key: cord-0814545-1kaa50mh
authors: ROBINSON, J.; Barton, M.; Papenburg, J.; Ulloa-Gutierrez, R.; Brenes-Chacon, H.; Yock-Corrales, A.; Ivankovich-Escoto, G.; Soriano-Fallas, A.; Mezerville, M. H.-d.; Bitnun, A.; Morris, S. K.; El Tal, T.; Yeh, E. A.; Gill, P.; Laxer, R. M.; Nateghian, A.; Haghighi Aski, B.; Manafif, A.; Lefebvre, M.-A.; Caya, C.; Cooke, S.; Dewan, T.; Restivo, L.; Theriault, I.; Trajtman, A.; Dwilow, R.; Bullard, J.; Sadarangani, M.; Roberts, A.; Le Saux, N.; Bowes, J.; Wong, J. K.; Purewal, R.; Lautermilch, J.; Leifso, K.; Foo, C.; Newhook, L. A.; Bayliss, A.; Petel, D.
title: Multicenter cohort study of children hospitalized with SARS-CoV-2 infection
date: 2021-02-23
journal: nan
DOI: 10.1101/2021.02.19.21251340
sha: db1babff52cc36181e1ebe3f4886ebe4a40bf94d
doc_id: 814545
cord_uid: 1kaa50mh

Background: A cohort study was conducted to describe and compare the burden and characteristics of SARS-CoV-2 infection in hospitalized children in three countries. Methods: This was a retrospective cohort of consecutive children admitted to 15 hospitals (13 in Canada and one each in Iran and Costa Rica) up to November 16, 2020. Cases were included if they had SARS-CoV-2 infection or multi-system inflammatory syndrome in children (MIS-C) with molecular detection of SARS-CoV-2 or positive SARS-CoV-2 serology. Results: Of 211 included cases (Canada N=95; Costa Rica N=84; Iran N=32), 103 (49%) had a presumptive diagnosis of COVID-19 or MIS-C at admission while 108 (51%) were admitted with other diagnoses. Twenty-one (10%) of 211 met criteria for MIS-C. Eighty-seven (41%) had comorbidities. Children admitted in Canada were older than those admitted to non-Canadian sites (median 4.1 versus 2.2 years; p<0.001) and less likely to require mechanical ventilation (3/95 [3%] versus 15/116 [13%]; p<0.05). Requirement for oxygen or ICU occurred in 64 (30%) and death in four, three of whom. had malignancies. Age < 30 days, admission outside of Canada, presence of at least one comorbidity and chest imaging compatible with COVID-19 predicted severe disease. Conclusions: Approximately half of hospitalized children with confirmed SARS-CoV-2 infection or MIS-C were admitted with other suspected diagnoses. Disease was more severe at non-Canadian sites. Neonates, children with comorbidities and those with chest radiographs compatible with COVID-19 were at increased risk for severe disease.

The coronavirus disease 2019 (COVID-19) pandemic caused by the Severe Acute Respiratory Syndrome coronavirus-2 (SARS-CoV-2) was designated by the World Health Organization (WHO) as a pandemic on March 11, 2020. It has spread worldwide with an unprecedented impact on the global health care system, with over 106 million confirmed cases and 2.3 million deaths as of February 9, 2021 [1] .

Previous publications consistently described less severe clinical manifestations of COVID-19 in children as compared to adults [2, 3] . It has been postulated that protective factors in children may include a more robust innate immune response, partial immunity from prior infection with seasonal coronaviruses, and/or the lower incidence of comorbidities in children [3] . Children younger than one year of age appear to be at increased risk of hospitalization compared to older children [3, 4] . It is not clear whether this is because they have more severe disease or because the threshold for admission is lower for infants. Comorbidities increase the risk of pediatric hospitalization [5] and critical disease [6] but socio-demographic risk factors for hospitalization or critical disease in children are yet to be described.

It is estimated that at least one-third of adults with SARS-CoV-2 infection remain asymptomatic [7] . An even higher percentage of children may be asymptomatic [8] .

