Journal Pre-proof Clinical characteristics and factors associated with severe COVID-19 in hospitalized children during the SARS-CoV-2 Omicron pandemic in Taiwan

Vital signs need close monitoring, early management and/or intensive care may be


Conclusions 24
Vital signs need close monitoring, early management and/or intensive care may be 25 applied in COVID-19 patients with congenital cardiovascular diseases, fever 26 lasting ≥4 days, seizures, desaturation and/or elevated procalcition since they are 27 at higher risks of severe diseases. 28

Introduction
As of September 2022, more than 600 million cases of severe acute respiratory 32 syndrome coronavirus 2 (SARS-CoV-2) infection were documented worldwide. More 33 than 6.4 million deaths have been caused by coronavirus disease 2019 . 1  Omicron wave reported that children with underlying diseases, such as congenital heart 47 outcomes. 5,6 However, data in the era of the Omicron variant are still lacking in 50 identifying predictors of severe  Our objective was to provide demographic characteristics, clinical manifestations, 52 laboratory data, and clinical outcomes of hospitalized children or adolescents with 53 SARS-CoV-2 infection under the Omicron predominance in Taiwan. We would like to 54 define clinically significant predictors associated with severe COVID-19 in children in 55 the Omicron wave so that early management of high-risk patients could be implemented 56 to improve their clinical outcomes. 57 J o u r n a l P r e -p r o o f patient or the parent. RT-PCR was performed by a Roche cobas SARS-CoV-2 assay 78 (Roche Molecular Systems, Branchburg, NJ, USA) with three unique TaqMan probes 79 targeting conserved regions within the ORF (open reading frame) 1ab and E genes. 7 Ct 80 values were provided by the testing platform based on manufacturer-provided 81 interpretation criteria. 82 An oxygen saturation value ≤94% measured by oximeter was defined as desaturation. 83 The oxygen saturation we documented was measured at emergency room or upon 84 admission to the ward. Abnormal creatinine was defined as a patient with creatinine 85 levels higher than the upper limit of age-appropriate values. 8 Neutrophilia was defined 86 as a patient with an absolute neutrophil count higher than the upper limit of age-87 appropriate values. Leukopenia was defined as a white cell count lower than the lower 88 limit of age-appropriate values. 9 Thrombocytopenia was defined as a platelet count 89 <150,000/microliter. The laboratory data were measured from the blood samples 90 obtained upon admission. 91 92

Definition of clinical severity and diagnoses 93
We categorized the hospitalized patients into "nonsevere" and "severe" groups. The 94 definition of "severe" was modified from the maximum Clinical Progression Scale 95 research. 10 Patients requiring intensive care (including shock needing aggressive fluid 97 supplement, vasopressors or inotropes, respiratory distress requiring invasive or non-98 invasive ventilator, neurological symptoms such as new-onset seizure with focal 99 neurological signs, change of consciousness, or status epilepticus), extracorporeal 100 membrane oxygenation (ECMO), or death were separated into the "severe" group. The 101 other conditions were sorted as the "nonsevere" group, including patients admitted to 102 the ICU for invasive procedures without requiring intensive care. 103 The definition of clinical diagnosis for croup was defined as hoarseness, stridor, and 104 barking cough. Pneumonia or bronchopneumonia was defined as an abnormality on 105 chest X-ray and crackles revealed by auscultation. Encephalopathy was defined as 106 altered mental status (defined as altered level of consciousness, lethargy, or personality 107 change) lasting ≥24 hours without alternative cause identified plus two or more minor 108 criteria. Minor criteriae included (1) generalized or partial seizures not fully attributable 109 to a preexisting seizure disorder or febrile convulsion, (2) new onset of focal neurologic The Kolmogorov-Smirnov test was used to test the normality of the sample distribution. 116 If the data were not normally distributed, the Mann-Whitney U test was used to assess 117 the continuous variables. We used the chi-square test to analyze categorical data. We 118 used ROC curves to determine the cutoff value of continuous variables. Binary logistic 119 regression analysis was used to identify the significant risk factors for severe COVID-120 19. Unadjusted and adjusted risk ratios are presented with 95% confidence intervals 121 (CIs). We regarded P values <0.05 as statistically significant. All statistical analyses 122 were performed by using Statistical Product and Service Solutions (SPSS) version 23. 123 124

