• data collection
• mortality
• qualitative research

International reports of the COVID-19 pandemic have described the relative sparing of children, both in case frequency1 and disease severity.2–4 The major presentation described in children is the paediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 (PIMS-TS)—many of these children do not have evidence of PCR-positive viral disease and therefore may represent a postinfectious phenomenon.5 We describe our single-centre paediatric intensive care unit (PICU) experience of children who tested positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in the first 10 weeks of the pandemic in the UK, excluding those who met PIMS-TS criteria as these have been described elsewhere.

A total of 313 children were admitted to our intensive care units (ICUs) from the 26 March 2020 to the 31 May 2020. Ninety-six children were suspected to have COVID-19 of which 24 (25%) children tested reverse transcription (RT-PCR) positive at admission with SARS-CoV-2 on nasopharyngeal aspirate. Thirteen children presented with a PIMS-TS phenotype; here, we describe the characteristics and course of the remaining 11.

The demographic and presenting characteristics are shown in table 1. Comorbidities were present in nine out of eleven children. Four (36%) children were from a black, Asian or minority ethnic background. The clinical characteristics are presented in table 2. Inflammatory markers were raised, but the range of maximum values during admission were widely variable.

Four children had respiratory disease fulfilling the 2015 PALICC (the paediatric acute lung injury consensus conference) criteria for paediatric acute respiratory distress syndrome. Hypoxaemic respiratory failure management included prone ventilation (n=4) and inhaled pulmonary vasodilators (n=3). Two infants were escalated to high frequency oscillation due to refractory hypoxia on conventional ventilation. Median duration of ventilation was 13 days (IQR 10–15.5 days).

The remaining seven children required admission to PICU for reasons other than respiratory failure. Three children presented in status epilepticus (two with known seizure disorders and one with an acquired head injury). All tested negative for SARS-CoV-2 on cerebrospinal fluid analysis. Three children presented with significant new diagnoses (congenital heart disease, leukaemia and diabetes mellitus) which would have required PICU admission regardless of their SARS-CoV-2 status. The other child was undergoing chemotherapy for a malignancy and was on established long-term ventilation but did not require a significant escalation in their ventilator parameters. Two of the cohort were immunocompromised.

Among other therapies, five (45%) children received vasoactive drugs. None received renal replacement therapy or extracorporeal membrane oxygenation. Five children (45%) received compassionate use of remdesivir following ethics review. Six children (55%) received prophylactic anticoagulation as part of a modification of our usual practice. The two infants developed line-associated thrombosis in the absence of prophylaxis requiring therapeutic anticoagulation.

All children survived to discharge from PICU.

While children can present to PICU with a pattern of illness similar to adult COVID-19 disease, this is rare and three-quarters of them had risk factors for respiratory infections. A larger number of children were found to be SARS-CoV-2 positive coincidentally. While a causal relationship between some presentations and SARS-CoV-2 infection cannot be ruled out, these cases will have implications for hospital infection control precautions in children with critical illness throughout the pandemic.