adjusted inpatient mortality from sepsis from 2007 to 2013.
Risk-adjusted hospital mortality declined 25% between 2007 and 2013. The number of hospital deaths from any cause decreased 17.6% between 2007 and 2013. The raw, all-cause hospital mortality rate decreased 12% in this time period.
Additionally, expected length of stay for patients with a principle diagnosis of sepsis decreased by 0.4 days from 2007 to January 2010. And risk-adjusted sepsis length of stay decreased 25% from 2007 to 2013.
"What's very clear is that we're seeing better outcomes in our hospitals," Whippy says. "This is very important because sepsis is the No. 1 cause of hospital mortality and there is ample evidence that it has been historically under-recognized, underestimated, and under-treated."
Whippy adds that leadership buy-in, to make sure that the ED staffs at all 21 hospitals as well as those who practice in the ICU were on board, was essential.
In the past three years, Kaiser has been developing programs in the hospitals that help patients who have survived sepsis to return to full function, because patients who recover from sepsis often endure long-term functional disabilities, perhaps because of brain damage that is as yet poorly understood.
Kaiser is reducing the use of sedatives previously used in patients with sepsis because it may reduce the amount of delirium and impairment that make it difficult for them to think; the teams work to get people out of bed and moving, too, Whippy says.
At Carolinas HealthCare System, the strategy is to train for recognition and treatment of sepsis by using expensive, high-tech mannequins, says Chief Academic Officer Mary Hall, MD. Medical students, residents, and even experienced physicians and nurses work with mannequins and staff who role-play as family. "Every aspect of sepsis care can be done in the simulation center," Hall says.
"We can quickly measure that the doctors know and feel much more comfortable and are able to recognize the sick patient and know what to do more quickly," she says of the simulation program.
Mark Bullard, MD, medical director for the Carolinas Simulation Center and an emergency physician explains that the dozen mannequins can mimic a variety of symptoms such as sweating, urinating, and drops in blood pressure. They have varying pupillary response to light, pulse, bowel, breath, and heart sounds, and can withstand placement of IVs.
And they can speak, aided by Bullard, who acknowledges acting like "the man behind the curtain in the Wizard of Oz," vocalizing how a patient with severe sepsis or septic shock might answer a question about the time of day or who is president. He gives, he says, "a good portrayal of an altered, confused state."
Bullard and Hall say the $1 million program has paid off by helping doctors and nurses respond to sepsis in the simulation center, rather than relying on real-time cases to learn the procedure. "It used to be that you saw one, did one, and then taught one. But those days are gone," Bullard says. "It's no longer safe to practice that way. Residents learn better in experiential learning environments."
Alan Heffner, MD, director of the Carolinas Medical Center ICU in Charlotte, says that rather than being seen as an inevitable condition that frequently raises hospital mortality rates, sepsis is increasingly perceived as a condition that requires a rapid response the same as one would expect for a patient with a heart attack or stroke.
"Sepsis ranks up there with the same magnitude, if not higher, in its danger as a public health concern. We've recognized that for years, but now we recognize it better because we're epidemiologically tracking people who die of this disease," Heffner says.
In a 2009 Carolinas study, emergency physicians decided to see whether implementing EGDT protocols for patients with sepsis seen at one large urban emergency department within the Carolinas system would save lives not just for the short-term, but for a one-year period. They tested one group of patients seen during a period before protocols were implemented, and another group of patients seen after implementation.
They tracked death of patients through internal patient records as well as the Social Security Death Index.
They found that 49% of patients treated before the protocol died within one year of their treatment, but after implementation of the protocols, just 37% died within one year; Heffner describes that as "compelling. It meant that patients had a survival benefit from the interventions performed within the first six hours of their emergency department visit that extended one year from the time they were treated."
During the past several years, most sepsis experts agree, the number of cases of sepsis nearly doubled, between 621,000 in the year 2000 to 1,141,000 in 2008, because it's much more rapidly recognized in the emergency department, and some have suggested more likely to be coded correctly for Medicare billing and tracking purposes.
Sepsis experts know that they need to do more to reduce sepsis progression and mortality. And many systems are looking at new strategies wherever they may be.
"I'm happy with what we've done so far," says Dowling. "But you can't be happy until you get to as close to zero as possible. The goal has to be zero mortality from sepsis. And I won't be happy until that's achieved."
Reprint HLR0614-10
This article appears in the June 2014 issue of HealthLeaders magazine.