AI model may help differentiate between MIS-C, Kawasaki disease
Click Here to Manage Email Alerts
An artificial intelligence tool called KIDMATCH may be able to accurately distinguish between MIS-C, Kawasaki disease and other similar febrile diseases, according to data published in the Lancet Digital Health.
“For 40 years, the Kawasaki research community has tried to create a diagnostic test for [ Kawasaki disease (KD)] and failed,” Jane C. Burns, MD, a pediatrician at Rady Children’s Hospital-San Diego, director of the Kawasaki Disease Research Center at the University of California San Diego School of Medicine, and study co-author, said in a press release.
Meanwhile, deep-learning algorithms have demonstrated “high performance capabilities in industrial applications, such as voice recognition or machine translation,” and are adept at identifying multiplicative risk factors, added Shamim Nemati, PhD, associate professor of medicine at UC San Diego School of Medicine and study co-author, in the release.
To develop and internally validate the Kawasaki Disease vs. Multisystem Inflammatory Syndrome in Children — or KIDMATCH — model, the Burns, Nemati and colleagues enrolled patients diagnosed with MIS-C from three hospitals between May 7, 2020, and July 20, 2021. Patients diagnosed with Kawasaki disease between Jan. 1, 2009, and Dec. 31, 2019, were also enrolled to increase the generalizability of the results.
The model was designed in two stages. In the first, the model was trained to differentiate between MIS-C and other pediatric febrile conditions due to the need to prioritize patients with the disease during the admission process, the researchers wrote. Stage two then focused on training the machine to categorize patients into other categories, such as Kawasaki disease or “other febrile illness.”
The performance of the model was evaluated using area under the receiver operating characteristic curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), all calculated to 95% sensitivity for MIS-C and Kawasaki disease, for stage 1 and 2 training, respectively.
A total of 1,583 patients were enrolled for internal validation of the model. The model achieved a median AUC of 98.8% during internal validation in the first stage, and 96% in the second, according to the researchers. During external validation testing, KIDMATCH correctly identified MIS-C in 76 of 81 patients (94%), in one group. In other groups, KIDMATCH demonstrated 96% and 90% accuracy, the researchers wrote.
“In just the space of 18 months, we have created a physician support tool that differentiates MIS-C from KD in children using simple test results and five features of the physical exam that any health care provider, clinic or hospital can do, with accuracy exceeding 90%,” Burns said in the press release.
Nemati added that the tool may potentially lead to earlier treatment, which could prevent severe outcomes.
“The hope here, of course, is that KIDMATCH can help front-line clinicians distinguish between MIS-C, Kawasaki disease and other febrile illnesses so that they can provide earlier, appropriate treatment and prevent severe complications,” Nemati said.
References:
How to Tell the Difference between Kawasaki Disease and MIS-C. https://www.newswise.com/articles/how-to-tell-the-difference-between-kawasaki-disease-and-mis-c?sc=dwhr&xy=10007702. Sept. 21, 2022. Accessed Oct. 12, 2022.
Collapse
Click Here to Manage Email Alerts