1 Introduction Primary health care is the first line of defence for the population and can therefore prevent or reduce unnecessary, expensive speciality care [1–3]. To prevent unequal access to primary health care, health service planners and policy makers need accurate measures of accessibility, so shortage areas can be identified, to award financial assistance to physicians who settle in such areas [3]. An example is Impulseo I, an incentive program in Belgium. Physicians receive €20,000 when they settle in a physician zone with a low physician-to- population ratio (PPR < 90 physicians/100,000 inhabitants, or PPR < 120 physicians/100,000 inhabitants and population density < 125 inhabitants/km 2 ) [4]. A physician zone is a contiguous geographic area of one or more municipalities. As can be deduced from the parameter in Impulseo I, and also from other incentive programs [5], policy makers currently use fairly simple methods to define health care accessibility and shortage areas. Geographical information systems (GIS) offer a wide range of more advanced methods to calculate physician density and shortage areas. Therefore, GIS can help policy makers to examine access needs, to better identify shortage areas, and to monitor the impacts of intervention policies [6], [7]. GIS are software tools for researchers or policy makers to input, store, manipulate, analyse, and visualise spatial information [8]. 2 Background 2.1 Potential spatial accessibility Health care accessibility can be classified into two categories: revealed and potential accessibility [9–11]. The former focuses on the actual use of health care services, while the latter focuses on the aggregated supply of available health care in an area and thus the potential use of services. Both can be further divided into spatial and non-spatial accessibility. Spatial accessibility is based on spatial factors, including the distribution of primary health care providers (supply) and population (demand), and the distance/time between supply and demand [12]. Non-spatial accessibility is based on non- spatial factors such as socio-economic factors, the health status of the population, and people’s knowledge about the health care system [9], [12]. In this paper, we will focus on potential spatial accessibility (henceforth briefly referred to as accessibility), because it is essential toward any effective government intervention program to identify where potential shortage areas are located [2], [13]. 2.2 Measures of potential spatial accessibility To calculate primary health care accessibility in general and physician shortage areas in particular, various methods can be used. Simple methods include distance/time to the nearest physician, and the number of physicians within a certain distance/time [14]. However, these methods often give only a rough estimation of accessibility. Distance to the nearest provider for example does not capture full accessibility, because it is often observed that people bypass the nearest service when there is more than one service to choose from [15], [16]. Physicians co-exist in a network of overlapping catchments, and people are free to choose health care wherever and from whomever. Therefore, physicians compete for the population’s use of their services [16]. Because there is no single pathway between the population and physicians, some methods are based on PPRs to measure accessibility in a certain area, as is the case in Impulseo I. The advantage of this method is that it is easy to implement and comprehend. However, these traditional PPRs have several limitations [17– 19]. First, PPRs are usually calculated with zonal data, which are based on administrative boundaries (e.g. municipalities). Financial assistance in areas with low access to primary health care: a review of policy methods in Belgium Bart Dewulf Ghent University Krijgslaan 281, S8 9000 Ghent, Belgium bartd.dewulf@ugent.be Tijs Neutens Ghent University Krijgslaan 281, S8 9000 Ghent, Belgium tijs.neutens@ugent.be Yves De Weerdt VITO Boeretang 200 2800 Mol, Belgium yves.deweerdt@vito.be Nico Van de Weghe Ghent University Krijgslaan 281, S8 9000 Ghent, Belgium nico.vandeweghe@ugent.be Abstract In some countries, financial assistance is awarded to physicians who settle in an area that is designated as a shortage area, to prevent unequal access to primary health care. Policy makers use fairly simple methods to define health care accessibility, for example physician-to- population ratios (PPRs) within predefined administrative boundaries. Our purpose is to check whether these methods give accurate estimations of health care accessibility. More in particular we test the potential of various floating catchment area (FCA) methods. The basic PPR method offers only a crude representation of health care accessibility, because large contiguous areas are considered. Local variations can therefore often not be detected. The enhanced two-step floating catchment area (E2SFCA) method is able to calculate accessibility at a small scale (e.g. census tracts), takes interaction between physicians into account, and considers distance decay. The resulting accessibility is much more geographically spread and offers a better representation of reality. Therefore this method is preferable used by policy makers to define shortage areas for awarding financial assistance. Keywords: Primary health care, accessibility, geographical information systems (GIS), floating catchment area (FCA)