Although neither symptomatic nor asymptomatic hospitalized children have been reported to be a source of hospital-acquired SARS-CoV-2 infections, it is possible that if proper precautions are not instituted prior to confirmation of infection, other patients, caregivers or health care workers could be exposed. Furthermore, infected caregivers . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ; https://doi.org/10.1101/2021.02.19.21251340 doi: medRxiv preprint accompanying their hospitalized children may serve as vectors of transmission in paediatric healthcare settings.

At the more severe end of the clinical spectrum, a small percentage of children develop a multi-system inflammatory syndrome (MIS-C) approximately 2 to 6 weeks following SARS-CoV-2 infection [9] . A systematic review of 655 MIS-C cases reported that 447 required intensive care (68%) and 11 died (1.7%) [10] .

The objectives of this study were to describe and compare the characteristics and outcomes of a cohort of hospitalized children with SARS-CoV-2 infection or MIS-C in three countries, one of which is high income (Canada) and two of which are middle income (Costa Rica and Iran) [11] and to describe risk factors for severe or critical COVID-19.

Fifteen pediatric hospitals (13 in Canada and one each in Tehran, Iran and San José, Costa Rica) entered consecutive symptomatic and asymptomatic children up to 17 years of age admitted February 1, 2020 through November 16, 2020 with laboratoryconfirmed SARS-CoV-2 infection (defined by detection of SARS-CoV-2 by molecular testing from any site and/or positive serology for SARS-CoV-2), including those with MIS-C as defined by WHO criteria [9] . Those who met WHO MIS-C criteria but did not have laboratory-confirmed SARS-CoV-2 infection were excluded to ensure that only cases due to SARS-CoV-2 were included. There were no other exclusion criteria.

Indications for testing and assays used varied over time and by hospital. Cases were identified through the microbiology laboratory, infection prevention and control and/or COVID-19 unit databases.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 23, 2021. ; https://doi.org/10.1101/2021.02.19.21251340 doi: medRxiv preprint Children who were symptomatic but did not require supplemental oxygen were classified as having mild disease. Severe disease was defined as the need for supplemental oxygen due to COVID-19 or MIS-C without either death or admission to ICU. Critical disease was defined as death or admission to an ICU (excluding children admitted to ICU for other indications). The Canadian Nosocomial Infection Surveillance Program definition was used to define healthcare-associated infection: Symptom onset equal to or greater than 7 calendar days after admission to hospital, using best clinical judgement if onset is sooner [13] . Viral co-infection was defined as laboratory detection of any virus concurrent with SARS-CoV-2 infection. Bacterial co-infection was defined as laboratory detection of one or more bacteria that was treated with antibiotics by the attending physician. Testing for coinfection was at the discretion of the attending physician. Medical complexity was defined as presence of a tracheostomy, feeding tube, home supplemental oxygen, or invasive or non-invasive ventilation prior to admission.

Neurological complications were defined as worsening or new onset of seizure, encephalopathy, dystonia, chorea, athetosis, hemiparesis and/or abnormal cerebrospinal fluid cell count for age. Hematologic complications were defined as disseminated intravascular coagulation, thrombosis and/or bleeding.

