Geomatics, epidemiology and biostatistics: An application to acute coronary syndrome

Abstract

The emerging field of Geomatics has found useful application in several research areas. A new phase of its development merges it into the realm of epidemiology and public health to bring insight into the regional disparities of disease incidence and for disease surveillance. New advances in biostatistics include spatial statistics methods which aim to specifically understand and model the spatial variability. Spatial statistics, when combined with geomatics, constitutes an excellent and powerful analysis approach to handle and better understand health issues, specifically in disease related prevention and intervention studies. The main goal of this paper is to explore the integration of geomatics and spatial statistics with an application to a specific health issue. The outcome of interest is acute coronary syndrome (ACS) incidence in the province of Quebec (Canada) between 1996 and 1998, and hospital readmission at one month post-discharge. It is an established fact that mortality and hospital-acquired infection rates are indicators of the quality of care [24] but more recently some studies are turning their attention to the early hospital readmission rate [2, 26]. Within this context, following specific questions are addressed: Is there spatial heterogeneity and/or spatial aggregation in the ACS incidence and early readmission rates? Is there any geographical trend in the rates? Is there an explanation for the spatial heterogeneity? By ACS, we mean the occurrence of myocardial infarction (MI) or unstable angina. By early readmission, we mean readmission of a patient for coronary heart disease (ACS and angina) 30-days post discharge. Although practice guidelines have been in circulation to standardize the treatment and follow-up of acute myocardial infarction [16, 22], regional variations are currently reported in the literature [20, 21, 25]. The presence of a complex network of factors influencing care quality [2], hospital readmission and the interaction between them have been put forward as the potential explanation of the observed spatial variability in hospital readmission rates. Most of these factors center on the patient while the data available and/or the interests of public policy makers focus on rates of local health units over a given time period. Over these administrative geographical units, interest centers on variables that could explain spatial variability, such as deprivation indices and other area specific characteristics, and help to understand inequalities within health care services and accessibility. © 2009 Springer Berlin Heidelberg.