Table 1.
Summary of epidemiological data.
Table 2.
List of model parameters and their values.
Figure 1.
Surveillance data and model estimates for weekly incidence of cases.
For each country, graphs show observed case incidence from surveillance data (black points), the 95% credibility region on incidence from the country-specific fits (grey region) and predicted incidence for the posterior median set of parameters obtained from the global fits (dashed lines) for model variants M1 (blue), M2 (green) and M3 (red). Weekly incidence from the models is plotted in all cases, with lines being drawn between weeks for visual clarity. Depending on the country, observed case incidence are either confirmed H1N1pdm cases (H1N1CC) or influenza like illness rate (ILI) - showing ILI rate per 100,000 population for Chile and New Zealand and ILI rate per 10,000 consultations for Australia and Victoria.
Figure 2.
Surveillance data and model estimates for the age-distribution of cases.
Observed cumulative cases distribution among age-groups (grey rectangles) and model median posterior estimates (coloured thin bars). The dark grey bars correspond to country-specific fits, whereas blue, green and red bars represent the results for M1, M2 and M3 model variants of the global model, respectively.
Figure 3.
(A) Estimated empirical R0-values derived from the early exponential growth rate of the epidemic versus proportion of children in the eight studied countries/states. R0-values estimated from data on H1N1 confirmed cases were used in the regression analysis except for Victoria for which only ILI data was available. (B) Distribution of estimated reproduction numbers by country obtained in country-specific and global fits. For each country, the posterior median estimates of R0 for country-specific and global fits are plotted with 95% credible intervals. The grey circles correspond to country-specific estimates, whereas blue squares, green stars and red triangles represent estimates for M1, M2 and M3 model variants of the global fits, respectively. For those countries where two datasets were available, the two estimates are plotted. For the global fits, because R0 differences among countries derived from population demography only, fitting resulted in one estimate only even when both ILI and confirmed case data were available.
Figure 4.
Estimated age-dependent susceptibilities.
Estimated susceptibilities (posterior median with 95% credible intervals) are plotted according to age in the 8 countries/states for (A) country-specific fits and (B) global fits (M1, M2 and M3).
Table 3.
Estimated parameters for country-specific model (median posterior with 95% credible interval indicated in parenthesis).
Table 4.
Estimated parameters for global model variants.