Admission serological markers can be used to predict significant cardiac involvement in children with paediatric multisystem inflammatory syndrome

ABSTRACT

Introduction: Paediatric multisystem inflammatory syndrome (PIMS) began to present in April 2020 midway through the covid-19 pandemic. Occurring 2-4 weeks after initial covid-19 infection, patients presented with persistent fever, evidence of inflammation and single or multiorgan dysfunction1. The Yorkshire and Humber congenital heart disease network is made up of the Leeds congenital heart unit and 18 peripheral hospitals2.With limited local paediatric cardiology availability, the vast majority of children presenting with PIMS required transfer to Leeds. This presentation aims to describe the cohort of children that were seen within the network as well as to identify any markers of significant cardiac involvement which could beusedto reduce the frequency of unnecessary inter hospital transfers. Methods: This was a retrospective case notes review of all patients treated within the Yorkshire and Humber network with symptoms of PIMS between 1st May and 30th November 2020. Patients were classified as to whether or not they had significant cardiac involvement (defined as at least one of inotrope requirement, ejection fraction <50%, pericardial effusion, coronary artery changes and significant ECG abnormalities). Cardiac markers were analysed at presentation and throughout the hospital admission including plasma NT pro-BNP, LDH, CRP, d-dimer and troponin. Statistical tests (Fisher's exact test for categorical variables, ttest for continuous variables) were used to identify which factors were indicative of significant cardiac involvement (SCI). Results: 22 patients met the inclusion criteria (Table 1). 14/22 patients (63.6%) were judged to have SCI. Markers that were found to be indicative of SCI included CRP and plasma NT pro-BNP (Table 2). Furthermore, when using a threshold of 2000ng/L, plasma NT pro-BNP was found to be 71% sensitive and 80% specific for SCI. In addition, when combined with a CRP threshold of 100mg/L, there was a positive predictive value of 85% and negative predictive value of 75%. Conclusions: PIMS is an important new syndrome affecting paediatric patients across the Yorkshire and Humber region. A significant proportion of the affected patients have cardiac involvement and require management in a specialist centre. Early identification of these patients using serological markers facilitates rapid treatment preventing long term sequelae whilst also reducing unnecessary interhospital transfers.