DISCUSSION
In this study, we compared CAP in patients with SARS-CoV-2 with patients positive for other viral infections. We found that the former was associated with less wheezing and work of breathing, a significantly lower lymphocyte fraction and lower CRP levels. During evolution, the SARS-CoV-2-associated CAP group had significantly higher MV use (almost ten-fold) but less requirement for oxygen. No significant differences were found in terms of days of hospitalization, PICU admission or CPAP/HFV use.
Several studies have reviewed the characteristics of adults with SARS-CoV-2 and pneumonia17–21, but these remain scarce in the pediatric population22,23. However, among hospitalized children and adolescents, pneumonia is a major cause of disease (approximately –gativizes after a median of 17-e PCR15%)5–8.
Similar to other studies22,23, the most frequent symptoms in our patients were fever, cough, wheezing or shortness of breath. We found that SARS-CoV-2-associated CAP occurred in older children (8 years on average, as reported previously24), with a longer duration of fever, more cases with chest and abdominal pain and fewer cases with cough, wheezing or dyspnea than in non-SARS-CoV-2-associated CAP. Symptoms often overlapped, making it challenging to discern between the two. Given that the median time to RT-PCR negativity for SARS-CoV-2 is 17–19 days, and can remain positive for several weeks up to 3 months25,26, a means of differentiating SARS-CoV-2 from other etiologies is important.
Many studies evaluating the utility of biomarkers in defining the etiology of pediatric CAP have been performed using FBC, neutrophil percentage, serum CRP or procalcitonin, although the cut-off points are not well defined. As reported in other studies, we found significant differences in terms of FBC and inflammatory markers6,22,23,27, with increased CRP in almost 50% of the patients, but less intense in the SARS-CoV-2 group than in the non-SARS-CoV-2 group. Some studies have reported very high rates of lymphocytopenia6,23, but it appears to be less common in children than in adults12,13. Variable ymphocytopenia values have been found in children, between 3–33%8,21,22,28,29. In our cohort, 42% of the children presented with lymphocytopenia. Both leukocytes and lymphocytes were significantly lower in patients with SARS-CoV-2-associated CAP than in other viral-associated CAP.
The evolution in children was usually good. Mortality in children with SARS-CoV-2-associated CAP is rare, less than 5% in different series7,23,30 (3% in our study). It seems that patients with comorbidities are at higher risk23,30: In this context, about 50% of the EPICO group had underlying conditions.
The radiological presentation of SARS-CoV-2 can be non-specific and indistinguishable from other pathologies. The first published studies of children reported few findings in radiographs15; however, subsequent reports have revealed a higher proportion of radiographic abnormalities11, which depend on the severity of pulmonary involvement31,32. In adults, most computed tomography studies in SARS-CoV-2-associated CAP show ground-glass opacities (25–60% according to different studies)33–38 that can progress to white lung. The proportion of white lung in children is low, and the mechanism is worth further study. Here, the proportion of other infiltrates was 58%, which might correspond to ground-glass opacities, but they did not often progress to white lung, perhaps because of the limited inflammatory response in children. The proportion of consolidations in SARS-CoV-2-associated CAP (42%) was lower than that found in the other viral CAP group (64%), and again the interpretation is unclear. It might reflect different possible patterns of the infection or different susceptibility to bacterial superinfection. Although bacterial coinfections are frequent in viral CAP4,38,39, they were very rare in our cohort (2%). However, a full work-up was not performed for most patients (due to laboratory overload during the pandemic). In some severe cases, pleural effusion may be found but it is rare10-11,34 (in our series we found it only in 6 patients, 4%).
Severe COVID-19 is rare in children, with variable PICU admission rates22,23,30. In our cohort, 15% of children with SARS-CoV-2-associated CAP required PICU admission and while this was almost double the proportion of the other viral CAP group, the differences were not significant, which may be due to a limited sample size.
Usually, viral-associated CAP in young infants resolves with oxygen therapy or CPAP. The evidence for an effective treatment for SARS-CoV-2 is evolving rapidly. Most reports mention supportive treatment12, including oxygen therapy and MV, but MV requirements are highly variable according to different studies in adults (18–42%40). Approximately, half of all our COVID-19 patients required oxygen therapy, up to 10% needed additional noninvasive respiratory support (CPAP, HFV), and 15% required MV. In our cohort, patients with other viral-associated CAP required oxygen more frequently, but SARS-CoV-2-associated CAP had more complications with a greater use of MV. Lung damage associated with SARS-CoV-2 appears different to other viruses, as patients need oxygen less often and MV more often. We hypothesize that part of the respiratory damage is vascular, neuro-muscular, or heart-mediated, rather than solely hypoxemia.
Although a high percentage of patients (74%) in our series received antibiotic treatment, it is often not needed in SARS-CoV-2-associated pneumonia owing to the viral etiology. The reason for this high proportion might lie in the fact that most of our cases were during the first wave of the pandemic when treatment was not yet well established, and by the high percentage of severe cases, as only hospitalized patients were included. Even so, most cases of viral-associated CAP receive antimicrobials at admission11.
Our study has some limitations. First, only hospitalized patients were included, and so the results are not representative for ambulatory CAP. Second, some cases had incomplete documentation of the exposure history or clinical features, and not all patients underwent a complete blood test or microbiological workup. The third limitation is the variation in the interpretation of radiographs depending on the observer, which can lead to different interpretations. Likewise, both cohorts were not paired in time, which can complicate the comparability of the results.