Re: SARS-CoV-2 testing and outcomes in the first 30 days after the first case of COVID-19 at an Australian children's hospital.

Dear Editor,

Ibrahim et al. highlighted a very low number of SARS‐CoV‐2 positive cases in their tertiary paediatric setting. We congratulate the authors on encompassing the whole presenting cohort of children who were tested, and would like to make a contribution from the vantage point at our hospital trust, providing emergency and urgent primary care in London, UK, where the incidence of COVID‐19 has been high and community transmission is well established.

The authors underlined that a high proportion of patients who were SARS‐CoV‐2 negative had comorbidities. We contrast this with our experience in London, where we found a sizeable proportion of paediatric cases had comorbidities. We carried out an internal prospective cohort study, tracking the clinical and demographic characteristics of all paediatric patients (age 0–18 years) with a positive combined pernasal‐throat swab for SARS‐CoV‐2 RNA during March and April 2020. We had 18 cases of a positive report, of which nine (50%) had one or more comorbidities, including fibrosarcoma, autoimmune neutropenia, anorexia nervosa, developmental disorder and bacterial co‐infections (aspiration pneumonia, perineal cellulitis, acute appendicitis). We wish to reiterate the vulnerability of children with pre‐existing illness or immunocompromised states.

We believe that adult caregivers may continue to pose a hazard to chronically ill children in periods of school closures. The UK lockdown came into force on March 17. Nevertheless, eight (44%) of our positive cohort had an adult family member who met the case definition prior to onset of illness, suggesting that either (i) in our local population adult carers may be a dominant disease vector during the implementation of school closures, or (ii) there was a greater delay from contamination to symptom‐onset in children compared with adults.

In the paediatric community the concern that children with inflammatory airway diseases would be worse affected by COVID‐19 has not been borne out. Of the children admitted to our centre with bronchiolitis, croup, viral‐trigger wheeze or asthma, we have seen none positive for SARS‐CoV‐2.

Fever was the predominant symptom in our cohort; 13/18 of our cohort had fever alone, and 3/18 had fever and cough. Increased work of breathing or tachypnoea was evident in 6/18, none of whom had hypoxaemia. Eight (44%) of 18 of our cohort were unwell febrile infants under 3 months old who were tested for SARS‐CoV‐2 as part of a septic screen. Three cases were late‐preterm neonates with mild respiratory distress syndrome born to mothers positive for SARS‐CoV‐2. Seven of 17 cases tested showed lymphopaenia and 1/17 had thrombocytopenia. Two of 17 of our cohort had elevated C‐reactive protein levels in the absence of bacterial co‐infection.

We note that the authors did not comment on potential skin manifestations of COVID‐19. In our cohort, 3/18, all of whom were under 6 months old, had persistent livedo reticularis in the absence of systemic shock. Skin manifestations may prove to be a useful clue in the diagnosis of COVID‐19 in children.

We thank the authors for their discernment of a distinguishing feature – headache – that is not yet included in the case definitions for COVID‐19. If the case definition remains narrow, children contaminated with SARS‐CoV‐2 with mild or atypical symptoms will continue to be potential invisible carriers who could thwart efforts to shield more vulnerable populations.

In summary, we believe the authors' findings combined with our own make a case for policymakers to broaden the case definition for COVID‐19 to include headache and skin rashes in children, as this would help containment efforts. Fever remained the predominant symptom but was not universally present, and therefore its use as a screening tool may generate false negatives.

Competing interests

None declared.