This study aimed to characterize the spatial distribution of hypertension (HT) clusters in a rural Australian city using self-reported HT data collected at a local health-screening clinic. HT status was recorded for 515 self-selected participants in a free health-screening program in Albury, New South Wales, Australia. We compared predictions of HT clusters computed using spatial scan statistic and Generalised Additive Model (GAM). We then implemented a new approach incorporating sensitivity analysis in GAM to combine cluster predictions at multiple span sizes. A statistically significant cluster for HT was identified in Albury centered to the north of the main urban center, with relative risk up to 2.29. The sensitivity analysis confirmed the cluster location and highlighted other potential HT clusters. Our approach allows detection of irregularly-shaped disease clusters and highlights potential clusters that may be overlooked using traditional methods. This is important in cases using local, small datasets where regularly-shaped or overly smoothed disease clusters may not provide enough detail to be suitable for targeting place-based interventions.
|Number of pages||4|
|Journal||Computing in Cardiology|
|Publication status||Published - 1 Jan 2017|
|Event||44th Computing in Cardiology Conference, CinC 2017 - Rennes, France|
Duration: 24 Sep 2017 → 27 Sep 2017