Accessibility to primary care physicians: Comparing floating catchments with a utility-based approach

Floating Catchment Area (FCA) methods are a popular choice for modelling accessibility to healthcare services because of their ability to consider both supply and demand. However, FCA methods do not fully consider aspects of travel and choicemaking behaviour as the only behavioural component is the impedance function. FCA approaches also tend to assign population demand to clinics and levels-of-service to population zones in an overlapping manner that has been shown to inflate/deflate supply and demand. While the adjustments proposed in the recent Balanced FCA method can rectify this, it apportions population and levels of service in a fractional manner. In response, this research proposes a utility-based measure of healthcare accessibility based on a multinomial logit (MNL) destination choice model that avoids the multiple-counting issue in FCA methods. It also considers additional behavioural aspects that define the appeal of clinics in addition to the travel time required to reach them, including their capacity and level of crowding. Comparisons of the MNL approach with the original and balanced FCA models using data for the City of Hamilton, Canada, suggest that while the accessibility patterns produced by each method are broadly similar, some key differences exist in the calculated accessibilities and their spatial patterns. The MNL model in particular estimates higher accessibilities in suburban and rural areas. After considering their strengths and weaknesses, we argue that both the FCA and MNL approaches offer merit for planning and policy.

Automated summary

This research proposes a utility-based measure of healthcare accessibility using a multinomial logit (MNL) model to avoid the multiple-counting issue in FCA methods. The MNL model estimates higher accessibilities in suburban and rural areas compared to the FCA models.