Risk Prediction Model Could Identify Pediatric MIS-C Cases


A team from Vanderbilt identified 4 key predictors from a retrospective analysis of cases since 2020.

Risk Prediction Model Could Identify Pediatric MIS-C Cases

Matthew Clark, MD, RhMSUS

A clinical diagnostic prediction model may be able to help clinicians distinguish multi-inflammatory syndrome in children (MIS-C).

In new late-breaking data presented at the American College of Rheumatology (ACR) 2021 Convergence, a team of Vanderbilt University and Johns Hopkins investigators highlighted a risk prediction model that uses clinical, laboratory and cardiac feature to help identify MIS-C cases within 1 day of observed symptoms.

The prediction model may be valuable to clinicians needing to quick and accurately identify rare MIS-C cases during the COVID-19 pandemic.

Investigators, led by Matthew Clark, MD, RhMSUS, Rheumatology Fellow at Vanderbilt, sought to develop a multifactorial MIS-C diagnostic prediction model that would use signs and symptoms observed in the first 24 hours of a possible patient. According to the Centers for Disease Control and Prevention (CDC), there have been approximately 5500 MIS-C cases in the US since May 2020; in Clark and colleagues’ state of Tennessee, cases have been in the higher percentile, at an estimated 200-249.

“MIS-C is a rare consequence of SARS-CoV-2,” they wrote. “MIS-C shares features with common infectious and inflammatory syndromes, and differentiation early in the disease course can be difficult.”

Investigators used data from a retrospective chart review of children ≤20 years old who were admitted to Vanderbilt Children’s Hospital and evaluated for MIS-C from June 10, 2020 to April 8, 2021. Clinicians collected standardized clinical, lab, and cardiac characteristics of pediatric patients at each presentation in the hospital. MIS-C was clinically diagnosed by the child’s primary service, retrospectively reviewed and confirmed by a pediatric rheumatologist and pediatric infectious disease physician.

Investigators selected candidate MIS-C predictors a priori and examined for collinearity Spearman correlations. They used logistic regression to identify the most important predictors for the rare disease, with variables selected in ≥80% of 500 bootstraps included in their final predictor model.

Their trial population included 127 children admitted to the Vanderbilt hospital with concern and evaluation for MIS-C during the observed time period. They selected statistically distinct variables that identified 45 cases of MIS-C versus 82 non-cases to build the risk preidciton model.

Eventually, Clark and colleagues included 4 predictors in the final model:

  • Hypotension
  • Abdominal pain
  • Rash
  • Serum sodium

Hypotension was defined as required fluid resuscitation, vasopressor support, or blood pressure under the 10th percentile for age, height, and sex.

The prediction model showed “excellent” discrimination, with a c-index of 0.90 (95% CI, 0.85 – 0.94) as well as good calibration.

“We used early clinical and laboratory features to inform the design of a clinical diagnostic prediction model with excellent discrimination to assist clinicians in distinguishing patients with MIS-C from those without,” investigators concluded. “We plan to test this model with external and prospective validation.”

The study, “A Prediction Model to Distinguish Patients with Multisystem Inflammatory Syndrome in Children,” was presented at ACR 2021.

Recent Videos
Signs and Symptoms of Connective Tissue Disease
Connective Tissue Disease Brings Dermatology & Rheumatology Together
© 2024 MJH Life Sciences

All rights reserved.