A Comparative Analysis of Clinical Characteristics and Laboratory Findings of COVID-19 between Intensive Care Unit and Non-Intensive Care Unit Pediatric Patients: A Multicenter, Retrospective, Observational Study from Iranian Network for Research in Viral Diseases (INRVD)

Background: To date, less is known about the clinical features of COVID-19 pediatric patients admitting to ICUs. Herein, we aimed to describe the differences in demographic characteristics, laboratory ndings, clinical presentations, and outcomes between COVID-19 pediatric patients admitting to ICU and non-ICU settings. Methods: This multicenter study involved 15 general and pediatrics hospitals on conrmed severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection by positive real-time reverse transcription polymerase chain reaction (RT-PCR) between March 19 and May 31, 2020, during the initial peak of the COVID-19 pandemic in Iran. Results: Overall, 166 patients were included, of which 61 (36.7%) required ICU admission, especially in <5 years old age group. Malignancy and heart diseases were the most frequent underlying condition. There was signicant decrease in platelet counts, PH, HCO3 and base excess as well as increases in creatinine, creatine phosphokinase and potassium levels between ICU-admitted and non-ICU patients. Dyspnea was the major symptom for ICU group patients. Acute respiratory distress syndrome (ARDS), shock and acute cardiac injury were the most common features among ICU-admitted patients. The mortality rate was substantially higher in the ICU than in non-ICU patients (45.9% vs. 1.9%, respectively; P<0.001). Conclusions: Underlying diseases were the major contributing factors in COVID-19 pediatric patients for the increased ICU admissions and mortality rates. There are few paraclinical parameters for differentiating pediatrics in terms of prognosis and serious outcomes of COVID-19. Healthcare providers should consider children as a high-risk group, especially those with younger age and underlying medical conditions and dene strategies to control and prevent COVID-19 transmission in this population.


Background
The current Coronavirus Disease 2019  pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) generally causes mild to moderate respiratory tract diseases in humans, and patients usually recover without any special treatment. However, older people and especially those with underlying medical conditions are at an increased risk of developing severe disease outcomes [1].
Several studies on COVID-19 infection among pediatric patients have revealed differences in clinical signs and symptoms, prevalence rates, and mortality rates compared to adults [2,3]. Children are less affected than older individuals by SARS-CoV-2 in the aspect of the number of infected persons and the incidence of serious adverse outcomes [4,5]. Based on the results of previous epidemiological investigations, the prevalence of children affected by COVID-19 was 2.2% and 1.7% in China and the USA, respectively [3,6]. SARS-CoV-2 is transmitted among the pediatric population mainly through direct contact, contaminated droplets, and perhaps aerosols [7,8].
Herein, we aimed to conduct a large multicenter study to compare demographic characteristics, laboratory ndings, clinical features, and outcomes between pediatric COVID-19 patients admitted to the ICU versus non-ICU cases.

Study design, setting, and participants
The current survey was a retrospective cross-sectional study carried on a total of 325 pediatric patients diagnosed with COVID-19 infection who admitted to 15 general and pediatrics hospitals collaborative to the Iranian Network for Research in Viral Diseases (INRVD) between March 19th and May 31th 2020.
Overall results have been submitted elsewhere (under revision). A proportion of this population in whom COVID-19 was con rmed by positive real-time reverse transcription polymerase chain reaction (RT-PCR) for SARS-CoV-2, according to WHO interim guidance [22] was selected for further analysis (n = 166). The major inclusion criteria were patients who needed ventilation support either invasive mechanical ventilation or extra corporeal membrane oxygenation (ECMO) and/or organ dysfunction development who were transferred into the ICU. All other patients were transferred to wards for specialist care and were included as non-ICU cases. This study was approved by the Institutional Review Board of Tehran University of Medical Sciences (Ethics code number: 1399.378) and was conducted in compliance with the principles of the Declaration of Helsinki. Written informed consent was obtained from parents of pediatric patients.

Molecular Detection Of SARS-CoV-2
Throat and nasal samples were obtained using ocked swabs immediately after admission. Laboratory con rmation of the SARS-CoV-2 was performed using the RT-PCR assay [22]. RT-PCR was carried out using the diagnostic kits approved by the Iranian Pasture Institute, targeting the E and RdP genes, along with the same protocol for all laboratories across the country. The samples were deemed positive if the cycle thresholds (Ct) value was ≤ 37 and negative if the Ct value was > 40. Samples with a Ct value between 37-40 was considered as a borderline result and were repeated.

