Using big data to search for possible geographic clustering of Congenital Heart Disease (CHD) across Australia.

Abstract

Several diseases show geographic clustering, giving insights into possible genetic and environmental causes. The pathogenesis of Congenital Heart Disease (CHD) remains largely unknown and analysis of geographic distribution of CHD cases lacks input from large, national-scale datasets. People with structural CHD were selected from the Australia and New Zealand CHD Registry. Of people known to be still living, from linkage with the National Death Index, addresses were geocoded and aggregated to standardised geographic regions with measures of the Australian population. Areas were described based on measures of their remoteness and driving time to hospitals. The relationship between the distribution of the CHD and Australian populations was compared with bivariate spatial correlation. Of 81,349 people with structural CHD in the Registry, 63,863 were still living and could be geocoded. Overall, most people lived in Major Cities, and within 1-hour drive from a hospital, with the proportion the same across the CHD population, the "complex CHD" population and the Australian population. Across the country, there was a strong positive correlation between the Australian population and the CHD population. There were only a small number of areas (6%) where the Australian and the CHD populations were proportionally different. Overall, there was clear evidence that the geographic distribution of the CHD population proportionally follows the general Australian population. This suggests that there is unlikely to be any spatial clusters that are driven by genetic or environmental causes.