The frontiers of sepsis care

Physicians in Kansas improve sepsis care at critical access hospitals

Where: Small hospitals in Kansas.

The issue: Improving sepsis care in critical access hospitals.


Years ago, physicians at the University of Kansas Hospital noticed that sepsis patients transferred in from smaller hospitals had significantly higher mortality rates than similar patients who had been directly admitted. When they looked into this issue, they found that sepsis was often occurring unrecognized at the small hospitals, Steven Q. Simpson, MD, FACP, told attendees at CHEST 2016 last October.

“We have 128 hospitals in the state; 83 of them are in towns where they have 25 beds or less in the hospital,” said Dr. Simpson, who is a professor of pulmonary and critical care medicine at the University of Kansas in Kansas City. “Our demographics and our epidemiology suggest that in Kansas we have about 20,000 cases of severe sepsis per year, and half of those cases come from places like this.”

With so many patients at risk of poor outcomes, Dr. Simpson and colleagues developed a program to help the small hospitals improve their sepsis care.

How it works

The Kansas Sepsis Project is a training program for rural clinicians, led by University of Kansas experts. “Our preferred mode is to find one of the larger towns that these critical access hospitals send folks to. We invite critical access hospitals in,” said Dr. Simpson. With area clinicians gathered at one spot, the project leaders conduct a training session about sepsis, including lectures and interactive activities.

“That's our preferred mode, because it's more efficient for us. The more effective mode for the critical access hospitals, it turns out, is to go there individually and speak to the doctors up close and personal and look at charts with them,” Dr. Simpson said, noting that the project now also provides such individualized training.

The project also offers a sepsis protocol, derived from the Surviving Sepsis Campaign, but simplified. “It focuses on the three-hour bundle,” said Dr. Simpson. “It doesn't even say calculate 30 mL per kilo [for fluid resuscitation]. It just says give 2 L of fluid. It says to track what you can to make a triage decision based on serum lactate and/or blood pressure response to fluids, and it says consider placement of a [central venous] catheter if that can be done without transfer.”

To measure improvements in care, the project also provides screening and data-collection tools. “They also have a tracker that keeps track of whether they have done the appropriate things for the patient and if the patient's been triaged and what the outcomes are,” Dr. Simpson said.


According to the collected data, sepsis care at the targeted hospitals has improved. At one hospital, the team began by pulling past charts looking for sepsis cases. They found 28. “None of the cases actually identified sepsis on the chart at the time of admission,” said Dr. Simpson.

After participating in the project, the same hospital identified 36 cases of severe sepsis within six months and showed improvements in markers of good sepsis care. “Blood cultures: from nonexistent to most cases. Antibiotics: from under half to nearly all. And IV fluids: from almost no one to about half,” he said.

More broadly, the project's database has revealed that the hospitals are getting better at recognizing severe sepsis the longer they participate. “Over time you spend in the project, your accuracy goes up and stays fairly high for as far as two years,” said Dr. Simpson.


An initial challenge of the effort was getting the small hospital clinicians to acknowledge the problem. “Many of them say, ‘Oh, we don't have sepsis at our place,’ until we convince them that they do,” Dr. Simpson said.

The hospitals also faced difficulty gathering the needed local data to improve care. (Dr. Simpson noted that critical access hospitals aren't subject to the CMS core measures on sepsis.) They also worried about potential negative impact on hospital revenue if more sepsis cases were identified and patients were then transferred out. But improving initial sepsis care can actually reduce the likelihood of transfer, he said. “They are able to keep people at home and derive the revenue that comes from that.”

Next steps

The project is now working with a commercial vendor to implement an app to analyze and assist with bedside sepsis care. “As you're doing the things you need to do for the patient, it records it and sends you an email instantly that says, ‘This is everything that you've done,’” Dr. Simpson said.

Words of wisdom

“Little hospitals collectively see a lot of sepsis patients. Providers there can be taught to recognize and treat severe sepsis on the spot...if you provide them tools and training and tracking their performance and consistent follow-up and needed education,” he concluded.