Sometimes referred to as blood poisoning, sepsis is the body’s often-deadly response to infection. As many as 1.6 million Americans are affected by sepsis, and more than 258,000 of them die from it, every year. Yet sepsis remains not only misunderstood by the public—a recent study by Harris Poll on behalf of the Sepsis Alliance in June 2014 found that only 44 percent of Americans had even heard of sepsis—it also remains challenging for physicians and other clinicians to work with. There are a variety of reasons that sepsis remains such a huge, unmanaged problem, in hospitals and emergency departments. But there are those who are working to change the situation. And there are things that IT professionals, especially clinical informaticists, can do, to turn things around.
One clinician leader who is in the thick of efforts to educate about sepsis and turn things around in his own organization is Jim O’Brien, M.D., vice president of quality and patient safety at Ohio Health Riverside Hospital, one of 12 hospitals in the Ohio Health system. O’Brien is involved in studying sepsis treatment patterns at Ohio Health Riverside Hospital and its sister hospitals in the Ohio Health system; what’s more, he is also chairman of the board of directors of the Sepsis Alliance. Dr. O’Brien spoke recently with HCI Editor-in-Chief Mark Hagland about his work both at the Ohio Health system, and through the Sepsis Alliance, in this area. Below are excerpts from that interview.
Jim O'Brien, M.D.
Where are we as a healthcare system on the journey towards minimizing sepsis?
My experience is that the majority of CEOs and CMOs think of sepsis from the perspective of being a healthcare-acquired infection, and that’s how it’s most often presented in the press, but the vast majority of patients with sepsis come into our facilities already with sepsis. So it’s very important to understand that. And one of the areas in which we’ve actually made the most progress is around eliminating or preventing sepsis in central line catheter infections, urinary tract infections, and surgical site infections; but in the majority of cases involving sepsis, the sepsis comes in with the patients. And if you went into the average CEO’s office and asked them what their door-to-balloon time was [the measure of the length of time required to get patients presenting with heart attack symptoms into a hospital’s cardiac catheterization lab], they’d know the answer generally. But ask them how long it takes to get antibiotics into someone with septic shock, and they won’t know. Still, the existing data suggests that there’s almost an 8-percent increase per hour in death due to a delay in antibiotic treatment for septic shock. So the length of time to get someone appropriate antibiotics is going to be critical.
In a large multi-system provision of care for septic shock—in a study of over 3,000 patients, the median time to get antibiotics into patients in hospitals, the majority of them presenting in EDs, was six hours. [A study published by the Center for Healthcare Research & Transformation at the University of Michigan noted that “[H]ospitalizations for septicemia rose sharply in both Michigan and in the United States from 2007 to 2011. This suggests an important shift in the disease burden of severe sepsis (one of several diagnoses categorized as septicemia.)]
Six hours is a terribly long time. Why so long?
I’ve got biases and opinions as to why the length of time involved here is so long. The first “why” is that we don’t treat sepsis and septic shock as a medical emergency. And if someone comes in with trauma or a stroke, most hospitals have a process that has been optimized for those issues. But in terms of how hospitals care for sepsis, we just consider sepsis to be a complication, and sepsis is mostly not treated as a medical emergency.
So if someone comes in with a heart attack, they are going to activate the heart attack team and whisk them off to the cath lab, because we know time matters. If someone comes in with septic shock, the doctor will write orders, etc. But the hospitals that treat this as a medical emergency will reduce this from about 25 percent mortality to about 10 percent. And the next question is, why don’t hospitals treat this as a medical emergency? There aren’t great pressures for them to do so. But it’s not part of core measures in HospitalCompare, nor is it an element in the annual U.S. News and World Report report on hospital quality; and there’s not much reporting on this. The Sepsis Alliance funds surveys, and our survey this year found that only 44 percent of Americans even knew of the existence of sepsis.
This makes me think of the sudden emergence of awareness around the Ebola outbreak, actually.
Yes, and what kills people with Ebola? Sepsis. It’s your body’s response to infection. There have been 2,800 people dying of Ebola in the last six months. Meanwhile, we have lost 2,800 Americans to sepsis in the last four days.
That’s a very dramatic statistical comparison. It’s reminds me of the Institute of Medicine’s 1999 report, “To Err Is Human,” that cited the 98,000 annual deaths that are caused by preventable medical errors every year. Some in the mainstream media noted that that figure was equivalent to a jumbo jet crashing every day.
Yes, and this is as if a jumbo jet crashed every day and people didn’t know it; and yet we have the ability to fix it! We believe that increasing public awareness of this is absolutely critical. To put this all into perspective, a mortality rate of 10 percent from septic shock—after considerable work to reduce its current level—would still be four times the rate of mortality from heart attacks. Of people who present with a true heart attack where we’re taking them off to the cath lab, it’s at about that number—about 2.5 percent—and it’s a credit to where we’ve come since the 1960s. The funding for every breast cancer death is about $80,000 per death, but it is only $190 per sepsis death. That’s not to say that breast cancer research is not important; it is. But the funding is so important.
Now, the CDC [the federal Centers for Disease Control and Prevention] is getting more active about this. Now, the number of discharges from hospitals, in that University of Michigan report that just came out, found a 56.8 percent increase in deaths in hospitals from septicemia, from 2007 to 2011. The federal government has reported that septicemia is the single most expensive reason for hospitalization; and the cost for it is $20 billion annually in hospital treatment alone, while the cost for the skilled nursing component has never been quantified.
