select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to study looking at children in the US aged 0 to 19 years whose underlying cause of death was COVID-19

A study published in JAMA Network Open looks at children and young people in the US from 0 to 19 years old with COVID-19 listed as the underlying cause of death.


Prof Sir Andrew Pollard, Director of the Oxford Vaccine Group, Ashall Professor of Infection and Immunity, Pandemic Sciences Institute, and Department of Paediatrics, University of Oxford, said:

“These data should be viewed with the context that mortality in childhood in high income countries like the USA and the UK is at a historic low level and has been falling because of advances in public health including provision of clean water, improved nutrition and vaccines.  Indeed, in 2023 in the UK, about 4 out of every 1000 children under 5 years of age will die compared with around 329 out of every 1000 in 1800.  This means that the ranking of a new disease like COVID19, despite its very low mortality in childhood, makes it appear to be more significant because there are so few deaths from other causes with which to compare it.

“It is difficult to compare directly between countries as the data collection is not identical, but these childhood COVID19 mortality rates, while low, do seem somewhat higher than in the UK and further robust investigation of the underlying reasons for this apparent international difference are warranted.  Anecdotally, colleagues in the USA have pointed to obesity as an important contributing factor and this seems consistent with the Centre for Disease Control observation that obese children are 3 times more likely to be hospitalised than non-obese children.

“The deaths here that stand out are those for accidents and homicide, which should perhaps be the headline from this publication and a call for action to defend the lives of American children against avoidable traumatic deaths.”


Dr Michael Absoud, Honorary Reader, Department of Women & Children’s Health, King’s College London, said:

“The first pre-print version of this paper which was written by UK based authors using US Covid data, first appeared in May 2022.  The original version was unfortunately widely critiqued after the CDC used the analysis in their slide deck for childhood vaccine authorisation1.  Although this is an improved journal published version, I think the authors’ interpretations are not consistent with their own results and there are significant methodological limitations:

• The paper makes a claim that a figure of only 2% represents “a leading cause of death” – I believe using a ‘ranking method’ is a misleading interpretation particularly because the proportion of children dying of any cause is small, except for the top few causes of deaths (which have been grouped in the manuscript).

• The paper includes 17-19 year olds as “children” which contributes to the overestimation of the true value for children.

• The cut off for data collection (31 July 2022) is 6 months ago, hence excluding more recent periods where children now have very high levels of immunity to SARS-CoV-2, after their first exposure from mid 2021 to early 2022 (Delta and initial Omicron waves).  The comparator data also does not use causes of deaths in 2022, and hence missed increases in injuries and significant waves of other viruses and infections which occurred due to an increased susceptible population as mixing levels returned to near normal levels.  These methodological choices result in an overestimation of rates, and make the claim that Covid is the leading cause of infectious disease deaths in children, a misleading one in my view.

• Whilst any childhood death is tragic, the paper also does not examine known risk factors for death from Covid such as severe neurodisability and immunosuppression, which would likely form the majority of cases.

• The paper makes conclusions about pharmaceutical and nonpharmaceutical policy interventions, but the study itself had not examined these aspects.

• The crude death rate reported in this study of 1.0 per 100,000 is not dissimilar to the England wide infection fatality rate of 0.7 per 100,000 reported recently2.  The true rate from more recent data estimated the risk of death due to COVID-19 to be one in a million omicron infections in children and young people, being at least 6-fold lower than preceding waves, given high levels of population immunity3.”

  1. Fact Check: Covid as a Leading Cause of Death in Children – COVID-19 in Georgia (
  2. COVID-19 deaths in children and young people in England, March 2020 to December 2021: An active prospective national surveillance study | PLOS Medicine:
  3. Significantly lower infection fatality rates associated with SARS-CoV-2 Omicron (B.1.1.529) infection in children and young people: active, prospective national surveillance, January-March 2022, England – Journal of Infection:


Dr Alasdair Munro, Clinical Research Fellow in Paediatric Infectious Diseases, NIHR Southampton Clinical Research Facility, and University of Southampton, said:

“This study attempts to address the contribution of COVID-19 to child mortality in the USA during the latter pandemic period, reassuringly finding it was responsible for only 2% of deaths between August 2021 and July 2022.  This is similar to data from England, finding COVID-19 to be responsible for 1.2% of deaths (  There are some important methodological considerations.

“Comparison of mortality statistics is inherently difficult, especially when attempting to create rank orders.  Decisions about grouping mortality risks (such as accidental injury) make a huge difference; given these risks so disproportionately outweigh most others (including COVID-19), simply separating them into different categories of accident could push other categories out of the “leading causes”.  Should one separate different causes of infection but group all accidents together?  This is a difficult methodological issue.  Due to these difficulties, stating risks in rank order is somewhat arbitrary and often of limited value.

“It is critical to stratify by comorbidities when assessing mortality risks from COVID-19, as previously studies have found the overwhelming majority of deaths to have occurred in a select group of children with underlying conditions (  This vulnerable group of children deserve special consideration in these analyses, particularly as making policy for the general population on this basis may not be appropriate.

“Despite these significant limitations, this descriptive study does highlight the low risk of death to children from COVID-19 during the pandemic period.  This is particularly noteworthy as the severity of illness across all except the very youngest age groups has fallen dramatically during the study period due to acquisition of immunity (through both vaccination and infection).  In addition, as mortality risk is heavily concentrated in vulnerable in children with significant comorbidities who now have access to vaccination, we can expect the risk to be even lower in future than the already low risk demonstrated in this study.”



‘Assessment of COVID-19 as the Underlying Cause of Death Among Children and Young People Aged 0 to 19 Years in the US’ by Seth Flaxman et al. was published in JAMA Network Open at 16:00 UK time on Monday 30 January 2023.

DOI: 10.1001/jamanetworkopen.2022.53590



Declared interests

Prof Sir Andrew Pollard: “This comment is made in a personal capacity.  Andrew Pollard is chair of the UK Department of Health and Social Care’s (DHSC) Joint Committee on Vaccination and Immunisation (JCVI), but does not chair or participate in the JCVI coronavirus committee, and was a member of the World Health Organization’s (WHO) Strategic Advisory Group of Experts until the beginning of 2022.  He has received research funding for coronavirus vaccine research from UKRI, CEPI and NIHR.  Oxford University has entered into a partnership with Astra Zeneca for the development of a coronavirus vaccine.”

Dr Michael Absoud: “I have no conflicts of interest.”

Dr Alasdair Munro: “No conflicts to declare.”

in this section

filter RoundUps by year

search by tag