Evidence-based diagnostic prediction score for pediatric NMDA receptor encephalitis.

Area of Science:

• Neurology
• Pediatrics
• Immunology

Background:

• Early diagnosis of anti-N-methyl-D-aspartate receptor encephalitis (NMDARE) is critical for favorable prognosis.
• Current diagnostic criteria for NMDARE are less effective in children compared to adults.
• Identifying causative autoantibodies in pediatric NMDARE presents diagnostic challenges.

Purpose of the Study:

• To develop and validate a novel diagnostic prediction score specifically for pediatric NMDARE.
• To improve the accuracy of diagnosing NMDARE in children using clinical and laboratory data.
• To facilitate earlier therapeutic interventions in pediatric autoimmune encephalitis cases.

Main Methods:

• Retrospective analysis of pediatric patients (0-18 years) with suspected autoimmune encephalitis from January 2015 to March 2023.
• Identification of significant clinical symptoms and laboratory findings through univariate and multivariate logistic regression.
• Development and validation of a scoring system using five-fold cross-validation and logistic regression analysis.

Main Results:

• A pediatric NMDARE prediction score was developed using 15 significantly different variables (p < 0.05).
• The score demonstrated 82.1% sensitivity and 82.2% specificity with an 8-point cutoff (AUC=0.888).
• Five-fold cross-validation yielded 95.6% sensitivity, 71.4% specificity, and a kappa coefficient of 0.670.

Conclusions:

• A novel, evidence-based diagnostic prediction score for pediatric NMDARE has been successfully developed.
• The score integrates key clinical features and laboratory findings to enhance diagnostic accuracy in children.
• This tool is expected to aid in the early diagnosis and prompt treatment of suspected autoimmune encephalitis in pediatric populations.