Data were analyzed using Epi-Info TM Version 7.2. Baseline and demographic characteristics were summarized using standard descriptive statistics. In the descriptive analysis two and three-way comparisons of demographic, clinical, laboratory and treatment data as well as outcomes were done to identify differences between children from the three participating countries. Categorical data were compared using chi square . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) Logistic regression analysis was performed using R version 3.5.2 (R Core Team, Vienna, Austria) to explore predictors of severe or critical disease, comparing those with severe or critical disease to those with asymptomatic or mild disease. Adjusted odds ratios (aOR) and associated 95% confidence intervals (95% CI) were calculated. Children with MIS-C were excluded from this analysis since the risk factors for severe or critical MIS-C likely differ from those for severe or critical acute COVID-19. Children in ICU for indications judged to be unrelated to COVID-19 were also excluded. Baseline characteristics of interest included biological sex, age, country, timing of admission (March-June versus July-November [14] ), and presence of one or multiple comorbidities (as listed in Table 1 ). Obesity and pulmonary disease (including asthma) were specifically analyzed as they are common in childhood. Clinical variables that were explored included presence of a coinfection, shortness of breath, fever for more than 5 days, presence of gastrointestinal symptoms (defined as vomiting, diarrhea or abdominal pain), chest imaging compatible with COVID-19 (the physician completing the case report form determined whether the report of the chest imaging on the chart fit with COVID-19), or abnormal laboratory markers including leukocytosis (>15 X 10 9 /L), leukopenia (<4 X 10 9 /L), neutropenia (<1.5 X 10 9 /L), lymphopenia (< 1.0 X 10 9 /L), thrombocytopenia (< 100 X 10 9 /L), thrombocytosis (> 450 X 10 9 /L), C-reactive protein above 50 mg/L, ferritin above 500 mcg/L or albumin below 29 g/L. Univariable analysis . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 23, 2021. ;

was conducted and variables with p-value ≤ 0.2 were added to the multivariable logistic regression model in a stepwise manner.

Three centers with a low community prevalence of SARS-CoV-2 infection had no patients that met the inclusion criteria. Of 294 cases enrolled from the other 12 centres, 21 with MIS-C were included while another 83 were excluded as they did not have confirmed SARS-CoV-2 infection ( Figure 1 ). Eight (4%) of the 211 cases met the definition for healthcare-associated SARS-CoV-2. Symptom onset ranged from day 7 to day 47 (N=7; one remained asymptomatic) with viral detection ranging from day 8 to day 55. One of the 8 was an infant with extreme prematurity born to a mother with COVID-19 at delivery; the infant had previously tested negative but tested positive on day 18 of life when new respiratory symptoms developed.

Eighty-seven (41%) of the 211 children had at least one comorbidity (Table 1) , including asthma (N=15; 7%), obesity (N=13; 6%), cancer (N=12; 6%), and other conditions requiring immunosuppressive drugs (N=10; 5%). Thirty-five (17%) had multiple comorbidities. Fourteen (7%) met the definition for medical complexity.

Laboratory confirmation for the 211 patients was via molecular detection (N=185), serological confirmation (n=21) or both (N=5). Molecular detection was from the nasopharynx (N=154), nose (N=18), endotracheal tube (N=4), and throat (N=20) with some having detection from more than one site.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 23, 2021. ;

Ages and symptoms are shown in Figures 2 and 3 respectively for patients with (N=21) and without (N=190) MIS-C. Fever was the most common symptom in both groups. Children with MIS-C were more likely to have fever, gastrointestinal symptoms, myalgias, conjunctivitis, cracked lips and rash (P<0.05) than were those without MIS-C.

The median duration of fever was 7 days (interquartile range [IQR] 5, 9) and 3 days (IQR: 1, 5; P<0.001) in those with and without MIS-C respectively.

One hundred eight of the 211 cases (51%) were not suspected to have or MIS-C on admission (Table 1 ). Eighty-one (76%) had asymptomatic or mild disease.

As mentioned, 8 later developed healthcare-associated SARS-CoV-2.

There were 23 bacterial and 16 viral coinfections (Table 1) 

Of the 211 cases, 140 (66%) did not require supplemental oxygen or ICU admission, 30 (14%) required supplemental oxygen but not ICU admission, one (0.5%) developed healthcare-associated COVID-19 while already in ICU, and 7 (3%) tested positive for SARS-CoV-2 on the day that they were admitted from home to ICU but the ICU admission was for other indications. The remaining 33 (16%) required ICU . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021.

