The potential of parallel computing for locating primary care services

This paper investigates the potential for grid and parallel computing for location-allocation modelling of health services. Previous work showed the potential for solving small scale location-allocation problems concerning hospitals within a grid or parallel computing environment. In this study, we look at a much larger problem and investigate how computation within a grid could facilitate a more in-depth understanding of the problem. Within the context of siting primary care services in Ireland, an algorithm and case study for a region within Ireland are presented. From the case study it can be seen that finding a nationwide solution is computationally intensive. It is also shown how spatial interaction modelling is readily parallelisable and that location-allocation modelling can be translated to a grid environment. The benefits of such an approach and how it might lead to more equitable service provision are also outlined.