Our study demonstrated similarities and differences in the 2009-2010 pandemic influenza season when compared to previous influenza seasons in the Region. During the 2009-2010 influenza season, influenza activity peaked in the Region much earlier than in five previous seasons. During the 2009-2010 season there were also significantly higher clinical consultation rates of ILI/ARI observed in younger populations. Similar to some of the previous seasons, the influenza virus followed a west to east direction in the western part of the Region, and the duration of the winter wave of the pandemic influenza was short, but similar to some other prior seasons.
The activity of pandemic influenza peaked very early in the Region most probably due to the emergence of the novel virus. This is consistent with the notion that susceptibility of the population may play a more important role that the other factors, i.e. absolute humidity [22, 23], variation in vitamin D levels related to the amount of sunlight , and patterns of social mixing , associated with the seasonality of influenza outbreaks in temperate climates . A similarly early peak in influenza season in the WHO European Region was observed during the 2003-04 season, when the Fujian strain of A(H3N2) influenza virus was the predominant virus circulating in Europe . During this season, the pre-existed immunity in the general population was also limited due to a reassortment event in the A(H3N2) influenza virus [28, 29].
The duration of epidemic period in the Region during pandemic influenza season was not shorter that during some other recent influenza seasons. This could be partially explained by the large size of the Region and a minimum time period required for the virus to spread to all countries. In addition, the secondary attack rate of pandemic A(H1N1) 2009 virus has been shown to be generally similar to the one usually observed for seasonal influenza viruses [30–32], contributing to the similar duration of influenza seasons.
We observed a moderate west- east spread of the influenza virus in the Western part of the Region. In the last eight years this direction was previously observed during 5 historical seasons with (co) dominant and generally more virulent A (H3N2) influenza virus . The fact that this direction was observed only in the western part of the Region during the pandemic influenza season, could be explained by more extensive air traffic in this part of the region and its links with USA and Mexico where the first cases of pandemic influenza occurred and spread first to the Spain and United Kingdom in the Region . In addition, the spread of influenza virus in the central and especially western parts of the region might be influenced by neighboring countries in the Middle East and Central Asia. We did not observe south to north spread of the virus in any parts of the Region; this direction of spread was previously found during a few, A(H3N2) dominated influenza seasons in Europe . Monitoring the direction of geographic spread of influenza viruses in the region provides useful information for planning the vaccination, allocation of health care resources and public campaigns.
Although in most of the countries the total consultation rates were not higher when compared to historical seasons, children in 0-4 age (in 10 countries) and in 5-14 age groups (in 11 countries) experienced significantly higher consultation rates compared to past data. Apparently, the younger population had more limited immunity for the pandemic A(H1N1) influenza virus compared to 65+ group in our analyses, who experienced significantly lower consultation rates compared to previous influenza seasons. This could be explained by the previous exposure of older population to the similar A (H1N1) viruses that circulated before the 1957 pandemic [35–39]. It is remarkable that during all three previous influenza pandemics of 20th-century, older populations appeared to be more protected compared to children, probably due to a similar mechanism of the presence of pre-pandemic antibodies .
We found that some countries in the WHO European region experienced higher total or age-specific rates while others did not. The health seeking behavior of the population in different countries might have affected the rates of outpatient consultations for ILI/ARI. However, the percentage of sentinel respiratory specimens tested positive for influenza virus during the pandemic season in individual countries is very similar to the positivity rate observed prior the 2009 influenza pandemic . This suggests that the differences observed between different countries are observed due to the stochastic nature of influenza spread. Differences in impact in different countries are typical for influenza pandemics , and could be explained by the heterogeneity in the degree of immunity in local populations to the pandemic influenza virus, as well as by differences in transmission factors such as local geographic conditions and patterns of social mixing.
Strengths and limitations of the study
To our knowledge, this is the first study comparing the timing, geographic spread and outpatient clinical consultation rates for ILI or ARI of the 2009-2010 pandemic influenza season against historical data for a large number of countries. We used routine sentinel surveillance data, and it has been shown that both ILI and ARI consultations rates provided by countries with sentinel surveillance systems are reliable estimates of influenza activity, e.g. peaks of influenza activity based on ILI or ARI rates are well supported by virological confirmations . Another strong point of the analysis is inclusion of data for 2003-04 influenza season, during which the predominant A(H3N2) virus caused higher consultation rates in children of 0-4 age in several countries .
Sentinel ILI/ARI surveillance data that were used for the current analyses, are collected systematically from a standard number of sites, thereby enhancing the comparability of the 2009-2010 data to those from historical seasons. Although three countries (Kyrgyzstan, the Russian Federation, and Ukraine) in these analyses provided non sentinel epidemiological data, considering that these countries have universal ARI surveillance systems that have been operating for many years, i.e. all ARI cases are being routinely reported, we expect the data provided by these countries to reliably represent influenza activity in their countries.
Although the case definitions for ILI or ARI used by countries are differing, it is important to highlight that for this analysis no between-country comparisons have been made. Instead, data from each of the included countries were compared over the time. When comparing outpatient consultation ILI or ARI rates in each of the countries, we assumed that there were no changes in health care systems of the countries that might have affected the consultation rates over the time. However, the ILI or ARI rates in some of the countries might have been influenced by media bias, i.e. attention that had been paid to the influenza pandemic by mass media could have affected the health seeking behavior resulting in higher consultation rates. Since we have based our analysis on the winter wave of the pandemic and we have excluded the ILI or ARI rates observed during the summer wave, when the consultation rates were most likely to be affected by media bias or panic , we expect less influence on health seeking behavior during the winter wave of the pandemic. On the other hand, a variety of prevention measures had been undertaken by countries during the pandemic influenza season, including vaccination, social distancing, school closures, antiviral treatment and prophylaxis that might have reduced the attack rates [43–45], and subsequently the outpatient consultation rates due to ILI or ARI. If these measures had not been undertaken, we might have observed a greater difference between pandemic and seasonal ILI/ARI rates in the countries of the Region.
The availability of surveillance systems for outpatient consultations rates allowed monitoring of influenza activity during the pandemic in the region and enabled this analysis to improve our understanding of epidemiology of the 2009 pandemic. However, there was no systematically collected data regarding hospitalization across the Region during the previous seasons to enable comparative analysis of risk factors for severe acute respiratory infections. Establishment of such systems in the countries of the WHO European Region is ongoing  and will improve greatly surveillance of future influenza outbreaks.