In this edition of Eurosurveillance there is an important report of one highly pathogenic avian influenza virus (HPAI type A/H7N7) that affected humans during the 2003 poultry epidemic in the Netherlands and Belgium [1]. The human infections were mostly among those working to control the infection, and their families. In response, ECDC, together with an expert group, has produced interim occupational guidance for Europe that will reduce the risk [2]. However, this experience also emphasises the variability in the influenza virus families. While H7N7 was quite infectious for humans and showed measurable person to person transmission, another better known avian influenza, A/H5N1, is quite different, as it currently seems to infect humans only rarely and human-to-human transmission seems to be even rarer [3].
The year’s end is traditionally a time for reflection, and I would like to propose five fundamental questions about pandemic risk for the start of 2006

Has the risk from avian and pandemic influenza been exaggerated?

The answer to this question must be both ‘Yes’ and ‘No’. In the autumn of 2005, when H5N1 appeared on the borders of Europe in Romania, Turkey and Croatia, there was suddenly massive public interest and gross confusion of three separate, although related, influenza types: Seasonal Influenza, Avian Influenza and Pandemic Influenza. [see ECDC website http://www.ecdc.europa.eu/influenza/factsheet_influenza.php for definitions] Even the serious media’s presentations of the situation gave the strong impression that a human influenza pandemic was about to start, and that the pandemic virus would probably be brought to Europe by migrating birds. If there is no pandemic in 2006, members of the lay public may reasonably feel that some authorities have been ‘crying wolf’. Although official statements have been mostly measured and accurate, the reporting of those statements tended to exaggerate the current threat from H5N1. The reality is thus far in its evolution, the family of H5N1 viruses available for study are avian viruses that are poorly adapted to humans, for whom they are not very infectious, but highly pathogenic in those few humans that they do infect [4,5]. That low risk (of becoming infected) - high risk (of severe disease if you are infected) message is a difficult one for risk communicators to convey. Occasionally, H5N1 transmits on from one human to another, but none of the viruses at present seem to represent a pandemic strain, as their reproductive rate in humans (Ro) is far below unity [3,6].

That is not to say that the H5N1 viruses are without social impact. In Thailand they have prejudiced that country’s economically important export trade in poultry products to Europe and Japan. For societies like China and Vietnam where poultry are key to food security, the threat to the rural communities is considerably greater [7]. It is not surprising that both China and Vietnam are turning to the potentially risky measure of poultry immunisation as a measure to protect their huge flocks. This is a massive task. It is estimated that at any moment China’s human population of 1.3 billion keeps around four billion domestic birds (point prevalence) and that each year they require fourteen billion domestic birds (period prevalence).

The main threat, however, is of a human pandemic. Any pandemic would represent a major risk to human health and a threat to social functioning worldwide. The two lesser pandemics of the 20th century (1957 and 1968) are each estimated to have killed between one and four million people worldwide. A pandemic on the scale of 1918-1919 (at least 20 million deaths) would be catastrophic [8). Arguably, the interconnected industrialised world of today is more vulnerable to a pandemic than it was even forty years ago. Not only is there much more international travel to spread infection, but societies are more dependent for daily existence on goods and services that are produced elsewhere. Efficient ‘just in time’ stockkeeping systems, e.g. for food, will be vulnerable to the sudden mass illness in production and distribution staff that would take place in a pandemic. It is estimated that for short periods at the height of a pandemic up to 20% of working adults might be unavailable for work, because they are ill with influenza, or caring for others who are ill, or simply out of fear of infection. Fortunately, these periods of intense illness will not occur everywhere at the same time, but the disruption could nevertheless be considerable.

Will the next pandemic be due to H5N1?