Ethics 125
This study was approved by the Institutional Review Board of National Taiwan 126 University Hospital (202206053RINC). All data were deidentified before being 127 analyzed. Informed consent was waived due to the nature of the retrospective study in 128 which the analysis was conducted with anonymous clinical data. 129 J o u r n a l P r e -p r o o f

Demography 131
A total of 355 children or adolescents with laboratory-confirmed SARS-CoV-2 132 infection were found. Fourteen were excluded due to identified evidence of SARS-133 CoV-2 infection more than 14 days before admission or false-positive results of SARS-134 CoV-2 PCR. Other two children were excluded due to asymptomatic infection 135 found by screening incidentally. Finally, 339 cases were enrolled in our study. We 136 separated them into "nonsevere" and "severe" groups. There were 22 cases (6.5%) in 137 the severe group. The median interval between disease onset and clinically severe 138 illness was 2 days (IQR: 0-4 days). Six patients developed severe disease after 139 admission with a median interval of 2 days (IQR: 2-3 days). 140 The demographic data and characteristics of the two groups are shown in Table 1. 141 The median age in patients with severe COVID-19 was older than that in patients in the 142 nonsevere group. Patients older than 72 months were significantly more common in the 143 severe group. Figure 1 illustrates the overall age distribution of the enrolled children. A 144 significantly higher proportion of patients in the severe group had underlying diseases, 145 including congenital heart diseases, neurological diseases, and multiple underlying 146 diseases.
Clinical manifestations between the two groups are shown in Table 2. Most of the 150 children had a fever (94.1%) with a median duration of two days (IQR: 2.0-3.0 days). 151 Children with severe COVID-19 presented with a longer duration of fever and a higher 152 agents. Intravenous immunoglobulin (IVIG) and anti-interleukin-6 (anti-IL6) agent 191 were given in some patients with severe diseases, showed in supplementary Table S4.  Table 4. 201 There were two deaths during hospitalization in our study, and neither patient 202 The results of multivariable analysis are shown in Table 5 reported a French cohort in which children younger than 90 days accounted for 37% of 228 cases, but only 3% of them had severe  This was different from two studies 229 in China and Italy, which reported a higher rate of severe disease among children 230 younger than 1 year and 6 months, respectively. 13,14 In our cohort, most of the 231 hospitalized patients under one year old were admitted due to young infant fever or 232 dehydration but rare severe illness caused by  In the present study, only 6.5% of hospitalized children and adolescents required 234 which was pre-Omicron wave, in the UK (15%) and in the U.S. (16.7% and 8.7%) in 236 the acute COVID-19 cohort without 15,16 In our cohort, only 3.2% of patients 237 needed mechanical ventilation, which was lower than the rate reported by Martin et al. 238 (6% in the acute COVID-19 cohort) and 8.9% reported by Swann et al. 11,17 Adeel et al. 239 also found that Omicron infected children were associated with significantly lower odds 240 of severe disease than Delta infected children. 18  Most of our patients presented with fever (94.1%), which was higher than data 258 reported in France (74%) and South Africa (61%) during the Omicron wave. 23,25 Most 259 of the patients in our cohort were admitted due to symptomatic SARS-CoV-2 infection, 260 and the two patients with asymptomatic SARS-CoV-2 infection were excluded. The 261 above may be why a high proportion of our patients presented with fever. Another 262 reason is that all patients in our study had no previous SARS-CoV-2 infection, and most 263 did not receive a COVID-19 vaccination. This may result in more patients presenting 264 with symptomatic SARS-CoV-2 infection. 26 We found that fever duration ≥4 days is 265 one factor associated with severe COVID-19. To our knowledge, there have been no 266 previous reports on fever duration associated with severe COVID-19. This finding 267 suggests that we should alert patients about fever duration ≥4 days to be related to 268 developing severe disease, and further survey and management are recommended. critically ill COVID-19, which was associated with worse functional outcomes reported 283 by Liotta et al. and Pun et al. 27,28 In our study, seizures at presentation noted in enrolled 284 patients were frequently associated with encephalopathy (15 of 25 patients), and 6 of 285 them required intensive care. As a result, seizure or desaturation at presentation is 286 important manifestations implying potential moderate to severe disease. 287 We found that an elevated level of procalcitonin (>0.5 ng/mL) upon admission 288 was a factor associated with severe COVID-19 rather than CRP. Procalcitonin is 289 produced from extrathyroidal sources and is inhibited by an increase in interferon 290 (INF)-γ during viral infection. The biomarker was extremely amplified during bacterial 291

