DISCUSSION
We conducted a systematic review and meta-analysis with the current evidence available regarding illness severity and clinical features in children (0-21 years), aiming to provide the most frequent COVID-19 characteristics in the pediatric population in order to bring useful information.
The majority of studies included in this systematic review are from China, and eighteen from other countries, this brought more variability in to our study, in comparison to other systematic reviews that only included Chinese population, nonetheless frequencies for epidemiological data, illness severity and clinical features are similar.  Most of the patients in the systematic review published by Castagnoli et. al. were >10 years old (553) [62]. Similarly, we found the higher number of cases in the teenage population (>10 years old) 44.23% (2505). The Korean Society of Infectious Diseases has reported a similar age distribution, this might be associated with teenagers being more active than other pediatric groups [64]. Sex distribution is nearly the same (55.25% vs 57.35% cases in males) in our study compared with the systematic review published by De Souza TH.et al. [65]. Above 85% cases had exposure with a confirmed case in our study. In contrast, a study from Tagarro. A et al. 2020 [49], described 365 suspected cases in children, 41 were positive, 56% of confirmed cases were females and 39% of all confirmed cases had previous contact with a confirmed case versus 61% cases of community transmission. Unlike other respiratory viruses, everything seems to indicate currently, that children are not superspreaders, but they may be the final part of the chain of contagion. This has important implications for school opening policies when the epidemic is controlled locally. [66]
Regarding the illness severity, 95% cases were non-severe, of whom, 6% were asymptomatic and 33% had abnormal findings in chest CT without any serious clinical manifestation, similarly in the systematic review of De Souza TH.et al 96.5% of patients were non-severe but 14.2% were asymptomatic [65]. Asymptomatic cases in the pediatric population possess several challenges to public health, however, it is not clear  how this cases in children could affect in the velocity of spread of the virus, besides it would be necessary to stratify the viral load of these patients in order to be able to estimate the risk of infection, in a recent the viral load of SARS-CoV-2 was analyzed  by patient age, finding no significance difference in viral load between children and adults, thus children could be as infectious as an adult [67]. On the other hand, our results have shown that children develop non-severe disease in most cases in comparison with adults, who present 81% of mild cases [8] versus almost 95% in pediatric patients.  Moreover, severe cases in adults represent 14% and critical cases 5% by contrast only 2% and 3% children cases were severe or critical, respectively. Differences in illness severity between children and adults had brought questions about why severe/critical disease frequency in children is low. Garcia-Salido in their letter to the editor discusses some hypotheses about children having more frequently a mild disease. Firstly, the expression of the angiotensin-converting enzyme (ACE) 2 receptor is lower in nasal fluid when the individual is younger, and it could explain the minor frequency of infection and severe disease in pediatric population. The endothelial damage is higher in adults with diabetes and cardiovascular diseases in comparison with healthy children where is practically absent. Finally, the innate immunity in children is well trained by viral community infections and viral vaccines. [68]
Although the focus of the pandemic is on adults, prevention in the pediatric population should not be neglected. Pathak EB et al. [69] estimated the projected numbers of severe pediatric cases in the USA, under different scenarios of cumulative pediatric infection proportion (CPIP), finding 10,993 projected severe/critical cases among children with a 5% CPIP. Despite non-severe pediatric cases will be the majority, the total number of severe cases will exceed the capacities of a health care system already impaired.
It is important to emphasize that children are not very symptomatic and not all meet the operational definitions, for example, abdominal pain, vomit and diarrhea have almost the same frequency as cough.Figure 4 . Unlike adults, in a meta-analysis published by Rodriguez-Morales AJ et al [70] found that 92% of adults presented with fever and 63% with cough, a much higher than children.   Other symptoms less frequent in children in contrast with adults are expectoration, fatigue/myalgia, sore throat and dyspnea whereas had equal frequency and gastrointestinal symptoms like diarrhea, vomit and abdominal pain were more frequent in children. Figure 4 . Hence COVID-19 in children represents a clinical challenge since it is less symptomatic and less severe, making it more difficult to identify. Although the children may spread the virus less, we do not know what will happen when they have more social contact. Unlike Castagnoli study, we observed an increase of severe cases, that are explained by appearance and description of PIMS-TS. Their main clinical features were fever, conjunctivitis, rash, lower limb edema and gastrointestinal symptoms, some of them developing shock and cardiac alterations. Patients presented elevated CRP, ferritin, D-dimer and cardiac enzymes; these clinical findings demonstrated a new phenomenon related to SARS.CoV-2 infection in children. [17,18]
Some clinical features of this new syndrome are KD-like but have some differences, for example this syndrome is related to a specific causal agent; SARS-CoV-2 unlike KD whose etiology is still unknown. KD mainly affects children under 5 years but the prevalence of this syndrome is higher in adolescents, the syndrome manifests itself with greater severity, shock, organ failure, heart dysfunction with decreased ejection fraction and higher prevalence in PICU admission compared to KD. [17,48].
We include five studies that referred specifically to this syndrome, the available information is very limited. These are clinical manifestations of severe COVID-19 in children that can lead systemic complications, admission to ICU and even death, therefore it is important that pediatrician take into account this possible presentation in pediatric patients with symptoms similar to KD and optimize early recognition and management to avoid death.
It is important to mention that our study has several limitations such as letters to the editor, including cases series and in-press articles, having patients diagnosed after a confirmed family member and lastly, heterogeneity among studies goes from medium to high. Nevertheless, this study brings useful epidemiological and clinical information about the pediatric population, for instance, children less than 1 year old could be a group of risk since patients <1y have twice the risk of developing severe or critical illness compared to other pediatric agesFigure 3. For this reason, public health authorities take to account that children can and will get severe disease and we must be aware for this target population. Children in most of cases will present a mild disease but with an atypical presentation, where clinicians must identify and discard COVID-19 in children, so testing in the pediatric population must increase.