We do not know. Pandemics occur through the emergence of a new strain of influenza virus which can infect and is pathogenic to humans, to which there is little pre-existing immunity and which can transmit readily from person to person. This is thought to happen by one of two mechanisms. Either through two pre-existing influenza virus types exchanging genetic material (recombination) or spontaneous genetic shift (mutation) from a single pre-existing influenza strain. Could H5N1 do either? It is certainly a candidate for a pandemic strain, as it can infect humans and is highly pathogenic. Some have argued that it only needs to make the final step of efficient person to person transmission, and WHO has set its global scale at Pandemic Alert Phase 3, the last phase before efficient human-to-human transmission. Others, however, consider that the next pandemic is equally if not more likely to come from a low pathogenicity avian influenza, such as H9N2 [5]. None of the three pandemics of the 20th century were based on a H5 strain, and H5N1 has been around at least since 1996 without a pandemic having resulted. It is also relatively uninfectious for humans, unlike the H7N7 strain observed by De Ry et al [1]. At the same time, H5N1 has spread massively, with the result that there are outbreaks in poultry in many East and South East Asian countries, including the huge bird populations of China. Although recombination involving H5N1 has not yet been detected, the possibility of it happening must have increased. H5N1 is not a uniform strain, but rather a large and complex family of viruses, and one of these may eventually mix and exchange genetic material with a transmissible human influenza [9 ]. However further risk assessments to determine whether or not H5N1 will cause a pandemic are of less value than making preparations for a pandemic due to H5N1 or any other influenza virus.

How bad will the pandemic be and what will be its characteristics?

Again, we do not know. Pandemics are not standard. The three 20th century pandemics varied not only in their driving viruses and scale, but also in their characteristics. For example, the 1918-1919 pandemics affected young adults in particular, while the later epidemics more often affected the elderly. We cannot assume that the next pandemic will be driven by transmission in particular groups, and data that can only be derived during the actual pandemic must guide interventions. It could be that workplace transmission will be crucial or that transmission among school-age or younger children will predominate. When a pandemic happens, the two most important investigations will be isolating the virus (to develop tests and the pandemic vaccine) and carrying out early quick, focused epidemiological studies at the sites of first outbreaks, both in Europe and beyond (to determine basic parameters such as mode of transmission, age-specific attack rates, and case-fatality rates, to guide countermeasures) The analogy with the evidence-based approach to controlling SARS is clear [10].

What role will antivirals play during a pandemic and how big a stockpile should countries have?

There is a danger that the availability of antivirals (especially oseltamivir) dominates thinking and preparations for a pandemic [11]. A detailed and rational approach to the use of antivirals in a pandemic has yet to be determined. Hospital doctors will, quite reasonably, expect to have available antivirals to treat those requiring hospitalisation, although it will be impossible to know ahead of time whether they will be effective at later stages in a patient’s illness. Some countries are planning to have national stockpiles. However, simply having a stockpile is not enough, and if one European country has a stockpile ten times larger than its neighbour, it cannot be therefore judged to be ten times better prepared. Since in order to be effective in treatment of influenza, antivirals must be delivered within 48 if not 12 hours of symptom onset, it can be seen that mass delivery to populations will be a major issue. A stockpile without a rapid delivery system will provide little protection. Some have proposed that there be a European Union stockpile of antivirals. A modest European stockpile could for example assist in protecting workers during poultry outbreaks close to Europe[1,2]. It would also be an asset in the unlikely event that the next pandemic started in or near Europe, so that WHO’s stamping out tactic could at least be attempted, supposing the existence of a practical plan to do so [12]. However, rapid development and production of a pandemic vaccine will probably be more important for the second wave, with the more distant hope of more cross-protective vaccines that would protect against pandemic first waves (so-called universal vaccines) [13]. Equally important and more immediately accessible will be the simple public health measures (early self-isolation of those with symptoms, handwashing, respiratory hygiene, etc.) that are already available, and will save lives [14].

Is Europe prepared for a pandemic?

Not as prepared as it could or should be. Six national assessments have been undertaken by countries using a standard assessment tool and working with teams from ECDC, the European Commission and WHO European Region. These assessments (which will continue in 2006) found that while all six countries were preparing rapidly, all also had considerable way yet to go. Major issues remain to be addressed, notably the need for preparations to extend outside the health sector alone and for plans to be made more operational [15].

In conclusion, the threat from a pandemic has not been exaggerated. It could happen in 2006 from H5N1, or, more likely, in the future, and with another strain. However, in 2005 most European authorities and politicians started to give the risks the serious attention they deserve, and to invest the necessary resources to develop countermeasures. It is to be hoped that as the media interest inevitably declines, those in authority will sustain the investment and the levels of preparatory activity. Certainly, the pandemic risk will not decline.

* Angus Nicoll is a Seconded National Expert at ECDC where he coordinates its influenza activity. He has been seconded from the Health Protection Agency (HPA) and his position is supported by the UK Government. The opinions above are personal and do not necessarily represent the views of either ECDC or the HPA