COVID-19 progression in adults in many studies and in a systematic review conducted 293
by Ahmed et al. 29 Compared with adults with COVID-19, abnormal procalcitonin was 294 more common, and higher values of procalcitonin were reported. 30,31 According to a 295 report from Xia et al., 31 coinfection was common in their cohort, but there was no 296 patient diagnosed with bacteremia or sepsis in our study. Hence, elevated procalcitonin 297 should be taken as a predictor of potential severe COVID-19 irrespective of bacterial 298

coinfection. 299
There were two deaths in the hospital due to severe COVID-19. Both patients had 300 underlying diseases (cyanotic congenital heart disease and multiple congenital 301 anomalies with epilepsy) but did not receive antiviral agents within the first five days 302 of the onset of illness. Remdesivir, an antiviral agent for SARS-CoV-2 infection that 303 can be applied to pediatric patients, has been identified as having an effect against death 304 or progression to ventilation. 32 Hence, antiviral agents may be considered in patients 305 with high-risk underlying diseases when they are infected by SARS-CoV-2. 306 The main strength of our study was that all the enrolled patients had no 307 documented previous SARS-CoV-2 infection, and most did not receive a COVID-19 308 vaccination before they were infected. These points make the analyses less affected by 309 the potential effect of previous infection and/or vaccination. Our study also has several 310 small number of severe COVID-19 cases and mortality in children. Second, the 312 demographic characteristics and clinical spectrum of hospitalized patients may be 313 affected by the strategy of disease control and medical resource allocation in different 314 stages of the pandemic. Third, due to the nature of retrospective studies, some 315 laboratory data may be missing or unavailable, which may underestimate their 316 predictive power for severe disease. In addition, our analysis did not include data on 317 radiologic findings that may be potentially associated with severe disease. 33,34 Some 318 previously reported risk factors, including prematurity, obesity, and diabetes, 15,23 were 319 not present in our analysis due to incomplete data or the small number of cases. 320 In conclusion, this study showed that pediatric patients with cardiovascular disease, 321 fever duration ≥4 days, desaturation, or seizure on admission, and procalcitonin >0.5 322 ng/mL were associated with severe COVID-19. When clinicians approach patients 323 with the above conditions, their vital signs, respiratory patterns, neurologic 324 symptoms and activities need close monitoring, early management and/or 325 intensive care may be applied. 326 327

Conflicts of interest 328
None of the authors declare conflicts of interest associated with this manuscript. 329

Acknowledgments 331
We thank the staff members of the emergency department of National Taiwan 332 University Hospital for the management of patients and the pediatric neurologists of 333 National Taiwan University Hospital for consultation. This study was supported by 334 the National Science and Technology Council (NSTC 112-2321-B-002-013 and 335 NSTC 111-2321-B-002-017) and National Taiwan University Hospital (NTUH 336 W1_112-03). The funders of this study had no role in the study design, data 337 collection, analysis, interpretation, or writing of the report.  Myalgia or muscle soreness 22 (6.5%) 21 (6.6%) 1 (4.5%) 0.707 N denotes case number; IQR denotes interquartile range; "duration of fever" is defined as the total duration of fever of the patients J o u r n a l P r e -p r o o f Data are shown as case numbers (percentages) or medians (interquartile ranges).
J o u r n a l P r e -p r o o f