Differing patterns of influenza activity in the southern hemisphere during and between the 2009 pandemic and the 2010 winter influenza season – the usefulness for EuropeArchived

ECDC comment

This is timely comparison carried out at the end of the 2010 southern hemisphere winter.

2009 versus 2010 comparison of influenza activity in southern hemisphere temperate countriesVan Kerkhove MD and Mounts AW Influenza and Other Respiratory Viruses; no. doi: 10.1111/j.1750-2659.2011.00241.x

This is timely comparison carried out at the end of the 2010 southern hemisphere winter. The authors looked back at the first ordinary influenza season was after the 2009 pandemic (the pandemic was officially declared over by the World Health Organisation (WHO) in mid 2010) and evidence of changes in behaviour and severity of the 2009 influenza A(H1N1)2009  virus itself. The authors, from the WHO and a European country, took advantage of the southern hemisphere being the first (ahead of the northern hemisphere) in experiencing a post-pandemic influenza season. They were also able to compare how pandemic virus was behaving in its second season in diverse countries where it first appeared at the start of the 2009 winter period of transmission. The methodology used by the authors consisted of using data from WHO’s FluNet and ministries of health in Argentina, Chile, South Africa, Australia and New Zealand to compare five different distinct features between two completed transmission seasons – the 2009 pandemic season and the 2010 winter epidemic season. These comparison features were:

  • type of circulating influenza viruses;
  • time course;
  • geographic distribution of activity peaks;
  • impact of the disease on health-care systems;
  • effect on different risk groups.

Virus circulationIn 2009 the influenza A(H1N1)2009 pandemic virus had circulated freely in the five  temperate countries at  the start of their usual seasons (early summer 2009) and quickly became dominant. In contrast, the second season showed a more diverse pattern among the five countries. In New Zealand and Australia, the 2009 pandemic virus was predominant, but it was not so in Chile and Argentina where mostly seasonal A(H3N2) and influenza B viruses were detected in the former and mostly influenza B viruses in the latter. The 2010 seasonal epidemic in South Africa had three smaller waves of influenza B, A(H3N2) and 2009 pandemic virus co-circulating simultaneously. Notwithstanding these differences there were common findings. All of the 2009 pandemic viruses detected in all five countries were antigenically similar to the original wild strain in April 2009 (A/California/7/2009) and had seemingly totally eclipsed the previously circulating seasonal A(H1N1) viruses. Also none of the many samples analysed had markers of resistance to oseltamivir.

Time courseThe peak activity during the 2009 influenza season – measured by influenza-like illness (ILI) reported cases per week (Chile), influenza cases per week (Argentina) or weekly ILI consultation rates (Australia and New Zealand) – occurred in a typical southern hemisphere winter pattern. it followed a biphasic pattern in South Africa.

Geographic distribution of activity peaksThe authors found the geographic distribution of symptomatic influenza cases in all five countries similar during the 2009 and 2010 seasons, with uneven geographic distribution reported throughout each country over the course of the influenza season. A notable finding in New Zealand was that areas relatively spared in the pandemic season seemed more affected in 2010.

Impact of the disease on health-care systemsThe laboratory confirmed cases were reported in all five countries in 2010 was an order of magnitude lower than the figures reported a year earlier. Similarly, the reported number of deaths due to the 2009 pandemic virus in 2010 has also been considerable lower than the number reported in 2009. Overall, the impact of influenza in terms of severity of illness, mortality and impact on health-care systems was lower in the 2010 season than in the 2009 pandemic.

Effect on different risk groupsThe 2009 pandemic in the southern hemisphere was also characterised by a high attack rate, especially in young people. Moreover, individuals with chronic illnesses, pregnant women, very young individuals and individuals belonging to certain ethnic minority and disadvantaged populations (Maori and Pacific Island populations in New Zealand) were particularly susceptible to developing severe complications during 2009. This was still the case during 2010. There was some evidence that the proportion of infected persons who were hospitalized, admitted to ICU or died among the influenza A(H1N1)2009 pandemic confirmed cases may have increased in 2010 compared to  2009. At the same time the overall attack rate and the impact in terms of severe illness was considerably lower in 2010. The authors suggest three possible arguments to explain these findings: the high rates of infection in 2009, relatively high vaccination coverage in many of the areas under observation during the performance of the study and the fact that the virus has not perceptibly changed antigenically since April 2009.