admission for COVID-19 (N=26) or MIS-C (N=7; 33% of MIS-C cases), of which 18 (55%) required mechanical ventilation for COVID-19 (N=14) or MIS-C (N=4; 19% of MIS-C cases). Of the medically complex children, 7 of 14 (50%) were admitted to ICU. This included the only child with a tracheostomy and the only child mechanically ventilated at home, 2 of 5 who were tube fed, 2 of 3 who were on home oxygen, and 1 of 4 who were on non-invasive ventilation on home. There were four deaths (three in Iran and one in Costa Rica): i) a child with a single kidney and ureteric reflux died of respiratory failure 5 days after admission with COVID-19, ii) a child admitted for chemotherapy for leukemia had healthcare-associated COVID-19 confirmed on day 21 of admission and died on day 34. iii) a child admitted with relapsed leukemia had healthcare-associated COVID-19 confirmed on day 16 of admission and died on day 19, and iv) a child with an untreatable brain tumour died 2 days after admission with COVID-19. SARS-CoV-2 was thought to have caused the first two deaths and hastened the latter two. Table 2 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ;

illness. Although ICU admission rates were comparable ( p= 0.06). The COVID-19 cases in Iran mainly occurred early in the pandemic so were commonly managed with therapies that are no longer recommended (azithromycin, hydroxychloroquine, and lopinavir-ritonavir).

In the multivariate analysis, age ≤ 30 days, hospitalization outside of Canada, the presence of at least one comorbidity and chest imaging abnormalities compatible with COVID-19 remained significant (Table 3 ).

Of 211 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ; admission; 25 (7%) were managed with mechanical ventilation and 4 (1.1%) died [15] .

Of 281 patients up to 22 years of age admitted to 8 US hospitals up to April 12, 2020, 70 of 212 (33%) with COVID-19 were admitted to ICU, 26 (12%) required mechanical ventilation and 7 (3%) died while 44 of 69 (64%) with MIS-C were admitted to ICU, 3 (4%) required mechanical ventilation and none died [16] .

The incidence and burden of asymptomatic and mild infections in hospitalized children is just beginning to be described; 34 of 315 infected youth and young adults It is now well recognized that comorbidities increase the risk of hospitalization for children with SARS-CoV-2 infection. A US series reported that 94 of 244 (42%) admitted children had comorbidities, with obesity being the most common ( [17] . In the . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ; https://doi.org/10.1101/2021.02.19.21251340 doi: medRxiv preprint current study, approximately 40% of admitted children had comorbidities but our methodology did not allow us to determine the risk of hospitalization for any given comorbidity. Establishment of risk factors for hospitalization is of potential relevance as the highest risk patients may eventually be suitable targets for immunization or novel early therapies [18] .

Identifying risk factors for severe disease in hospitalized children is perhaps of even greater importance as novel early therapies become available. In the current study, neonatal age group, hospitalization outside of Canada, the presence of at least one comorbidity and chest imaging abnormalities compatible with COVID-19 were risk factors for severe or critical COVID-19. A systematic review of 285 014 children with SARS-CoV-2 infection reported that 5.1% with comorbidities and 0.2% without developed severe disease [5] but as with hospitalization, the risk with any given comorbidity is unknown. A study of 48 patients up to age 21 years admitted to ICUs in the United States reported that 40 (83%) had comorbidities (versus 56% in the current study) of which 19 were medically complex (defined as long-term dependence on technological support associated with developmental delay and/or genetic abnormalities) [6] . In the current study, 7 of the 33 children (21%) admitted to ICU due to COVID-19 fulfilled a similar definition of medical complexity. In the previously mentioned European study, risk factors for ICU admission with either COVID-19 or MIS-C included age younger than 1-month, male sex, comorbidities, and presence of lower respiratory tract infection signs or symptoms at presentation (not further defined) [15] . In the previously mentioned study from the US, risk factors for ICU admission for 48 hours of more for COVID-19 included younger age, obesity, hypoxia, higher admission white . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ; blood cell count and bilateral infiltrates on admission chest radiograph [16] ; risk factors for ICU admission for MIS-C were not reported. A previous pediatric systematic review also reported obesity as a risk factor for severe disease [19] while it did not appear to be a risk factor in the current study. In a study from Oman, leukocytosis for age, elevated CRP (optimal cut-off was >100 mg/L), and anemia for age were risk factors for ICU admission but race or ethnic background were not available in our cohort. MIS-C cases are underreported as i) MIS-C was yet to be described when the study commenced so relevant clinical questions were not included in the original version of the case report form and ii) SARS-CoV-2 serology was not available at any sites early in the study and remained unavailable by the end of the study at some sites.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ;

