KIDMATCH: A Tool for Diagnosing MIS-C and Kawasaki Disease
KIDMATCH is a clinical decision support tool developed to help doctors distinguish between Multisystem Inflammatory Syndrome in Children (MIS-C), Kawasaki Disease (KD), and other febrile illnesses. MIS-C is a new condition identified during the COVID-19 pandemic that can lead to severe health issues in children, similar to KD. Since neither MIS-C nor KD has a definitive lab test, accurate diagnosis relies heavily on clinical judgment.
How KidMatch Works
KIDMATCH uses an artificial intelligence (AI) model that considers a child's age, specific symptoms, and results from common laboratory tests. These include:
*White blood cell count
*C-reactive protein (CRP)
*Erythrocyte sedimentation rate (ESR)
*Gamma-glutamyl transferase (GGT)
The tool processes this information to predict whether a child has MIS-C, KD, or another febrile illness. If the tool encounters an unfamiliar case, it flags it as indeterminate, prompting further review by clinicians.
Clinical Examples
Case 1
A 3-year-old with fever, vomiting, and a history of MIS-C in the family was evaluated. Although the initial symptoms were nonspecific, KIDMATCH helped confirm the child had a febrile illness rather than MIS-C, allowing for a safe discharge.
Case 2
A 2-year-old with fever and a neck mass was suspected of having KD. KIDMATCH identified the condition as a bacterial infection, guiding the doctor to treat with antibiotics instead of for KD.
Case 3
A child with fever and symptoms not initially suggestive of MIS-C had concerning lab results. KIDMATCH, used retrospectively, could have flagged MIS-C earlier, potentially preventing the progression to severe cardiac issues.
Feedback and Refinement
Clinicians provided feedback on KIDMATCH’s ease of use and effectiveness. Most found it user-friendly but suggested better integration with electronic health records for smoother access. The tool’s AI-driven analysis was praised for its ability to simplify complex clinical data into actionable insights.
Participants and Outcomes
Trained on data from 1,517 patients (including 135 MIS-C cases).
Used by four experienced clinicians in real-time scenarios.
Demonstrated the potential to improve diagnostic accuracy and patient outcomes in pediatric care.
Conclusion
KIDMATCH represents a significant advancement in the early detection and differentiation of serious pediatric conditions like MIS-C and KD, supporting healthcare providers in making timely and accurate diagnoses.
**STUDY POPULATION :
The KIDMATCH clinical decision support tool was developed and trained using data from a diverse cohort of 1,517 pediatric patients evaluated at Rady Children’s Hospital. This cohort included 69 patients diagnosed with Multisystem Inflammatory Syndrome in Children (MIS-C), 775 patients with Kawasaki Disease (KD), and 673 patients with other febrile illnesses (FC). To enhance the tool's robustness, additional MIS-C patient data were incorporated from Children’s Hospital of Los Angeles (50 patients) and Connecticut Children’s Hospital (16 patients), bringing the total to 135 MIS-C patients. All MIS-C patients were enrolled between May 14, 2020, and June 18, 2021, and were defined according to the CDC case definition. Notably, none of the MIS-C patients had received a SARS-CoV-2 vaccine, but all tested positive for SARS-CoV-2 antibodies. KD patients, enrolled from 2009-2019, met the American Heart Association's criteria for either complete or incomplete KD and were diagnosed by two experienced KD clinicians to avoid misclassification. The FC patients, also enrolled from 2009-2019, were previously healthy children with fever for at least three days plus at least one clinical criterion for KD. The final FC diagnoses were adjudicated by two pediatric clinicians based on comprehensive medical record reviews and test results. Written consent or assent was obtained from parents and subjects, and the study received approval from the Institutional Review Boards (IRBs) of the University of California San Diego, Children’s Hospital Los Angeles, and Connecticut Children’s Hospital.