University of Pittsburgh School of Medicine researchers studied EHR data and developed an algorithm that identifies and assigns patients into four separate types of sepsis, according to a May 19 study published in JAMA.
Sepsis, a life-threatening condition that occurs when a patient's bodily response to an infection harms its own tissues and organs, has traditionally been treated with a "one size fits all," approach, said Christopher Seymour, MD, lead study author and member of University of Pittsburgh's Clinical Research, Investigation, and Systems Modeling of Acute Illness Center, according to a news release.
"These protocols ignore that sepsis patients are not all the same… Hopefully, by seeing sepsis as several distinct conditions with varying clinical characteristics, we can discover and test therapies precisely tailored to the type of sepsis each patient has."
For the study, which was funded by the National Institutes of Health, Dr. Seymour and his research team used computer algorithms to analyze 29 clinical variables reported in the EHRs of more than 20,000 UPMC patients. The patients were diagnosed with sepsis within six hours of arriving at the hospital from 2010 to 2012.
The algorithm organized patients into four distinct sepsis types:
· Alpha: 33 percent of patients; individuals had the fewest abnormal lab results, least organ dysfunction and lowest in-hospital death rate.
· Beta: 27 percent of patients; older individuals with most chronic illnesses and kidney dysfunction.
· Gamma: patients with similar frequency as Beta individuals, but with increased inflammation and primarily pulmonary dysfunction.
· Delta: 13 percent of patients; most deadly type, often with liver dysfunction and shock and the highest in-hospital death rate.
Researchers found the algorithm to be successful on additional tests of EHRs from 43,000 UPMC sepsis patients as well as clinical data from almost 500 pneumonia patients at 28 U.S. hospitals. The research team then tested their findings on numerous recently completed international clinical trials that analyzed different therapies for sepsis. The team discovered that when participants in these trials were separated by the four sepsis types, some trials that previously failed ended up succeeding.
The study authors concluded that the next steps in sepsis therapy are to target treatments that apply to the specific types of sepsis and design new clinical trials to test their effectiveness.