Dartmouth-Hitchcock Experts Target Sepsis

Using knowledge gained from a national network of medical experts, Dartmouth-Hitchcock (D-H) has dramatically improved the outcomes of patients with sepsis by focusing on early detection and rapidly delivering recommended care – doubling the chance of survival.

Sepsis is the leading cause of deaths in U.S. hospitals. It affects nearly one million patients and costs an estimated $20 billion to treat annually. Beyond being deadly, sepsis is notoriously difficult to detect and treat. It often mimics other illnesses, such as the flu. Defined as a whole-body inflammation caused by infection, sepsis can progress very quickly, setting off a reaction that can lead to shock, organ failure and death.

Dartmouth-Hitchcock – which treats the sickest patients in the region and is in the top 4 percent nationally for high acuity patients – was concerned with a statistically high mortality rate from sepsis identified through national research. The challenge: how to more effectively identify one of the most time-sensitive, hard-to-detect illnesses in medicine, manage the condition, and save more patients. The solution was discovered through a new information-sharing collaborative, where data and best practices drove better outcomes.

"The answer involved being transparent about the problem, seeing our own performance clearly, learning from others, implementing change, and spreading that learning out to other systems," says Dartmouth-Hitchcock CEO and president Dr. James N. Weinstein.

Dartmouth-Hitchcock has long been at the forefront of using big data to study medical mysteries and determine the best outcomes for patients. The desire to broadly share medical knowledge and improve patient outcomes led Weinstein to become a founding member – with Mayo Clinic, Intermountain Health, and Denver Health – of the High Value Healthcare Collaborative (HVHC), a partnership of 17 health systems across the country striving to improve the quality of care while lowering costs.

With a $27 million grant from the Centers for Medicare and Medicaid Innovation, HVHC now has more than 70,000 physicians, treating more than 70 million patients and sharing best practices and data. Sepsis was one of the targets of that grant and volumes of data collected by HVHC's partners provided D-H clinicians with a clear vision and the practical steps for improvement.

D-H created a multi-disciplinary team to focus solely on the sepsis problem, says Dr. Andreas Taenzer, a Dartmouth-Hitchcock anesthesiologist who led the effort.

"It was very clear to us where we needed to go," Taenzer says, "but the challenge to us was to do it in a very rapid timeframe." Taenzer and his colleagues crunched the HVHC data, visited other hospitals and returned with a plan. Most hospitals have spent years targeting sepsis; D-H set a goal of changing outcomes within 90 days.

The team applied an industrial-scale rapid improvement process – used by business giants GE, Toyota and Motorola – "to define, measure and analyze the problem fairly quickly and get into improvement and control, which is where a lot of the heavy lifting in this project was," says Samuel Shields, D-H's director of performance improvement.

Their aggressive goal was to improve the reliability of what is known as the sepsis three-hour bundle: performing blood tests, pushing fluids, and starting broad-spectrum antibiotics within three hours of arrival. The HVHC data showed delivering every one of those steps consistently and quickly saved a majority of the sickest patients.

The bottom line: D-H's team had no more than three hours to keep a sepsis patient alive. A one-hour delay in treatment could mean a significant increase in mortality.

But who were the sickest patients? An important tool to help identify sepsis patients is the schedule of what is called SIRS (systemic inflammatory response syndrome) triage criteria. Sepsis patients show the symptoms of SIRS, but the trouble is that patients with common ailments such as strep throat also meet SIRS criteria. Tweaking the criteria numbers a bit, the D-H sepsis team developed a schedule of "Super SIRS" triage criteria more specific to sepsis.

"We knew this would give us more bang for the buck," says Dr. Patricia Lanter, director of quality in D-H's Emergency Department. The Super SIRS criteria were printed on hot-pink fliers that were posted throughout the Emergency Department where physicians and nurses discuss the day's cases. Having a multidisciplinary team of nurses, doctors, lab technicians and pharmacists made the process work.

The team also developed an Emergency Department scorecard for sepsis, focusing on the delivery of the three-hour bundle elements. By reviewing the scorecard data, changes were made in the lab, in timing of fluids and, most importantly, antibiotic delivery. Scorecard results were quickly communicated to nurses and physicians.

They learned that these small steps – done sequentially, every time and for every patient – proved to be the key to success.

The results speak for themselves. Since May, D-H has consistently delivered the three-hour-sepsis bundle to more than 80 percent of patients, surpassing published results by more than twofold and matching the HVHC's highest performing hospitals. Dartmouth-Hitchcock improved sepsis intervention dramatically and reduced patient mortality by more than 50 percent.

Following its success in the Emergency Department, the team is now working throughout the hospital's inpatient units to improve the detection and treatment of sepsis across all general care units.

"The final component will be to roll the effort out to all of the hospitals and care centers in the region," Taenzer says. "To improve the outcomes of all patients across the state, we need to work with other providers to intervene earlier. We're not there yet, and it's going to take a lot of work to sustain this effort, but we've come an astonishingly long way in a very brief period of time."

Source: Dartmouth-Hitchcock

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish