Download 39188-2-12118

International Biometric Society
USING DIFFERENT MODEL SELECTION TOOLS TO ELUCIDATE THE TRANSMISSION POTENTIAL OF
VZV IN EUROPE FROM A SOCIAL CONTACT PERSPECTIVE
Santermans, E.1, Goeyvaerts, N.1,2, Melegaro, A.4, Gay, N.4,5, Edmunds, J.6, Aerts, M.1,
Beutels, P.2,3, and Hens, N.1,2
1
Interuniversity Institute for Biostatistics and statistical Bioinformatics, Hasselt University,
Diepenbeek, Belgium
2
Centre for Health Economics Research and Modelling Infectious Diseases, Vaccine &
Infectious Disease Institute, University of Antwerp, Antwerp, Belgium
3
School of Public Health and Community Medicine, The University of New South Wales,
Sydney, Australia
4
DONDENA Centre for Research on Social Dynamics, Bocconi University, Milan, Italy
5
Modelling and Economics Unit, Public Health England, London, United Kingdom
6
London School of Hygiene & Tropical Medicine, London, United Kingdom
The basic reproduction number 𝑅0 and the effective reproduction number 𝑅 are pivotal
parameters in infectious disease epidemiology, quantifying the transmission potential of an
infection in a population. We estimate both parameters from 13 pre-vaccination serological
data sets on varicella zoster virus (VZV) in 12 European countries under the assumption of
endemic equilibrium. This means that varicella may undergo cyclical epidemics, however
fluctuating around a stationary average over time. Under this assumption, the expected
value of the effective reproduction number is 1. Estimating transmission rates for an
airborne infection such as VZV requires assumptions on the underlying age-specific mixing
patterns. The basic reproduction number 𝑅0 has been shown to be highly sensitive to these
mixing assumptions. Serological surveys, however, do not provide complete information
about these mixing patterns, since they reflect the rate at which susceptible individuals
become infected, but not who is infecting them. We address this unidentifiability by informing
the mixing pattern with data from population-based social contact surveys, assuming
transmission rates are proportional to contact rates [1,2]. Rates of close contact lasting at
least 15 minutes are estimated using a bivariate smoothing approach, while non-parametric
bootstrap is used to assess variability. Further, we evaluate how constant and age-specific
proportionality assumptions affect the estimated 𝑅0 values and the fit to the serology using
the inferred effective reproduction number as a model eligibility criterion combined with AIC
as a model selection criterion.
In all countries, primary infection with VZV most likely occurs in early childhood, but there is
substantial variation in transmission potential with 𝑅0 ranging from 2.8 in England and Wales
to 7.6 in the Netherlands. Two non-parametric methods, the maximal information coefficient
(MIC) [3] and a random forest approach, are used to explore these differences in 𝑅0 by
means of country-specific characteristics as for example childcare attendance, population
density and average absolute humidity. The results indicate that higher day care participation
rates and higher pre-school attendance rates relate to larger varicella transmission potential.
This illustrates the need to consider epidemiological differences between European countries
when parameterizing mathematical models to inform VZV vaccination. They also reveal for
the first time based on serological data, which factors are important in explaining
epidemiological differences between European countries.
[1] B. Ogunjimi, N. Hens, N. Goeyvaerts, M. Aerts, P. Van Damme, and P. Beutels. Using empirical social contact
data to model person to person infectious disease transmission: an illustration for varicella. Mathematical
Biosciences, 218:80-87, 2009.
[2] J. Wallinga, P. Teunis, and M. Kretzschmar. Using data on social contacts to estimate age-specific
transmission parameters for respiratory-spread infectious agents. American Journal of Epidemiology, 164:936944, 2006.
[3] D. N. Reshef, Y. A. Reshef, H. K. Finucane, S. R. Grossman, G. McVean, P. J. Turnbaugh, E. S. Lander, M.
Mitzenmacher, and P. C. Sabeti. Detecting novel associations in large data sets. Science, 334:1518-1524,
2011.
International Biometric Conference, Florence, ITALY, 6 – 11 July 2014