Toward a clinically based classification of disease severity for paediatric COVID-19.

In their Article, Haiyan Qiu and colleagues described 36 children with coronavirus disease 2019 (COVID-19) using the Chinese classification for paediatric COVID-19 severity: asymptomatic infection, mild disease, moderate disease, severe disease, and critical illness. Herein, we focus on the definition of moderate disease. This definition is based on clinical criteria (pneumonia with fever and cough in the absence of signs of hypoxaemia), radiological criteria (because “some cases may have no clinical signs and symptoms, but chest CT shows lung lesions, which are subclinical”), or both. In another study of COVID-19 in children, Lu and colleagues also used CT scans to identify lung lesions and classify the severity of COVID-19.

Since the study by Qiu and colleagues, we have evaluated 59 children with suspected COVID-19 in our paediatric emergency department. None of these 59 children had a CT scan; 14 (24%) had a chest x-ray. COVID-19 was confirmed by nasopharyngeal RT-PCR in a child aged 13 years who was exposed to a family member with COVID-19 and who later presented to our emergency department with a fever and dry cough. Blood tests, including inflammatory markers, were within normal ranges. Because the child appeared reasonably well, we did not do an x-ray or a CT scan, but did a lung ultrasound using a procedure we have previously described. The lung ultrasound showed a small, subcentimetric subpleural consolidation with vertical artefacts. The child recovered without treatment and the case of COVID-19 was classified as mild. However, we cannot know whether the CT scan would have showed subclinical lung lesions and thus allowed a classification of moderate disease.

Although chest CT scans improve diagnostic accuracy, doing these scans in children comes with disadvantages, such as high costs, the need for sedation, and radiation exposure. Therefore, a classification of severity that includes the radiological diagnosis of pneumonia is not appropriate for children. Guidelines state that medical history and examination are the determinants of pneumonia severity and appropriate levels of care, and that clinicians should reserve imaging to compromised children needing admission to hospital. Moreover, the high reported number of paediatric patients with asymptomatic and mild COVID-19 suggests that imaging should not be routinely used.1, 3

Therefore, we propose the definition of moderate disease in the paediatric classification of COVID-19 severity should be changed to a clinical diagnosis of pneumonia, frequent fever and cough (mostly dry cough, followed by productive cough), presence or absence of wheezing but no obvious signs of hypoxaemia (eg, shortness of breath), abnormal breath sounds on auscultation, and dry or wet snoring. The section “some cases may have no clinical signs and symptoms, but chest CT shows lung lesions, which are subclinical” should be removed because these asymptomatic or paucisymptomatic children should not have imaging scans done and should be classified as having mild COVID-19.

Qiu and colleagues and others should provide a new description of their patients using the clinically based classification we have suggested, showing how the mild and moderate classes would change. This description would provide a more appropriate clinical picture of the disease to paediatricians looking after children with suspected COVID-19. Reclassifying these data will also help clinicians to properly allocate resources.