Study population and setting
The Duke Biospecimens from RespirAtory Virus-Exposed Kids (BRAVE Kids) study is a prospective cohort study of participants < 21 years old who tested positive for laboratory-confirmed SARS-CoV-2 or lived in the same household with an adult with laboratory-confirmed SARS-CoV-2 with whom they had close contact — unprotected exposure within six feet between two days before and seven days after symptom onset of a positive test result — from April 7 and July 16, 2020. Participants were recruited through a review of positive SARS-CoV-2 test results in the Duke University Health System, followed by a phone call to establish if any individuals < 21 years old lived in the same household as the positive individual. Participants reported exposure, sociodemographic, and clinical data (symptoms from 14 days prior to enrollment) through a directed caregiver questionnaire over the phone and author review of the electronic medical record at baseline. Participants then answered a phone questionnaire seven days after enrollment to assess COVID-19 symptoms, with additional questionnaires at 14 and 28 days in patients experiencing symptoms at day seven. SARS-CoV-2 infection was confirmed with PCR laboratory testing via clinical care for patients who presented to their health care provider with symptoms, home testing visits by research staff, or self-collected nasopharyngeal swabs. They described participant characteristics by SARS-CoV-2 infection status, stratified by age categories to assess symptoms among those infected, using Chi-squared or Fisher exact tests and Wilcoxon rank sum or analysis of variance tests for categorical and continuous variables respectively. They compared nasopharyngeal viral loads by age, symptom status, and collection date relative to symptom onset with analysis of variance or linear regression.
Summary of Main Findings
The study enrolled 382 SARS-CoV-2 infected and/or exposed children/youth less than 21 years-old, 293 (77%) of whom were classified as infected with SARS-CoV-2 via either clinical or research testing. Infected children/adolescents tended to be younger (median age 10.4 years with an interquartile range (IQR) 5-14.4 years versus 10.4 years with an IQR of 4.8-16.4 years), were more likely to identify as Hispanic (88% versus 57%), and were more likely to have a sibling close contact (49% versus 29%) but less likely to have provider-diagnosed asthma (6% versus 17%). Among infected children/adolescents, the majority had symptoms regardless of age category (61-76%), however, respiratory symptoms (cough, difficulty breathing, congestion, or a runny nose) were more common among younger (0-5 years, 48%) and older (14-20 years, 60%) participants as compared to 6-13 year-olds (29%), whereas influenza-like symptoms (headache, muscle pain, sore throat) became more common as participants increased in age category (6% in 0-5 years, 39% in 6-13 years, and 61% in 14-20 years). Symptom duration also increased by age category with the median (IQR) duration of symptoms increasing from 4 days (3-7.5) (0-5 years) to 4 days (3-8) (6-13 years) to 7 days (4-12) (14-20 years). Viral load did not differ significantly by age or presence of symptoms, and was more likely to decrease as the number of days relative to symptom onset increased.
The authors gathered information about SARS-CoV-2 testing and COVID-19 symptoms very close to or before the onset of symptoms among 382 individuals less than 21 and followed them for up to 28 days from a household contact’s positive SARS-CoV-2 test or their diagnosis and measured viral load among 178 infected children/adolescents.
The description of participant recruitment was ambiguous regarding the inclusion of those identified by their own positive test; if these participants were eligible while those who tested negative during the same period were not eligible, the positivity rate of the sample would be biased upwards and would not be informative regarding the risk of transmission. The authors did not describe the refusal rate among eligible participants, or whether individuals who declined to participate differed from those who enrolled in this study. The authors also do not detail how many participants were lost to follow up after the baseline questionnaire, or how this might bias their findings. In addition, the uptake, timing, number of sequential tests, and modality of PCR swab collection for viral load measurement were not reported, and it was unclear whether any of these factors differed among the groups being compared. There is also a group of participants who tested negative for SARS-CoV-2 on clinical tests and were not followed up with research viral load testing and classified as negative, and subsequent cases might have been missed in this group. Finally, the authors did not collect information on where individuals were exposed, which limits any policy implications of these data.
This study presents evidence that COVID-19 symptoms among children/adolescents differ by age category, despite similar SARS-CoV-2 viral loads across age ranges.
This review was posted on: 22 January 2021