ECDC Comment (01st April 2011):This invaluable and timely summary from the five southern hemisphere countries for their 2010 winter shows some important consistencies while at the same time there are notable country-specific patterns and differences.  As long as the Southern Hemisphere is running ahead of the northern hemisphere it will be very useful for Europe and the rest of the northern hemisphere.  The most relevant surveillance analyses for 2010 have been illustrated graphically in a series of ECDC global updates over the summer of last year the last of which was published on the 14th October 2010. However this is the first international summary in a journal. These analyses have also been summarised in Table 1 below and contributed to ECDC’s Forward look 2009 pandemic risk assessment published on its website in October 2010. Across all the countries, levels of transmission in the community as evidenced by primary care surveillance were considerably down on what was seen in the pandemic winter. This is unlikely to be simply due to fewer people seeking care as the number of influenza-related hospitalisations reported and deaths were also considerably down on what was seen in 2009. The volume of laboratory isolations needs to be interpreted cautiously since there was intense care seeking and testing in 2009, far less so in 2010. However, beyond that, there are a number of important national variations:

Australia – In this country, in 2010, the age distribution of influenza laboratory reports were intermediate between those of 2008 and 2009 being relatively consistent across all age groups but decreasing with increasing age so that, as in both the pandemic winter and the 2010 winter, older people were under-represented (at least in laboratory reports). After a quiet initial period, the rates of consultations with ILI confirmed cases and the proportion of samples from sentinel physicians all rose in parts of Australia in late winter, suggesting a late season rise in transmission. Chile – Chile is one of the few countries outside Europe that reports on testing for respiratory syncytial virus (RSV). As in some other seasons, it experienced early winter season epidemics of RSV in young children resulting in considerable illness in the community and hospital admissions. As in Australia, there was an early season but followed by a late season rise associated with influenza A(H3N2).New Zealand – New Zealand had a particularly interesting pattern as it seemed to anticipate what Europe experienced [1]. While, like other countries, overall rates of transmission in the community were lower than in the pandemic winter (as were reported hospital admissions associated with influenza), there was intense influenza transmission in places with numerous hospital admissions in those localities. Overall, there were considerable numbers of hospital admissions but the reduction was not as notable as in the four other countries. It was observed that these ‘hot spots’ tended to be areas that had had less influenza transmission during the 2009 pandemic winter season. Some support for this came from a national serosurvey that showed heterogeneity of immunity, more so than in other countries [2, 3]. New Zealand has also published its own useful summary of its 2010 experience [1]. Unlike in Australia and Chile, there was no late season rise in transmission.South Africa – The pattern in this country was unusual in that there were hardly any isolations of the pandemic virus in 2010. Influenza A(H3N2) and B viruses predominated, a pattern that the WHO reported as consistent, at least in relation to A(H3N2), with other parts of southern and eastern Africa for which data are available.

Relevance for Europe – The relevance of this diversity of experiences is hard to speculate on. It is a reminder of the important differences between the northern and southern temperate zones, and especially Europe. The southern hemisphere temperate countries are somewhat isolated from each other and can, like this, show different patterns; in contrast, such variation is less common within Europe, although it can be seen between Europe, North America and North Asia. Finally the finding of little oseltamivir resistance in the Southern Hemisphere needs to be interpreted cautiously as this can emerge rapidly [4, 5].

Table 1 – Influenza pattern in five southern hemisphere temperate countries in their 2010 winter


Community epidemiology and impact 


Argentina (limited data)

Lower observed rates than in 2009 – fewer severe cases

Mostly A(H1N1)2009


Lower observed rates than in 2009 – fewer severe cases Late season rise due to a mix of viruses

A(H1N1)2009 then some B & A(H3N2)


Lower observed rates than in 2009 – fewer severe cases Late season rise mostly due to A(H3N2)

A(H1N1) 2009 then B & A(H3N2) Early epidemics of RSV

New Zealand

Local observed rates higher than in 2009 – fewer severe cases than in 2009 but straining hospital services in some areas

Almost entirely A(H1N1) 2009

South Africa

Lower observed rates than in 2009 – fewer severe cases

A(H3N2) & B – no A(H1N1)2009