Human influenza A (H1N1) 2009 has caused severe epidemics in many countries, but its dynamics in spatial and temporal contexts have so far been poorly appreciated.

A total of 24,414 laboratory confirmed human influenza A (H1N1) 2009 cases reported from May to September 2009 in Hong Kong were evaluated, using a Geographic Information System (GIS)-based approach. Spatio-temporal clustering was assessed after dividing Hong Kong into 4 geographic sectors, 18 districts and 500 m × 500 m cells. Global Moran’s I, Local Moran’s I and SaTScan™ were used in the exploration.

Spatio-temporal clusters first appeared on Hong Kong Island at week 3, alongside multiple foci suggestive of infection nidus introduced from abroad. The clusters grew rapidly and became confluent in urban areas, lasting till week 22. Separately, local clusters emerged in the North representing second or third generation infections, which died down over a relatively short period of time.

Heterogeneity of spatio-temporal clustering of H1N1 was demonstrated during the epidemic, despite the small area of the territory of Hong Kong. The dynamics could have been shaped by population mobility at a local level. The application of GIS in epidemiology studies can add value to standard surveillance activities.