In conclusion, approximately half of children with SARS-CoV-2 infection or MIS-C diagnosed during hospitalization were admitted with another diagnosis. Clinicians must consider the diagnosis of community-acquired or healthcare-associated SARS-CoV-2 infection in all admitted children. Neonates, children with comorbidities and those with a chest radiograph compatible with COVID-19 were at highest risk of requiring supplemental oxygen or ICU admission during SARS-CoV-2 hospitalization. Death was rare, but children with malignancies appear to be over-represented. Future studies are needed to examine socio-demographic factors that may drive hospitalization rates and outcomes related to SARS-CoV-2 infection in children.

Funding: No funding was received for this project.

is related to the content of this manuscript.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ;

Confirmed COVID-19 -COVID-NET, 14 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ; Table 1 28 (27) 23 (22) 10 (10) 10 (10) 44 (41) 22 (20) 16 (15) 15 (14) 11 ( . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ; Death or Any ICU 25 (23) 17 (16) Oxygen supplementation, ICU admission related to COVID or death 46 (45) 18 (17) 1 Age missing for one child admitted because of COVID-19 2 Indicates that COVID-19 infection probably started more than 10 days prior to day of testing 3 Children will be listed more than once if they had more than one comorbidity. 4 Bacterial coinfections included Enterococcus faecium bacteremia, VAP with Acinetobacter, bacterial pneumonia in a child with CF (P. aeruginosa and methicillinresistant Staphylococcus aureus grown from sputum) and 2 UTIs. Viral coinfections . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. ; is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. 1 Children already in ICU for other reasons (N=7) or with MIS-C (N=20) or both (N=1) were excluded 2 After adjusting for multiple comparisons, a p value <0.0016 should be considered to be statistically significant 3 Shortness of breath was not included in the model presented but was significant with wider confidence intervals in alternate models when chest imaging compatible with COVID-19 was not included in the model. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 23, 2021. 

Acute SARS-CoV-2 infection N=190

Met study eligibility N=211

Excluded as met MIS-C criteria but did not have laboratoryconfirmed SARS-CoV-2 infection . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 23, 2021. ; https://doi.org/10.1101/2021.02.19.21251340 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 23, 2021. ;

Coronavirus Resource Center Global Map

Epidemiology of COVID-19 Among Children in China

Novel coronavirus disease (COVID-19) in children

Coronavirus Disease 2019 in Children -United States

Severe COVID-19 Infection and Pediatric Comorbidities: A Systematic Review and Meta-Analysis

Characteristics and Outcomes of Children With Coronavirus Disease 2019 (COVID-19) Infection Admitted to US and Canadian Pediatric Intensive Care Units

The Proportion of SARS-CoV-2 Infections That Are Asymptomatic : A Systematic Review

Clinical characteristics of pediatric SARS-CoV-2 infection and coronavirus disease 2019 (COVID-19) in Kuwait

Multisystem inflammatory syndrome in children and adolescents temporally related to COVID-19

A Systematic Review of Multisystem Inflammatory Syndrome in Children Associated With SARS-CoV-2 Infection. The Pediatric infectious disease journal 2020

The STROBE guidelines

Surveillance for COVID-19 and other viral respiratory infections among inpatients in

Trends in COVID-19 Risk-Adjusted Mortality Rates

COVID-19 in children and adolescents in Europe: a multinational, multicentre cohort study

Severe Acute Respiratory Syndrome Coronavirus 2 Clinical Syndromes and Predictors of Disease Severity in Hospitalized Children and Youth

Hospitalization Rates and Characteristics of Children Aged <18 Years Hospitalized with Laboratory