New, transparent AI tool may help detect septicemia

New, transparent AI tool may help detect septicemia

Ten years ago, 12-year-old Rory Staunton got a ball in gym class and scraped his arm. He woke up the next day with a fever of 104 degrees Fahrenheit, so his parents took him to the pediatrician and eventually to the emergency room. They were told it was just a stomach flu. Three days later, Rory died of sepsis after scraping bacteria infiltrated his blood and caused an organ to fail.

“How does that happen in a modern society?” His father, Kieran Staunton, said in a recent interview with Undark.

Every year in the United States, sepsis kills More than a quarter of a million People – more than a stroke, diabetes or lung cancer. One reason for all this carnage is that sepsis is poorly understood and, if not caught in time, is essentially a death sentence. As a result, much of the research has focused on getting sepsis early, but the complexity of the disease has plagued current clinical support systems—electronic tools that use pop-up alerts to improve patient care—with low accuracy and high rates of false alarm.

That may change soon. Back in July, Johns Hopkins researchers published three studies in temper nature medicine And the digital medicine npjIt displays an early warning system that uses artificial intelligence. The system detected 82 percent of sepsis cases and reduced deaths by about 20 percent. While artificial intelligence — in this case, machine learning — has long promised to improve healthcare, most studies demonstrating its benefits have been conducted on historical data sets. Sources told Undark that, to their knowledge, when used on patients in real time, no AI algorithm has shown widespread success. Sochi Sariya, director of the Machine Learning and Healthcare Laboratory at Johns Hopkins University and senior author of the studies, He said What’s new in this research is how “artificial intelligence is applied to the bed, which is used by thousands of providers, and where we see lives being saved.”

The Targeted Real-Time Early Warning System, or TREWS, scans hospitals’ electronic health records — digital copies of patients’ medical histories — to identify clinical signs that predict sepsis, alert providers about patients at risk, and facilitate early treatment. Leveraging massive amounts of data, TREWS provides real-time patient insights and a unique level of transparency into their causes, according to study co-author and Johns Hopkins University internal medicine physician Albert Wu.

The system also offers a glimpse into a new era of medical electronics, Wu said. Where Introduction to the SixtiesHe added that electronic health records have reshaped how doctors document clinical information, but after decades, these systems have become an “electronic diary.” With a series of machine learning projects on the horizon, both from Johns Hopkins and other groups, Sarria said using electronic records in new ways could transform health care delivery, providing clinicians with an extra set of eyes and ears — and help them make better decisions.

It’s an attractive vision, but one in which Sariya, as CEO of the company developing TREWS, has a financial stake. This insight also reduces the difficulties of implementing any new medical technology: providers may be reluctant to trust machine learning tools, and these systems may not perform well outside of controlled research settings. Electronic health records also come with many Existing problemsfrom burying providers under administrative work to risking patient safety due to software glitches.

Sariya is optimistic though. “The technology is there, the data is there,” she said. “We really need high-quality care-enhancing tools that allow caregivers to do more with less.”

Currently, there is There is no single test For sepsis, health care providers therefore have to compile their diagnoses by reviewing the patient’s medical history, performing a physical examination, performing tests, and relying on their clinical impressions. Given this complexity, Over the past decade Doctors have increasingly relied on electronic health records to help diagnose sepsis, Mostly by recruitment rule-based standards-If this is then.

One example, known as the SIRS criteria, says a patient is at risk of sepsis if two of the four clinical signs — body temperature, heart rate, respiratory rate, and white blood cell count — are abnormal. This breadth, while useful in recognizing the different ways sepsis may present itself, leads to a myriad of false positives. Take a patient with a broken arm. “The computerized system might say, ‘Look, fast heart rate, fast breathing.'” “It might sound a warning,” said Cyrus Sharia, MD, an intensive care unit physician at Washington Hospital in California. It’s almost certain a patient doesn’t have sepsis, but nonetheless it might sound the alarm.

These alerts also appear on service providers’ computer screens as a pop-up, forcing them to stop whatever they’re doing to respond. Therefore, despite these rules-based systems sometimes Reduce mortalityThere is danger alert fatigue, as health care workers begin to ignore the flood of annoying reminders. According to Michael Chabot, a trauma surgeon and former clinical officer at Memorial Hermann Health System, “It’s like a fire alarm going off all the time. You tend to be sensitive. You don’t pay attention to it.”

Indeed, electronic records are not particularly common among clinicians. In a 2018 survey, 71% of doctors He said the records contribute significantly to fatigue and 69 percent that they take valuable time away from patients. Another 2016 study found that physicians should devote every hour they spend caring for patients 2 more hours To electronic health records and office work. James Adams, chief of emergency medicine at Northwestern University, described electronic health records as a “crowded swamp of information.”

Leave a Comment