A PERSPECTIVE ON THE 2009 A/H1N1 INFLUENZA PANDEMIC IN MEXICO

. In this article, we provide a chronological description of the 2009 H1N1 inﬂuenza pandemic in Mexico from the detection of severe respiratory disease among young adults in central Mexico and the identiﬁcation of the novel swine-origin inﬂuenza virus to the response of Mexican public health authorities with the swift implementation of the National Preparedness and Response Plan for Pandemic Inﬂuenza. Furthermore, we review some features of the 2009 H1N1 inﬂuenza pandemic in Mexico in relation to the devastating 1918-1920 inﬂuenza pandemic and discuss opportunities for the application of mathematical modeling in the transmission dynamics of pandemic inﬂuenza. The value of historical data in increasing our understanding of past pandemic events is highlighted.

1. Introduction.Seasonal influenza epidemics generate substantial morbidity and mortality rates every year around the world [1].Novel influenza viruses represent a challenge to public health as they have the potential to trigger influenza pandemics, which are defined as global epidemics associated with substantial morbidity and mortality impact relative to seasonal epidemics.The impact of the 20th century influenza pandemics in 1918-1920 (Spanish influenza), 1957-1959 (Asian influenza) and 1968-1970 (Hong Kong influenza) is not well understood.In particular, the 1918 (H1N1) influenza pandemic has been by far the most devastating of recent history with an estimated global death toll of 20-50 million [2], yet our understanding of its impact in the American continent is still limited although it has been improved by a few quantitative reports from the US [3], Canada [4] and Mexico [5,6,7].
Mexico, the epicenter of the 2009 A/H1N1pdm influenza pandemic, is a geographically and climatologically diverse country with a wide variety of ecological systems ranging from coastal plains, temperate highlands, humid tropical forests, and deserts.Mexico is divided into 31 states and a Federal District (Figure 1).
Past pandemics have exhibited signature features [10], and the ongoing 2009 H1N1 influenza pandemic is not an exception.These signature features include an "age shift" in the proportion of influenza-related mortality towards younger age groups compared to seasonal influenza epidemics.Overall, the age-specific mortality profile can be explained as the result of population immunity patterns arising from prior exposure to related influenza viruses during the population life history [1].The 1918 influenza pandemic in Mexico produced a W-shaped age-specific excess mortality rate curve which highlighted elevated mortality among young adults and senior populations [5].In contrast, early data of the 2009 H1N1 influenza pandemic in Mexico suggested relative protection for persons who were ∼ 60 years of age or older, the age group at risk of exposure to H1N1 strains during childhood before the 1957 pandemic [11], a pattern found to be consistent in other countries (e.g., [15,12,13,14]).
Another signature feature of influenza pandemics is the multiple wave profile of morbidity and mortality over short periods of time often occurring unexpectedly during spring/summer seasons in temperate regions.For instance, the 1918 influenza pandemic was characterized by a multiple wave profile (e.g., New York [3], Geneva, Switzerland [16], England and Wales [17], Copenhagen [18], and Sydney, Australia [19]).Similarly, the recent 2009 H1N1 influenza pandemic also exhibited multiple waves in some countries including Mexico [21], the US, and Japan [37].However, most reports from the southern hemisphere have exhibited a single winter wave (e.g., Chile [23] and Argentina [24], Australia [25], and New Zealand [26]).The wave-like patterns in the temporal distribution of cases are most likely associated with the timing of introduction of the novel virus to different nations relative to school activity periods [27,28,29], climatic conditions [30,31], the strength of the connectivity within countries [32] as well as temporary effects of interventions [33].Significant changes in the virus composition (increasing antigenic drift) could also explain the emergence of multiple pandemic waves.
The ability of a novel influenza virus to spread in a population can be measured by the basic reproduction number (R 0 ), a transmissibility measure defined as the average number of secondary cases generated by a primary infectious individual in a completely susceptible population [34].The transmissibility of the 1918 influenza pandemic has been reported to lie in the range 1.5 to 5.4, depending on the serial interval, location, and pandemic wave considered [35,16,18].Compared to the 1918 influenza pandemic, transmissibility estimates of the early spring wave of the 2009 influenza pandemic in Mexico were relatively low (R 0 ∼ 1.6) [39] within the variability observed for transmissibility levels of inter-pandemic influenza in temperate countries [40], in agreement with a previous estimate for the summer wave of the 1918 influenza pandemic in Geneva [16] and in the lower range for Copenhagen [18].The basic reproduction number of pandemics can be expected to be influenced by the time of the year when the novel virus is introduced in the population with lower effective transmissibility levels when the virus is introduced during school vacation periods due to reduced contact rates.Furthermore, experimental studies suggest that survival and aerosol transmission of the influenza virus improves substantially as absolute humidity levels decrease [30,31].These factors could partially explain the lower disease impact associated with the Spring 2009 H1N1 influenza pandemic wave in Mexico compared to that of the fall wave.
