Seroepidemiology aims to understand population-level exposure and immunity to infectious disease. Serological results are normally presented as binary outcomes describing the presence or absence of antibody, despite the fact that many assays measure continuous quantities. A population's antibody titers may include information on multiple serological states - naiveteĢ, recent infection, non-recent infection, childhood infection - not just seropositivity or seronegativity. In the first part of this thesis, I investigate 20,152 general-population serum samples from southern Vietnam collected between 2009 and 2013 from which I report antibody titers to the influenza virus HA1 protein using a continuous titer measurement from a protein microarray assay. I describe titer distributions to subtypes 2009 H1N1 and H3N2, and using a model selection approach for mixture distributions, I determine that 2009 H1N1 is best described by four titer subgroups while H3N2 is best described by three titer subgroups. For H1N1, my interpretation is that the two highest-titer subgroups correspond to recent and historical infection, which is consistent with pandemic attack rates. For H3N2 however, right-censoring of titers makes interpretations difficult to validate. To move beyond this stationary interpretation of titers, I developed two methods for analyzing this serum collection as a time series. First, I attempted to analyze mixture categories in individual time windows. This approach did not lead to a consistent temporal picture of titer change; results differed by site and were sensitive to assumptions in the mixture fitting. Second, I attempted to fit a hybrid-dynamical model with free incidence parameters. This inference was robust to parameter assumptions, consistent across sites, and in agreement with the incidence reported in Vietnam's influenza surveillance network. In addition, my new approach showed evidence that there was a second silent wave of the 2009 influenza pandemic that was not recorded in national surveillance, which is this thesis' main novel result.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:736149 |
| Date | January 2017 |
| Creators | Nhat, Nguyen Thi Duy |
| Contributors | Boni, Maciej Filip ; Simmons, Cameron ; Baker, Stephen |
| Publisher | University of Oxford |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | https://ora.ox.ac.uk/objects/uuid:f8eba4e1-9c68-4750-bfed-e6f0de8dd2de |
Page generated in 0.002 seconds