Very Interesting 1st Step A Letter to the Editor For the Article “Searching for Sharp Drops in the Incidence of Pandemic A/H1N1 Influenza by Single Year of Age” published in
PLoS ONE Volume 7 Issue 8 in August 2012
This article “Searching for Sharp Drops in the Incidence of Pandemic A/H1N1 Influenza by Single Year of Age” looks at the historical relationship between past H1N1 encounters and the recent 2009 Pandemic. This article has 40 references. Most of these articles, being recently published, with the eldest being used solely for statistical reference for the “SiZer for exporation of structures in curves” (Chadhuri and Marron). A topic this student while interesting was completely confused by. I look forward to looking at this topic further. I would have liked to have seen more explanation of this in this article. For know SiZer meaning significant zero crossing of the derivatives is a graphical device to assess which observed features are `really there' and which are just spurious sampling artifacts. (Chadhuri and Marron). Most of the articles sited for this article are related to serological studies written after the pandemic, often written about specific populations (i.e. New Zealand, India, UK, specific cities). 40 articles are sited as referenced, but further research shows a lot more information was available. Several of the referenced articles are solely surveillance on Hospitalizations and the spread of the pandemic. I believe the two points of measure were not as well characterized as they may be for future studies. Several are related to how the disease can be reduced from lessons learned. One in particular used knowledge of the age-specific prevalence of immunity and incidence of infection with 2009 pandemic influenza A/H1N1 virus to discuss modeling the future burden of disease and the effectiveness of interventions such as vaccination. (Miller, Hoschler and Hardelid) The topic being very relevant to this article in that the age demographic for pH1N1 has shifted the at-risk population to the young, rather than the elderly. This article finds a significant decrease in risk for populations after age 57. (JH, BN and MG) This had been acknowledged by the WHO in 2009. (WHO) This article though shows quite well how this sharp drop can be interpreted to be true and provides a by-age risk ratio. There conclusion being that the reduced incidence of pH1N1 disease in older individuals shows a detailed age-specific pattern consistent with protection conferred by exposure to influenza A/H1N1 viruses circulating before 1957. (JH, BN and MG) Of particular interest is the methods used by the authors of this article to data obtained from confirmed cases of pH1N1 infection by single year of age and hospitalization status from Argentina, Australia (Queensland), Hong Kong, New Zealand, South Africa, Thailand, and the United States (Wisconsin and New York City) is impressive and shows that the ability to collect data globally and usefully is significantly been enhanced with global surveillance systems. . (JH, BN and MG) Though there methods eliminated blinding-related bias it also eliminated the ability to collect several more data points that could have been retrieved consistently across the population they sampled. The information collected was in theory nice to know, but population immunity is based on the life exposures of the cases. In this the limited number of countries and geographic regions is just nice to know. The numbers of each population were not as well characterized other than from this country or from this city. No other population information was used. There was no socioeconomic information, there was no standard of care information, the majority did not have sex information and one data set did not have hospitalization data. The hypothesis searched for a change in one population by age, but that age may be related to other factors as well. The researchers for this work used routine surveillance for pH1N1 conducted by the Ministries/Departments of Health in each location, and were reported to them anonymously as aggregated data covering many months (length of time varied by location). (JH, BN and MG) The authors also figured out a way to bring in to their study undiagnosed influenza cases after reviewing probability, but not confirmed data in an attempt to increase their case numbers. Reading what they say about it makes it sound logical, but with limited knowledge this student is unclear how it is the same as using confirmed cases. The investigators sub-analysis did allowed them to explore possible alternative mechanisms for any significant changes in incidence by age, including gender related exposure to pH1N1 or biological differences between the sexes in immunologic response to pH1N1. (JH, BN and MG) The researchers were also able to pull information from three populations that had information stratified by age and sex. For this the researchers used provided information from Argentina, Hong Kong and Wisconsin. This information was able to be evaluated separately to provide information in supplement on sex of cases. This data though being only from those three locals has little use other that for those interested in those locations. The data was compiled across the 8 populations and a weighted incidence risk ratio was derived for each age by year. All data in this study was compiled to focus on age of the case and there hospitalization status. This focus limited the final output of data. In understanding how they processed there data their approach to smoothing and statistical inference because of their use of SiZer, which removes the bias inherent in selecting a bandwidth and allows an inspection at a wide range of smoothing bandwidths to see which features are insensitive to bandwidth selection and likely to be true features. This method of presenting data was intriguing and visually helped show the plateau (ages 45-50) before the sharp drop of data by age from mid 50s to 70s. This student is intrigued by this and looks forward to learning this in the near term. This article was well written, interesting and the topic relevant to the field of epidemiology. This article is a first step in seeing the significance in exposure to H1N1 by age and hospitalization, further studies should be considered with more socioeconomic factors, standards of care and further review of each cases end state.
Works Cited
Chadhuri, P and J S Marron. "SiZer for Exporation of Structures in Curves." Journal of the Americal Statistical Association 94.447 (1999): 807-823.
JH, Jacobs, et al. "Searching for Sharp Drops in the Incidence of Pandemic A/H1N1 Influenza by Single Year of Age." PLoS one 7.8 (2012): e42328. <http://www.plosone.org/ar...>.
Miller, Elizabeth, et al. "Incidence of 2009 pandemic infl uenza A H1N1 infection in." Lancet 375.March 27, 2010 (2010): 100-08. <http://ac.els-cdn.com/S01...>.
Reed, C, et al. "Prevalence of Seropositivity to Pandemic Influenza A/H1N1 Virus in the United States following the 2009 Pandemic." PLoS One 7.10 (2012): e48187. Online. <http://www.ncbi.nlm.nih.g...>.
WHO. H1N1 in Post-Pandemic Period. 10 08 2010. <http://www.who.int/mediac...>.
—. Influenza-like illness in the United States and Mexico. 24 April 2009. World Health Organization. 23 11 2012. <http://www.who.int/csr/do...>.