Last week, the New York Times covered the story of a 12-year-old boy whose death this spring sheds light on one of the most time-sensitive and tricky-to-detect illnesses in medicine. The pre-teen died of severe sepsis in New York City apparently missed his infection before it spiraled out of control. The boy was seen by his pediatrician and in an emergency room, but clinicians thought he had a stomach bug, treated him for dehydration, and sent him home. Three days later, he died, after bacteria that snuck into his body through a cut on his arm – a small battle wound from a gym class basketball game – prompted a massive inflammatory response that caused his circulation and heartbeat to falter and his organs to shut down.
Sepsis researcher David Gaieski, MD, an assistant professor of Emergency Medicine in the Perelman School of Medicine and clinical director in Penn’s Center for Resuscitation Science, spoke this week with several news outlets about the issues raised by the case. ABCNews.com explored the reasons why these infections can be so difficult to identify when they’re easiest to treat:
The problem is that, in its early stages, sepsis causes symptoms that aren't much different from those of a viral infection that will go away on its own. Consequently, sorting out who can go home from the emergency department and who needs quick hospitalization can be tough.
"The thing that people are always looking for is the crystal ball that will tell whether this kid who doesn't look all that sick really is," says Dr. David Gaieski, assistant professor of emergency medicine at the Hospital of the University of Pennsylvania. "Right now there's nothing that is 100 percent able to do that."
Nor are there any drugs specifically approved to treat sepsis – Xigris, which was developed for this use, was pulled from the market last fall amid concerns it wasn’t actually effective. A new analysis published this week showed that the drug reduced patients' risk of dying by 18 percent, even though, as ABCNews.com noted, doctors never fully identified the type of patient who would benefit the most from Xigris, and the new study suggests that perhaps only the sickest patients are likely to improve after taking the drug.
Cliff Deutschman, MD, MS, a professor of Anesthesiology and Critical Care and director of the Stavropoulos Sepsis Research Program, told ABCNews he was never really convinced that the drug was better than the standard methods of treating patients with antibiotics and fluids -- but now that it’s not available anymore, the question will never be answered. He also said he fears the Xigris story will have a chilling effect on further sepsis drug development efforts, making pharmaceutical companies unwilling to invest in this type of lifesaving research.
On early detection front, efforts in the Hospital of the University of Pennsylvania’s emergency department have focused on early measurement of serum lactate – which can help indicate whether enough oxygen is being delivered to tissues in the body -- as a marker of impending shock. The results are used to identify sick patients more quickly and then deliver a bundle of care known as early goal-directed therapy during their first six hours in the hospital. Since these new protocols were adopted in 2005, Gaieski told NBC News, deaths among severe sepsis patients admitted to HUP through the ED dropped from 24 percent in 2005 to 11 percent in 2009.
But at the national level, answers about the true toll of sepsis remain elusive.
During the Society for Academic Emergency Medicine annual meeting in May, Gaieski presented research showing wide variability in sepsis incidence. Four different methods of counting annual sepsis cases from a national data sample revealed four different incidence estimates, ranging from 894,013 to more than 3.1 million. Despite those differences, one pattern emerged: Sepsis is on the rise. During the six years analyzed, his team found an increase in cases – an average jump of about 13 percent each year – no matter which counting method they used. Mortality rates ranged from 15 to 30 percent depending on which data abstraction methods was used, but overall, there was a bright spot: The authors observed a decrease in the percentage of sepsis patients who died during their hospital stay during the six years studied, falling from 35 percent to 26 percent.
“We believe that greater awareness of the dangers of sepsis is helping clinicians to take a closer look at patients they might once have discharged and start treatment right away, but a lot of work still remains to be done, including more accurate methods of identifying at-risk patients early, tailoring of therapy to an individual’s specific immune response to an infectious insult, and better methods of objectively measuring the adequacy of resuscitation,” Gaieski says. “In addition, it’s very difficult to measure improvement of a problem we don’t at present accurately count, so it will be important to develop a uniform, consistent definition of severe sepsis that is used nationally. Then we can begin to properly assess the effectiveness of new strategies used to treat it – and put them to work saving more people.”