Laboratory And Imaging Indicators
Laboratory examinations including routine blood tests, hematological, biochemical tests and assessment of biomarkers for monitoring lung, liver, and renal functions along with blood saturation parameters were performed in hospitals' laboratories. According to the guidelines issued by The Ministry of Health, all COVID-19 con rmed cases with any respiratory symptoms underwent a chest X-ray upon admission. A chest CT scan usually applied in the case of either the presence of any abnormality in their X-ray or for the patients who progress to the severe form of the disease.

Data Collection
The epidemiological and demographic data, comorbidity, clinical symptoms, and signs on the admission of all laboratory-con rmed COVID-19 pediatric patients were retrospectively extracted from electronic medical records, according to prede ned standardized data collection forms provided by INRVD.

Statistical analysis
Statistical analyses were performed using R software (R Foundation for Statistical Computing, Vienna, Austria; http://cran.r-project.org/). Continuous variables were presented as medians and interquartile ranges (IQR), and categorical variables were expressed as counts (%). Fisher's exact test and chi-square test of independence were used to compare categorical variables, and Wilcoxon's rank-sum test was used for continuous variables. For comparative analyses, a P-value less than 0.05 was considered statistically signi cant.

Results
Demographic characteristics and associated comorbidities Table 1 presents demographic characteristics and related comorbidities of pediatric patients with COVID-19 at hospital admission. Out of 166, 61 (36.7%) were admitted in ICU, and 105 (63.3%) were non-ICU cases. According to the age group, the highest and the lowest numbers of admitted cases to ICU were in the age group of 1-5 and 5-10 years, respectively (31.1% vs 14.7%, P = 0.02, Table 1). Males were more infected in both ICU and non-ICU settings compared to the females (62.8% vs 37.1% and 59.0% vs 40.9%, respectively, P = 0.7, Table 1). Malignancies and heart were the most common underlying conditions; each affecting 11.4% of ICU-admitted patients (Table 1). 21.3% of ICU-admitted patients and 39.0% of non-ICU patients had histories of antibiotic use (P = 0.02, Table 1). However, there were no signi cant differences in in uenza vaccination history, corticosteroid therapy, and chemotherapies between two groups (Table 1). History of chemotherapy 15 (9.0%) 9 (8.5%) 6 (9.8%) 0.9

Clinical Characteristics
Fever and cough were the most frequent clinical symptoms among both groups. However, dyspnea was more prevalent among the ICU than non-ICU cases (50.8% vs 26.6%, respectively, P = 0.003, Table 1). Signs such as grunting and nasal aring were seen more frequently in ICU versus non-ICU patients (0.01 and < 0.001, respectively, Table 2). Among a list of complications, the acute respiratory distress syndrome (ARDS), shock and acute cardiac injury were the most common features among ICU-admitted patients compared to non-ICU patients with signi cant associations (P < 0.001, P = 0.07 and P = 0.01, respectively, Table 2). The mortality rate was signi cantly higher in the ICU than in non-ICU patients (45.9% vs. 1.9%, respectively; P < 0.001, Table 2).

Laboratory Data
There were no substantial differences between both groups in terms of WBC counts, serum in ammatory indexes (CRP and ESR) and other hematological parameters (Table 3). Exceptionally, 19 (32%) of patients admitted to the ICU had normal platelet counts compared to 46 (51%) of those with non-ICU patients (P = 0.08, Table 3). Regarding biochemistry parameters, signi cant increases in creatinine, creatine phosphokinase (CPK) and LDH levels were observed in ICU compared to non-ICU admitted patients, (P = 0.06, 0.01 and 0.06, respectively, Table 3). An elevated level of potassium was seen in 25% and 6.9% of ICU and non-ICU admitted patients, respectively, while it was decreased in 7.0% and 2.3% of ICU-admitted and non-ICU patients, respectively (P = 0.002, Table 3). (34%) of non-ICU cases (P = 0.002, Table 3). Moreover, HCo3 was decreased in the former group compared with the latter group (76% and 36%, respectively, Table 3). Lastly, Base Exess decrease observed in ICU (85%) more than non-ICU groups (64%) (P = 0.09, Table 3).

