[ad_1]
Complication of it is an infection known as sepsis the first killer In US hospitals. So it’s no surprise that more than 100 healthcare systems use the early warning system offered by Epic Systems (the leading provider of electronic health records in the US). The system generates alerts based on a proprietary formula that continuously examines signs of the condition in the patient’s test results.
A new study using data from nearly 30,000 patients at University of Michigan hospitals suggests that Epic’s system is not working well. The authors say they lost two-thirds of sepsis cases, rarely found doctors in cases where staff did not, and often gave false alarms.
Karandeep Singh, an assistant professor at the University of Michigan who led the study, says the findings show a broader problem for the owner. algorithms it is increasingly used in health care. “They are widely used, and yet very little has been published on these models,” says Singh. “It’s amazing to me.”
The study was published on Monday JAMA Internal Medicine. A spokesman for Epic discussed the findings of the investigation, saying the company’s system had “helped clinicians save thousands of lives.”
Epic is not the first health algorithm that raises concerns that technology does not provide for health care or is not actively harmful. In 2019, a system used to prioritize access to special care for people with complex needs was discovered needs of patients below compared to white patients. That asked some Democratic senators to require federal regulators to investigate bias in health algorithms. A examination published in April found that the statistical models used to predict the risk of suicide in mental health patients favored white and Asian patients well, but black patients badly.
The way it treats septic hospital rooms has become a special target of algorithmic support for medical staff. Instructions They encourage health care providers from the Centers for Disease Control and Prevention to the use of electronic medical records for surveillance and predictions. Epic has a number of competitors that offer commercial warning systems, as well as a number of US research hospitals. they built their tools.
Automated sepsis warnings have great potential, Singh says, as key symptoms of the condition, such as high blood pressure, can have other causes, making it difficult for staff to detect early. Treatment for sepsis, such as antibiotics, should begin an hour earlier great influence for the survival of the patient. Hospital administrators often take a special interest in the response to sepsis, mainly because it helps U.S. government hospital assessments.
Singh runs a lab researching applications in Michigan machine learning patient care. A machine was asked to chair a committee in the university health system set up to oversee learning uses about Epic’s sepsis warning system.
As Singh learned more about the tools used in Michigan and other health care systems, he was mostly concerned that the vendors were coming, knowing how they worked, or how they worked. His system was licensed to use Epic’s sepsis prediction model, and the company told customers it was very accurate. But there was no independent validation of his performance.
Colleagues from Singh and Michigan tested Epic’s prediction model in nearly 30,000 patient records, covering nearly 40,000 hospitalizations in 2018 and 2019. The researchers noted how many of the people who developed sepsis, defined by the CDC and the Centers for Medicare and Medicaid Services, described Epic’s algorithms. And they compared the alerts that the system would trigger with registered sepsis treatments by staff when they did not see epic sepsis alerts for patients included in the study.
The researchers said their results suggest that Epic’s system would not be much better for the hospital to catch sepsis and that staff can load unnecessary alerts. The company’s algorithms did not identify approximately two-thirds of the 2,500 sepsis cases in Michigan data. It would alert 183 patients who developed sepsis but were not given timely treatment by staff.
[ad_2]
Source link
