Did socio-ecological factors drive the spatiotemporal patterns of pandemic influenza A (H1N1)?: Fid-Bau Portal

Did socio-ecological factors drive the spatiotemporal patterns of pandemic influenza A (H1N1)?

Hu, Wenbiao / Williams, Gail / Phung, Hai / Birrell, Frances / Tong, Shilu / Mengersen, Kerrie / Huang, Xiaodong / Clements, Archie

Abstract

Abstract Background Pandemic influenza A (H1N1) has a significant public health impact. This study aimed to examine the effect of socio-ecological factors on the transmission of H1N1 in Brisbane, Australia. Methodology We obtained data from Queensland Health on numbers of laboratory-confirmed daily H1N1 in Brisbane by statistical local areas (SLA) in 2009. Data on weather and socio-economic index were obtained from the Australian Bureau of Meteorology and the Australian Bureau of Statistics, respectively. A Bayesian spatial conditional autoregressive (CAR) model was used to quantify the relationship between variation of H1N1 and independent factors and to determine its spatiotemporal patterns. Results Our results show that average increase in weekly H1N1 cases were 45.04% (95% credible interval (CrI): 42.63–47.43%) and 23.20% (95% CrI: 16.10–32.67%), for a 1°C decrease in average weekly maximum temperature at a lag of one week and a 10mm decrease in average weekly rainfall at a lag of one week, respectively. An interactive effect between temperature and rainfall on H1N1 incidence was found (changes: 0.71%; 95% CrI: 0.48–0.98%). The auto-regression term was significantly associated with H1N1 transmission (changes: 2.5%; 95% CrI: 1.39–3.72). No significant association between socio-economic indexes for areas (SEIFA) and H1N1 was observed at SLA level. Conclusions Our results demonstrate that average weekly temperature at lag of one week and rainfall at lag of one week were substantially associated with H1N1 incidence at a SLA level. The ecological factors seemed to have played an important role in H1N1 transmission cycles in Brisbane, Australia.

Highlights ► Temperature and rainfall at lag of one week were substantially associated with H1N1. ► Weather factors have played a more important role than social factors in H1N1. ► There was an interactive effect between temperature and rainfall on H1N1. ► Bayesian spatiotemporal methods can incorporate spatial correlation and uncertainty. ► Spatiotemporal model with covariates can be used to develop an EWS for H1N1.