That really is a huge number.
Yes, that is the part of this for me that creates so much cognitive dissonance. We have therapies that would make this better. But the stories are not being told.
Do physicians see treatment for sepsis as too routine to really pay attention to as a broader issue? For example, a lot of attention has been paid to the wonderful progress being made in creating programs to get patients presenting with stroke symptoms to have tPA [tissue plasminogen activator] administered to them.
Yes, I think there is a shininess factor involved. For example, it’s harder to construct a pro forma around sepsis care versus buying a surgical robot. And if I put out an ad saying we’re going to improve sepsis mortality, and if people don’t know what it is, how effective will that be? Also, what kinds of doctors care for stroke? Neurologists. In contrast, with sepsis, you don’t have a single type of physician involved. And there’s no medical director of sepsis care, so that creates barriers for providing reasons to come up with innovations, so even if you had something shiny, it’s hard to pitch. Eli Lilly developed a drug to care for sepsis, and they had a terrible time marketing it, and I saw them go through different strategies, and they never could get it right. It was withdrawn as a business decision.
But it was clinically effective?
We don’t know. There was one study that provided benefit for kids, not so much for adults, and in the follow-up study, it didn’t help adults that much.
In terms of measuring and optimizing sepsis treatment, is there a role for IT people, particularly clinical informaticists, in conjunction with medical directors?
I think informaticists can help supply data—if a medical director says, I think we have a problem with sepsis, and informaticists can help frame questions and analyze results. The return on investment is largely cost reduction. We recognized Intermountain Healthcare as a hero organization for its sepsis reduction this year. But the majority of hospitals are still reimbursed based on volume, not value. And if I develop a program such that people never develop septic shock, they get paid less. And indeed, I get paid more if people develop septic shock and need additional care as a result. So strategically, it’s a more difficult challenge to make.
Clinician leaders need to step up to the plate. But in a survey recently, only one of every 20 physicians could define sepsis. And only one out of five of us ICU physicians could define sepsis. And I blame the definition. It’s too vague. We are defining sepsis, similarly to if we just told them they had cancer with no further definition or information. We say someone who’s 18 with influenza and respiratory failure has sepsis; and we say that an 80-year-old with a urinary tract infection with kidney failure, also has sepsis. But if I put those patients next to each other, and asked, do they have the same thing? They’d say no. So the definition is based on symptoms.
Some physicians think you need a positive blood culture to determine sepsis; that’s true in less than one-third of cases. And here, when patients came into the ICU [intensive care unit] with sepsis, we would take blood samples. And we would always check with the clinical team, and they would say, they have pneumonia with respiratory failure and shock, not sepsis—but they had sepsis! More often, we describe it as a complication from an underlying disease. And in the press, it’s described as, “a celebrity died of complications from pneumonia or from cancer.”
What are you doing at Ohio Health Riverside Methodist in this area?
We started by focusing on patients presenting to our ED with septic shock—low blood pressure due to sepsis. And we started there because since there’s no single sign, symptom or test involved. But with patients who come in with low blood pressure, it’s most likely they have sepsis, unless, for example, they’re bleeding because they were shot. So we wanted to start with those patients who were the sickest, because we thought there would be the least confusion about this.
So we’ve put in place a more systematic screening process for patients presenting. And we’ve developed a trigger to get the physician working in the ED to see that patient. And we’re trying to emphasize very simple standards of care—checking the lactate level, which determines whether the patient is getting enough oxygen in their tissues; obtaining blood cultures; getting them antibiotics; and giving them intravenous fluids. So we’ve just been focusing on those four things, to try to get those right. What we’ve seen is a gradual increase in the quality of care provided and a gradual decrease in mortality. We have some ways to go, but we’ve made progress.
What is the timeframe on this activity?
I’ve been in my current role for two years, and it took one year of convincing people that this was a problem, and that it was a problem they could and should do something about.
So this process has been place about a year?
Yes. And mortality on these patients has come down from 35 percent to 20 percent. Other hospitals have shown they’ve gotten mortality down to 10 percent. We started with a very select group of patients.
For clinical clarity?
Correct. And Riverside is only one hospital in Ohio Health. That result is across all 12 EDs in our 12 hospitals.
How low do you think you can bring it?
The literature suggests that we should be able to bring it down below 10 percent. And we need to expand our population, too. There are more patients than the types in our focus study. But this is at least showing proof of concept that’s gotten the physicians to believe that we can make dramatic improvements. We’ve also been doing a lot of work around central line infections and urinary tract catheter-related infections. And a lot of the results in that latter area have been having the ICU nurse asking the doctor whether the patient really needs a urinary catheter or not. And it’s the same principle around central-line catheters. We need to be intentional about it. And we’ve done a lot of work around increasing hand hygiene in the hospital. And that continues to be something that’s really hard to accomplish, but we’re hitting about 90 percent in terms of clinicians doing the right thing. We need to get even higher, but it’s a significant improvement.
And that would affect sepsis for sure, right?
Absolutely. If we can get everyone to wash hands consistently, it will decrease the sepsis that occurs in our hospitals.