Swine flu deaths examined

Friday December 11 2009

Swine flu is less lethal than feared, according to reports on the BBC News. The website quotes research led by Sir Liam Donaldson, the chief medical officer, which found that only 0.026% of those infected with swine flu have died.

This was a well-conducted investigation into all deaths in England attributed to swine flu up to November 2009. It found that there were 138 deaths from an estimated 540,000 cases, or around 26 deaths per 100,000 people. Because of the nature of swine flu and the difficulties in identifying cases, there remains the unavoidable possibility that these mortality rates were slightly over- or underestimated. For example, some deaths associated with swine flu may not have been correctly identified, which would lead to an underestimation of the mortality rate. Equally, underestimations of the total number of swine flu cases could lead to overestimation of the mortality rates. However, the figures calculated in this study give the best estimate based on available information.

This research also suggests that mortality rates are much lower than in 20th century flu pandemics. However, vigilance and the prompt use of antivirals when indicated, in addition to vaccination, remain highly important.

Where did the story come from?

This research was conducted by Sir Liam Donaldson, chief medical officer for England, and colleagues of the Department of Health and Health Protection Agency. This work was conducted as part of the public health response to pandemic influenza in England, and no additional funding was sought. The research was published in the peer-reviewed British Medical Journal.

This research received coverage from The Guardian, the Daily Mirror and BBC News, which have all accurately reported the figures calculated and the nature of this research.

What kind of research was this?

This was a cross-sectional investigation of all reported deaths related to the H1N1 (swine flu) pandemic. The data was obtained through compulsory reporting systems used by general practices and hospitals.

This research has investigated all deaths considered to be related to flu, which may together be expected to provide accurate projections of flu-related mortality rates. Broadly speaking, overall mortality rates are calculated by dividing the total number of deaths by the total number of cases. It is possible that the death rates calculated could have some unavoidable inaccuracy. They could either be a slight underestimate if there have been other deaths that have not been correctly attributed to swine flu, or a slight overestimate if the total number of swine flu cases has been underestimated. 

This study used data from medical reporting systems. An alternative way of measuring flu-related mortality would be solely to examine death certificates. However, this would probably involve some delay, meaning that statistics derived from them would not be up to date or may not be available soon enough to aid planning. Also, the accuracy of the figures obtained in this way would be subject to correct identification of flu as the primary cause of death as well as the accurate completion of death certificates.

What did the research involve?

Since July 2009, hospitals have been required to report all suspected and confirmed cases of swine flu death to the Department of Health. Deaths in hospitals prior to this time were identified using the Health Protection Agency’s flu reference centre and public health department records. From August 2009, a separate reporting system has been used for deaths occurring in the community, such as those occurring in the home.

All identified deaths were followed up by contacting the senior physician involved in reporting the death. Deaths were considered to have been caused by swine flu if this was the cause of death listed on the death certificate, or if swine flu had been confirmed by a laboratory before or after death. Underlying medical conditions, duration of illness and the use of flu drugs were taken into account. The researchers also considered how many of those who had died would have been eligible for the vaccination once it had arrived.

The researchers estimated the number of cases of flu per age group based on the proportion of suspected cases that were confirmed in laboratory tests, GP consultation rates for flu, population estimates from the Office for National Statistics and the estimated numbers that would not consult a GP. The researchers then calculated the specific case fatality rate per age group.

This was well-conducted research that has used the best possible methods to gain an accurate estimate of death rates of flu in England.

What were the basic results?

There were reportedly 138 deaths in England definitely attributable to swine flu between June and November 2009. An estimated 540,000 people had symptomatic flu during this period in England (around 1% of the population). From these figures, the estimated mortality rate was 26 deaths (range 11 to 66) per 100,000 people who had swine flu, or 0.026% of those affected. There was no difference between males and females.

The lowest death rate was in children aged five to 14, at 11 deaths per 100,000 cases. The highest rates were for those aged 65 years or over, at 980 deaths per 100,000 cases.

Of the 138 who died, the average age (median/middle) at time of death was 39 years. Fifty of these, or just over a third, (36%) were in previous good health with no, or only mild, pre-existing illness. However, two-thirds had either severe underlying disease (33%) or incapacitating prior illness (30%). Three quarters of those who died (108; 78%) had been prescribed antiviral drugs prior to death, but of these, 82 (76%) did not begin taking them within the first 48 hours of illness, as is recommended.

Death was on average 12 days after symptoms began. There was a peak of deaths in July, and a second wave in October and November. Had these people survived, 67% of them would have been eligible for the newly developed flu vaccine.

How did the researchers interpret the results?

The researchers conclude that mortality levels in this pandemic compare favourably with rates of death in the influenza pandemics of the 20th century. However, they say that public health measures are still required even though mortality rates appear low, and that the vaccination of high-risk groups remains a priority. They also say that as a third of the deaths occurred in previously healthy people, there is a case for extending the vaccination programme and for continuing to make early antiviral treatment widely available.


The H1N1 flu pandemic was announced by the World Health Organization on June 11 2009, after the strain first emerged in Mexico in March. So far, this well-conducted investigation into all deaths in England known to be attributable to swine flu from July-November 2009 has found 138 deaths out of an estimated 540,000 cases, which is about 26 deaths per 100,000 people (0.026% of those affected).

Other points to note:

  • As the researchers say, the swine flu mortality rate (0.026%) seems much lower than in 20th century flu pandemics. They report that the rate in the 1918-9 H1N1 pandemic was 2-3%, and about 0.2% in the subsequent pandemics (1957-8 and 1967-8).
  • About two-thirds of those who died had significant underlying illness, but there was still a third who did not have underlying illness, which supports the case for vigilance against flu across all age groups and for all health conditions.
  • The high fatality rate in the elderly and in people with conditions that increase the risk of death (comorbidities) suggests that these groups are appropriate priority groups for vaccination. Additionally, although there is a low proportion of child swine flu cases that prove fatal (0.011%), a large number of children have been affected by swine flu.
  • The fact that the majority of the patients who died despite taking Tamiflu had received this medication more than 48 hours after the onset of illness appears to support the timely use of antivirals (within 48 hours). However, as the authors point out, this conclusion is limited by the absence of a control group who did not take an antiviral.
  • The research has used good methods to try to obtain accurate mortality numbers and accurate estimates of the total number of people who will be affected by swine flu. However, there remains the unavoidable possibility of mortality rates either being slightly underestimated if there have been deaths associated with swine flu where the virus was not recorded as the primary cause of death, or a slight overestimate due to underestimation in the total number of swine flu cases. Any deaths occurring in the private sector would also not have been reported through NHS systems (although this number could be expected to be very small).

This research does appear to show that death rates in the swine flu pandemic are lower than previously anticipated, but that vigilance and the prompt use of antivirals when indicated, in addition to the vaccination schedule, remain important.

Analysis by Bazian
Edited by NHS Choices