Pandemic influenza in Papua New Guinea: a modelling study comparison with pandemic spread in a developed country

Milne, George J; Baskaran, Pravin; Halder, Nilimesh; Karl, Stephan; Kelso, Joel

To achieve simulations for a particular basic reproduction number R0, βv was adjusted; details of the procedure for estimating βv and R0 are given in [20] . A pandemic with a basic reproduction number of 1.5 is used in this study, and corresponds to some estimations of the 2009 H1N1 pandemic [36] [37] [38] [39] and the 1957 and 1968 pandemics [40, 41] . Since it is virus intrinsic, we assumed that βv was similar in Madang and thus calculated R0 in the Madang and Madang-nnh models using the βv corresponding to an R0 of 1.5 in the Albany model. Details of parameter settings used in each model are given in Table A2 in the Supplementary Data File. Individuals in each household and contact hub make contacts within a close-contact mixing group, taken to be the entire household or a subset of larger hubs, and also make additional non hub-based random community contacts. The individual-toindividual transmission probability in larger households was moderated by household size using data from a study by Cauchemez et al [42] . The reduction of the probability of within-household transmission with household size is also shown in Figure A1 in the Supplementary Data File. This reflects the assumed reduction in pairwise contact in households with greater than 3 members and is significant in the Madang context, where the average household size is 6.32 compared to 2.54 in Albany.