Bulletin of the World Health Organization

The classical definition of a pandemic is not elusive

Heath Kelly a

a. Victorian Infectious Diseases Reference Laboratory, Department of Epidemiology, Locked Bag 815, Carlton South, Vic. 3053, Australia (e-mail: heath.kelly@mh.org.au).

Bulletin of the World Health Organization 2011;89:540-541. doi: 10.2471/BLT.11.088815

Doshi argues cogently that the definition of pandemic influenza in 2009 was elusive but does not refer to the classical epidemiological definition of a pandemic.1 A pandemic is defined as “an epidemic occurring worldwide, or over a very wide area, crossing international boundaries and usually affecting a large number of people”.2 The classical definition includes nothing about population immunity, virology or disease severity. By this definition, pandemics can be said to occur annually in each of the temperate southern and northern hemispheres, given that seasonal epidemics cross international boundaries and affect a large number of people. However, seasonal epidemics are not considered pandemics.

A true influenza pandemic occurs when almost simultaneous transmission takes place worldwide. In the case of pandemic influenza A(H1N1), widespread transmission was documented in both hemispheres between April and September 2009. Transmission occurred early in the influenza season in the temperate southern hemisphere but out of season in the northern hemisphere. This out-of-season transmission is what characterizes an influenza pandemic, as distinct from a pandemic due to another type of virus.

Simultaneous worldwide transmission of influenza is sufficient to define an influenza pandemic and is consistent with the classical definition of “an epidemic occurring worldwide”. There is then ample opportunity to further describe the potential range of influenza pandemics in terms of transmissibility and disease severity. The emerging evidence for A(H1N1) is that transmissibility, as estimated by the effective reproduction number (R, or average number of people infected by a single infectious person) ranged from 1.2 to 1.3 for the general population but was around 1.5 in children (Kathryn Glass, Australian National University, personal communication). Some early estimates of R for pandemic influenza H1N1 2009 may have been overestimated.3

Severity, as estimated by the case fatality ratio, probably ranged from 0.01 to 0.03%.46 These values are very similar to those normally seen in the case of seasonal influenza.7,8 However, the number of deaths was higher in younger people, a recognized feature of previous influenza pandemics.9

It is tempting to surmise that the complicated pandemic definitions used by the World Health Organization (WHO) and the Centers for Disease Control and Prevention of the United States of America involved severity1,10 in a deliberate attempt to garner political attention and financial support for pandemic preparedness. As noted by Doshi, the perceived need for this support can be understood given concerns about influenza A(H5N1) and the severe acute respiratory syndrome (SARS). However, conflating spread and severity allowed the suggestion that 2009 A(H1N1) was not a pandemic. It was, in fact, a classical pandemic, only much less severe than many had anticipated or were prepared to acknowledge, even as the evidence accumulated.

In 2009 WHO declared a pandemic several weeks after the criteria for the definition of a classical pandemic had been met. Part of the delay was no doubt related to the nexus between the formal declaration of a pandemic and the manufacture of a pandemic-specific vaccine. If a classical pandemic definition had been used, linking the declaration to vaccine production would have been unnecessary. This could have been done with a severity index and, depending on the availability and quality of the emerging evidence on severity, a pandemic specific vaccine may have been deemed unnecessary. Alternatively authorities may have decided to order vaccine in much smaller quantities.

The response to A(H1N1) has been justified as being precautionary, but a precautionary response should be rational and proportionate and should have reasonable chances of success. We have argued that the population-based public health responses in Australia and, by implication, elsewhere, were not likely to succeed.11 Similarly, the authors of the draft report on the response to the International Health Regulations during the 2009 pandemic note that what happened during the pandemic reflected the activity of the virus and, by implication, not the interventions.10

Risk is assessed by anticipation of severity and precaution should be calibrated to risk. As Doshi has argued, we need to redefine pandemic influenza. We can then describe the potential severity range of future pandemics. Finally, we need to use evidence to assess severity early to anticipate risk.

Competing interests:

None declared.