3. The 2009 H1N1 influenza pandemic in Mexico.Early pieces of epidemiological indicators suggested an unexpected rebound of the inter-epidemic influenza season in early March 2008-2009.Several signs alerted Mexican Health Officials about the circulation of atypical respiratory infections in the population when influenza epidemics are not expected to occur.In particular, there were reports of severe respiratory disease during late March and early April in Mexico City and the states of Mexico, San Luis Potosi, Tlaxcala and Oaxaca [41,42].The fact that young adults were being disproportionately affected in the population (Figure 2) was perhaps the most concerning sign to Mexican public health authorities, which is in stark contrast with the typical age-specific risk pattern of seasonal influenza epidemics [11].
An outbreak investigation initially suggested that the beginning of the 2009 influenza pandemic could potentially be traced back to the Town of La Gloria, Perote, in the state of Veracruz, where a large number of pig farms are located.An outbreak of respiratory disease was suspected to have spread through the population of La Gloria from early March through April 6th with an associated attack rate of about 25% [43].However, hospitalizations or deaths were not notified during this epidemic.On March 5th, 2009 health officials of the state of Oaxaca reported the death of a diabetic woman who developed severe pneumonia and initially thought to be infected with Severe Acute Respiratory Syndrome (SARS).This fatal case was later confirmed to be positive for novel H1N1 influenza on April 23rd.On March 27th, the National Institute of Respiratory Diseases (INER) reported the first fatal case of influenza in Mexico City.From March 24th through April 24th a total of 98 people were hospitalized for acute respiratory illness in the INER in Mexico City.It was particularly relevant that most of these hospitalizations at INER were previously healthy young persons [41].
Increases in acute respiratory disease in several states of central Mexico together with the unexpected elevation in severe respiratory disease rates among previously healthy young adults triggered an epidemiologic alert on April 17th, 2009 by the Mexico Ministry of Health [11].From this point on, the Mexican public health authorities maintained a communication channel not only with the Mexican people but also with the world through daily press releases led by the Secretary of Health on the current state of the pandemic and preliminary findings of ongoing research activities.Mexico's efforts were praised by disjoint groups such as the World Health Organization for the transparency and responsibility in responding to the 2009 H1N1 influenza pandemic [44].
On April 18th, the media spread the news about Mexico's epidemiologic contingency and special field work was carried out in search for cases of severe pneumonia in the country.On April 20th, preventive measures for health care workers were issued and laboratory tests initially revealed non-typeable infuenza A virus from samples of Veracruz and Oaxaca raising concerns for the circulation of a novel infectious disease.Specimens were immediately sent to the National Microbiology Laboratory in Canada and soon after Canadian researchers informed Mexican authorities about the detection of a novel influenza virus A/H1N1 of swine origin on April 23rd [20].These important international cooperation activities with Canada and the United States led the Mexico Ministry of Health to issue a pre-pandemic alert and activate a National Preparedness and Response plan [45].
On April 24th, Mexican President Felipe Calderon Hinojosa empowered the Federal Ministry of Health to coordinate all the national activities of prevention and control of the unfolding epidemic.Education activities were cancelled in the Federal District and the metropolitan area including the state of Mexico on April 24th.School closings were extended to the rest of the country three days later until May 11th when schools were re-opened.Other temporary measures to curb the spread of the epidemic in Mexico City included the closure of movie theaters and restaurants as well as the cancellation of mass-gathering events including soccer games and church services [45].
As of January 20th, 2010, a total of 69,607 confirmed novel H1N1 influenza cases and 944 deaths had been reported by the Mexico Ministry of Health [46].The temporal course of the pandemic closely resembled that of the 1918 influenza pandemic in Mexico [5].Roughly, three spatial waves of influenza morbidity and mortality have been documented during this pandemic in Mexico: the initial spring wave during the months of April and May mainly confined to the greater Mexico City area; a second summer pandemic wave during June and July localized in the southern states including the Yucatan Peninsula and a third fall wave that followed the return of students from the summer vacation period to classrooms, with higher morbidity and mortality than the earlier outbreaks.This third wave did not show a unique spatial pattern with outbreaks occurring in central states and northern states.There is a need for studies that carefully address the spatio-temporal characteristics of the 2009 influenza pandemic in Mexico in relation to demographic characteristics, school cycles, influenza seasonality and other geographic information relevant for Mexico.These analyses could suggest improvements to current influenza preparedness plans.