Radiological Features
The CT scan results performed on admission showed bilateral and unilateral ground glass opacity among 32% and 7.7% of non-ICU-admitted patients, respectively, and 25% and 6.8% of ICU admitted patients, respectively (P = 0.7). Bilateral and unilateral lung consolidation were also observed in 12% and 6.7% of non-ICU-admitted patients, respectively, and 21% and 12% of ICU admitted patients, respectively (P = 0.3). Bilateral and unilateral pleural effusion was found in 5.1% of non-ICU admitted patients, and 7% and 2.3% of ICU admitted patients, respectively (P = 0.8). In ICU admitted cases, unilateral and bilateral white lung were seen in 2.3% of cases and these ndings were not observed in non-ICU cases (P = 0.2).

Discussion
Iran is considered one of the most affected countries by COVID-19 globally, with high incidence and mortality rates. To date, several studies have reported clinical parameters associated with COVID-19 infection in children; however, data on pediatric patients in Iran are still scarce. On the other hand, there are minimal published data on pediatric patients requiring ICU worldwide. This descriptive cross-sectional country-wide investigation compared the epidemiologic and clinical features of ICU admitted and nonadmitted pediatric patients with con rmed COVID-19 in Iran.
Present survey showed that malignancies and cardiac disorders were the most common underlying disease in ICU-admitted pediatric patients. Similar results were found in the surveys conducted by Shekerdemian et al. [23] and Alfraij et al. [24], where malignancy was the most frequent underlying disease among children with COVID-19 admitted to the ICU. Prata-Barbosa et al. and Alfraij et al. also reported that heart disease was amongst the most frequent comorbidities among COVID-19 pediatric patients admitted to the ICU [24,25]. According to these results, children with cancers and heart diseases were associated with increased risk of severe complications of COVID-19.
Investigators observed that COVID-19 patients with cancer had higher ICU admission rates, severe complications, invasive mechanical ventilation, and mortality rate compared with COVID-19 patients without cancer. The higher susceptibility of cancer patients to severe COVID-19 infection might be explained in part by their systemic immunocompromised status induced by the underlying malignancy and anticancer therapy. Furthermore, most childhood malignancies have aggressive behavior and require prolonged periods of intensive therapy, which are potentially associated with long-term side-effects such as severe impairment of innate and adaptive immunity [26][27][28]. On the other hand, some other studies suggested that children with cancer are not more susceptible to severe COVID-19 infection than other children [12,[28][29][30][31]. This controversy stems at least in part from the fact that different kinds of cancers have distinct clinical features such as different growth rates, different responses to treatment, and different prognoses. Unfortunately, we could not be able to nd the data about the type of tumor in our patients due to the retrospective design of our study and so, further conclusions cannot be drawn at this stage.
A proportion of ICU admitted patients were children less than ve years old. Similar ndings were reported in two studies conducted by Dong et al. [2,9], in which infants and younger children were more likely to develop severe clinical manifestations of COVID-19 infection compared with children higher than ve years old. A potential explanation for this phenomenon might be the immaturity of the immune system.
The immune system of neonates and young children is underdeveloped and subdued, which might render them more susceptible to most infections, including SARS-CoV-2.
Among different clinical symptoms, dyspnea was signi cantly frequent in pediatric patients admitted to ICU than the non-ICU patients. This nding was consistent with the result of a survey conducted in New York City [32]. failure is signi cantly associated with pediatric death in their study cohort [24]. It has been reported that heart failure and acute cardiac injury are signi cantly associated with in-hospital death [35]. Taken together, these ndings indicate that acute cardiac injury and acute kidney injury are associated with severe COVID-19 infection among children.
With exception of few blood test parameters abnormalities (such as decrease in platelet, increases in creatinine, LDH, CPK and potassium) and also some atypical blood saturation indexes, the difference in the rest of indexed were not substantial between ICU and non-ICU individuals. These ndings shed important light on the nature of the disease in this population. Unlike in the adults, no signi cant laboratory parameters could be characterized for the prognosis of COVID-19 clinical outcomes in pediatrics worldwide.

Conclusions
In conclusion, this multicenter study demonstrates that underlying diseases were the major contributing factors in COVID-19 pediatric patients for the increased ICU admissions and mortality rates. There are few paraclinical parameters for differentiating pediatrics in terms of prognosis and serious outcomes of COVID-19. Our ndings emphasize that healthcare providers should consider children as a high-risk group, especially those with younger age and underlying medical conditions and de ne strategies to control and prevent COVID-19 transmission in this population. 1399.378) and followed the Declaration of Helsinki. Written consent was obtained from the guardians of the patients.

Consent for publication
Not applicable Availability of data and materials All data generated or analyzed during this study are included in this article.

Competing interests
The authors have no con ict of interest.