Background Multisystem inflammatory syndrome in children (MIS-C) remains one of the most complex post-infectious conditions associated with SARS-CoV-2. In Ukraine, where the healthcare system operates under uneven resource availability due to the ongoing conflict, diagnostic and therapeutic decisions often require simplified, adaptable tools for early risk assessment. This study aims to develop MIS-C severity scoring systems based on clinical and laboratory predictors and differentiate them according to healthcare facilities resource availability. Methods This retrospective cohort study included 51 children (aged 7 months to 17 years) with confirmed MIS-C. Patients were divided into severe (PICU+) and non-severe (PICU–) groups. MLR was used to identify independent predictors, followed by internal bootstrap validation. Two resource-adjusted models were proposed: MIS-C Severity Score Basic (MIS-C-SS-Basic) and MIS-C Severity Score Advanced (MIS-C-SS-Advanced). Results The Advanced model included four-system involvement, hepatomegaly, and IL-6 ≥ 310 pg/mL and showed AUC = 0.936 with sensitivity 91.7% and specificity 88.9%. The Basic model included ESR ≥ 22 mm/h, hepatomegaly, and free abdominal fluid and showed AUC = 0.826 with sensitivity 87.5% and specificity 70.4%. Both models were converted into point-based scores. Conclusion Two resource-adapted MIS-C severity scores were developed to predict PICU admission in children with MIS-C. External validation in independent cohorts is needed.
Bodnarchuk-Sokhatska et al. (Thu,) studied this question.