4.
Opportunities for modeling transmission dynamics of pandemic influenza.Understanding local risk factors and mechanisms that drive the spread of pandemic influenza is essential in the development of novel effective intervention strategies.Mathematical models provide a unique way to analyze influenza transmission patterns.It is worth mentioning that the history of the mathematical modeling of infectious diseases goes back to the work of Sir Ronald Ross in 1911 [47] who discovered the vector mechanism of transmission of the malaria parasite and explored the effects of controlling the mosquito population using simple mathematical models.Also relevant is the well-known work of Kermack and Mckendrick in 1927 [48] who introduced the classical SIR (susceptible-infectious-removed) epidemic model.
A crucial problem faced by decision-makers responsible for epidemic control is to monitor whether interventions are effective or, if not, to devise alternative measures.In particular, for infectious diseases characterized by rapid dissemination (such as pandemic influenza), it is important to identify its implicit profile of propagation.The profile of propagation is defined as the number and sites of cases characterized by random or non-random spatial distributions and, if non-random distribution is suggested by the data, whether cases are uniformly distributed over space or they constitute "clusters."Once the epidemic profile is determined, it is then relevant to identify whether all cases possess an equal influence on epidemic spread or, alternatively, whether some cases may contribute more to epidemic spread (and, therefore, should be prioritized in control measures).If differences in epidemic spread contributions are observed among cases, that would imply the importance of incorporating spatial information into transmission models rather than assuming that all cases are equal in their effects.
Since spatial features (e.g., human mobility patterns, roads, mountains, and weather patterns) are rarely uniformly distributed over space, determining the specific spatial scale(s) within which epidemic dynamic events develop, is crucial.Pandemic influenza spread, like the spread of forest fires, may be influenced by spatial attributes such as local demographic differences and topographic features.Other factors, like winds, human traffic, and seasonal attributes may also influence epidemic spread.Consequently, epidemics tend to be unique.That is, even when the biologic knowledge of the causative agent is known, the role of spatial factors (unique to a specific site and time) will greatly shape the transmission dynamics of infectious diseases.In this regard, a significant amount of prior work has addressed spatio-temporal propagation aspects of influenza (e.g., [49,50,51,52,53,32]).In this special issue, Herrera-Valdez and colleagues use a spatial transmission model to underscore the role of social distancing and school closures in shaping the temporal patterns of pandemic influenza in the context of the 2009 H1N1 influenza pandemic in Mexico.

Conclusion.
Overall, the 2009 influenza pandemic in Mexico shares a number of characteristic features with the 1918 influenza pandemic in the same country including 1) the timing of the pandemic onset in the Spring season [5,39], 2) a multiple wave profile of spread [5,21], 3) a relatively higher mortality rate among young adults compared to seasonal influenza epidemics [5,11], and 4)similar transmissibility levels in Mexico City [39,5].Yet, these two pandemics also differ in a number of aspects including the speed of propagation across countries with the 1918 influenza pandemic reaching most regions of the world in the order of months while the 2009 influenza pandemic reached most countries in the order of only a few weeks [54].Also, the mortality burden of the 1918 influenza pandemic was much higher than that of the 2009 influenza pandemic with some Mexican states experiencing mortality rates, circa 5000 per 100, 000 [6].By contrast, the overall mortality burden and case fatality rate of the 2009 influenza pandemic has not been found to be significantly higher than that of seasonal influenza epidemics [55,56,57].There are other clear differences in certain social and political components between the 2009 and the 1918 influenza pandemic.For example, the 1918 influenza pandemic took place in the middle of the Mexican revolution at relatively low sanitary conditions.In contrast, hygienic and sanitary conditions have significantly improved since then; antibiotics to treat bacterial complications of influenza are now available and influenza preparedness plans have been put together to consider the implementation of social distancing measures including temporary school closures and basic public health measures.
There is potential for vaccination to play a major role during future severe influenza pandemics.Current vaccine technologies are inadequate for the rapid deployment of vaccines that is required to control a fast propagating pandemic influenza virus.This and other issues related to vaccines and vaccination against influenza and other infectious pathogens are discussed by Curtiss in this same issue.Furthermore, several articles in this special issue use mathematical modeling to address aspects associated with vaccination strategies against pandemic influenza including the role of limited vaccine resources and delays in vaccine development.[21]) influenza pandemics in Mexico, respectively.For each pandemic wave, the grayscale range was set relative to the highest number of cases across states so that the state(s) with the highest number of cases/deaths are indicated in black.

Figure 1 .
Figure 1.Map of Mexico with state divisions.

Figure 2 .Figure 3 .
Figure 2. The age-specific weekly mortality of all respiratory deaths in the Federal District as recorded by the Civil Registry, Jan-2007 to Sep-2009.The clusters of deaths among young adults in 2009 can be readily identified.

Table 1 .
Events in chronological order relevant to the fall wave of the 1918 influenza pandemic in Mexico retrieved from national newspapers (October 09-November 27, 1918).

Table 2 .
[9]tality attributed to the Spring and fall waves of the 1918 influenza pandemic across Mexican states according to reports of the General Council of Hygiene[9].