- May 13, 2022
This tool could potentially serve as an early warning system for doctors by enabling them to rank patients based on their risk of serious disease and assist accordingly, the researchers said.
Blood infections are one of the leading causes of morbidity and mortality worldwide. In the study, published in the journal Scientific Reports, the researchers explained, “Bloodstream infections (BSI) can lead to prolonged hospital stays, and life-threatening and aggressive complications, in addition to high costs to the health care system.” Infection with a bacterium or fungus in the blood system can occur during surgery, or as the result of complications from other infections, such as pneumonia or meningitis. The body’s immunological response to the infection can cause sepsis or shock, dangerous conditions that have high mortality rates. For example, sepsis, the body’s extreme response to infection, is said to have caused roughly 20 percent of annual global deaths in 2017, according to the World Health Organization (WHO) and it can kill in a matter of hours, if not treated immediately.
“Therefore, timely and critical assessment of available microbiology results are necessary to ensure that individuals with BSI receive prompt, effective, and targeted treatment, for optimal clinical outcomes. However, the current standard of care, which mostly depends on blood culture-based diagnosis, is often extremely slow,” the team said. The researchers – students Yazeed Zoabi and Dan Lahav from the laboratory of Prof. Noam Shomron from Tel Aviv University’s Sackler Faculty of Medicine, in collaboration with other researchers, including Dr. Ahuva Weiss Meilik, head of the I-Medata AI Center at Ichilov Hospital – developed an AI tool studies the electronic medical records of patients who were found to be positive for blood infections. These records included demographic data, blood test results, medical history, and diagnosis.
After studying each patient’s data and medical history, the program was able to automatically identify risk factors. “The physicians would input the clinical information, which was collected as routine upon hospitalization, and the output of the algorithm would be a risk factor. This risk factor would be evaluated together with the physician’s advice and would be used for taking action in saving lives,” Prof. Shomron tells NoCamels. “We entered the medical files into software based on artificial intelligence,” he said in a statement, “We wanted to see if the AI would identify patterns of information in the files that would allow us to automatically predict which patients would develop serious illness, or even death, as a result of the infection.”
The team worked with the medical files of about 8,000 Ichilov patients who were found to be positive for blood infections between the years 2014 and 2020, during their hospitalization and up to 30 days after, whether the patient died or not, Prof. Shomron said. Prof. Shomron says the AI program reached an accuracy level of 82 percent in predicting the course of the disease, even when it ignored obvious factors such as the age of the patient and the patient’s hospitalizations. After the researcher entered the data of the patient, the algorithm knew how to predict the course of the disease.
This suggests that it will be possible to rank patients in terms of the danger posed to their health, ahead of time, according to the statement. “Using artificial intelligence, the algorithm was able to find patterns that surprised us, parameters in the blood that we hadn’t even thought about taking into account,” said Prof. Shomron.
“Albumin and creatine [levels] in the blood are strong markers of potential mortality due to infection. Also monocytes, platelets volume. In fact, a few of these were indicators of infection severity; however, the quantitation and the combination of them all in one algorithm were not performed,” he tells NoCamels.
The team is “now working with medical staff to understand how this information can be used to rank patients in terms of the severity of the infection,” Prof. Shomron says. “We can use the software to help doctors detect the patients who are at maximum risk,” he adds. He reiterates that while the accuracy level was important, the AI program is “a tool to assist the clinical team and call their attention to relevant or high-risk cases.”
The system will soon go live at Tel Aviv Sourasky Medical Center – Ichilov. This is a large project at Ichilov, Prof. Shomron says. “We are testing it there in the coming year,” he explains, “and then moving to suggest it to other centers.” He’d also like to use the program to test other conditions, diseases, and illnesses. “I’d like to eventually be able to extend it to other types of infections, not only bloodstream. I’d like to segregate it to types of bacteria (gram-positive/negative) and to test various algorithmic models to improve the decision taken,” he says. “Given the limited resources at the hospital, if one can segregate the high risk from the low-risk patients, attention can be focused on the ones that need it most.”
REFERENCE: No Camels; 22 DEC 2021; No Camels Team