Heart and lungs

Child swine flu deaths analysed

“The flu pandemic in England killed 70 children in 2009,” The Guardian has reported. The newspaper says that “most of those who died had pre-existing health problems but one in five were healthy before they caught the virus”.

The news story was based on a report that examined all of England’s swine flu deaths in children under 18 during the 2009 pandemic. The research found that there was a higher risk of dying from the influenza A H1N1 strain of flu among certain groups of children, such as those with pre-existing conditions and those of Pakistani or Bangladeshi backgrounds. However, the reasons behind this ethnic bias were not determined.

This report has illustrated that there may be children who are more at risk of dying from certain flu strains, a phenomenon that will require further research. Also, this research was carried out in England where the overall proportion of children who died during the pandemic was low, at a rate of six per million people. The researchers have now called for an analysis of global data on childhood mortality to help us further understand and prevent childhood deaths from seasonal and pandemic flu.

Where did the story come from?

The study was carried out by researchers from the peer-reviewed medical journal The Lancet.

This research was covered fairly by The Guardian and The Daily Telegraph. Both papers discuss the implications for vaccination programmes and for the early use of antiviral treatments.

What kind of research was this?

This was an observational, population-based study that aimed to analyse child deaths related to pandemic influenza A H1N1 (swine flu) in England in order to inform clinical and public health policies relating to seasonal and pandemic flu.

The researchers said that the overall death rate from seasonal flu is low, and that it predominantly affects people above 65 years old. However, the recent swine flu pandemic affected children disproportionately and, despite global reports of the complications associated with the pandemic influenza A H1N1 virus, they are not aware of a detailed analysis focusing on the effects that it had on children.

What did the research involve?

During the flu pandemic, reporting systems were established so that all suspected and confirmed deaths from pandemic influenza A H1N1 in England could be recorded. Further deaths were identified through cross-checking of records held by the Regional Directors of Public Health and by the Health Protection Agency’s influenza reference centres.

All cases of death where the influenza A H1N1 virus was suspected were assessed by a member of the Chief Medical Officer’s clinical team. A death was related to influenza A H1N1 if there was laboratory evidence of infection with this virus or if H1N1 infection was recorded on the death certificate.

From the records, the researchers identified all deaths in children under 18 years old. A paediatrician from the Chief Medical Officer’s team interviewed the child’s doctor about pre-existing disorders and medical history of the child, their symptoms and the clinical course of their flu. The researchers also collected demographic information about the child.

What were the basic results?

A total of 70 child deaths related to pandemic influenza A H1N1 occurred in England between June 2009 and March 2010. All of these cases were confirmed by laboratory testing. This corresponds to a rate of six per million of the population.

There were a similar number of boys (31) and girls (39) who had died. Deaths were reported in children aged between 3 months and 17 years, with an average (median) age at death of 7 years.

Six of the children who died were Bangladeshi or British Bangladeshi. This corresponds to a rate of 47 deaths per million of the Bangladeshi population in the UK. There were also 11 deaths in Pakistani or British Pakistani children, which corresponds to a rate of 36 per million of the population, plus 37 deaths in white British children (4 per million of the white population). There were no differences in pre-existing health conditions between the children from these three ethnic groups.

Analysis of the 70 deaths also show that:

  • 25 deaths (64%) were in children with severe pre-existing disorders
  • 15 deaths (21%) were in children who were previously healthy
  • half of the children who died had either pre-existing chronic neurological, gastrointestinal or respiratory disease
  • 19 of the children had spastic cerebral palsy affecting all of their limbs
  • 11 children had stomach problems
  • 41 of the children had conditions that required regular feeding through a tube
  • 5 children had asthma
  • 8 children had a pre-existing heart condition
  • After taking into account the different prevalence of pre-existing disorders, having a chronic neurological disorder was associated with the greatest risk of death.

Among the 70 deaths, 19 occurred before the children could be admitted to hospital. Children in this group were more likely to have been healthy or had only mild pre-existing disorders than those who died after admission to a hospital.

Forty-five of the 70 children received the antiviral drug oseltamivir (Tamiflu). Seven of the children received Tamiflu within 48 hours of the onset of their symptoms. On average (median), the children received Tamiflu five days after the onset of their symptoms. The latest that Tamiflu was given was on the seventh day after symptom onset. Two of the 45 children who received Tamiflu had swine flu that was resistant to the drug.

How did the researchers interpret the results?

The researchers said that of the 70 deaths in children in England related to pandemic influenza A H1N1, “mortality disproportionately affected ethnic minorities and those with pre-existing disorders”. They also say that “many deaths occurred before hospital admission and in healthy children or those with only mild pre-existing disorders”. They highlight the fact that the 70 child deaths recorded is greater than the number of children killed by leukaemia each year.

The researchers said that the high population mortality rates observed in Bangladeshi and Pakistani Britons might be attributable to clustering of the virus in London and the West Midlands. But they also highlight that there was a lower proportion of these ethnic groups in other areas that had high numbers of flu cases, such as the East Midlands and Yorkshire. The researchers called for further investigation as to why the death rates were higher in these groups.

The researchers said that although antiviral use for the treatment of influenza in children is controversial, the drugs are most effective if given within 48 hours of treatment. They say that their study was not designed to assess antiviral use but suggest that “early treatment with antiviral therapy may maximise the effectiveness of the treatment”, and that “further investigation into the contribution of pre-hospital antivirals to the outcome of affected children is needed”.

The researchers also suggested that their findings support the vaccination of children against pandemic influenza A H1N1.

Conclusion

This is a useful report, which has analysed child deaths related to pandemic influenza A H1N1 in England. It identified that there may be certain children who were more at risk from this strain of flu than others. However, the researchers acknowledge that overall there was a small number of child deaths in the UK related to the pandemic, therefore an international study that pooled data from a larger, world-wide number of children who had died would be greatly informative. Such as study might improve understanding of what factors increase the risk of death in children following exposure to similar flu strains.

The authors point to a few limitations of the research, noting that correctly recording deaths can be difficult, especially when classifying deaths as having occurred pre- or post hospital admission. They took steps to record this as accurately as possible and to take into account those children who may not have sought medical help.

This report has raised questions that require further follow-up, such as why particular ethnic groups had an increased mortality rate, and what is the most appropriate antiviral treatment plan for children. It also highlighted that some pre-existing conditions carried a higher risk of death from this flu strain. This will need to be addressed when planning a response to future pandemics.


NHS Attribution