by Luke Shulman

Food Access is loosely defined as “people’s ability to find and afford food”. However, there are a number of related concepts and terms depending on the mode of study being employed. A recent Tufts University study, categorized ten variations on the term including the colorful geographic nomenclature “food desert”, “food swamp”, and “food hinterland”. Food access is a difficult issue to categorize because of tremendous variation in prices, cultural preferences, and skill sets. Access is generally studied using spatial distance of a resident to a source of food or a density calculation of food within a certain geographic boundary such as a precinct or city. There are even greater differences when comparing urban and rural regions where the expectations of distance are wildly different.

Why is this important?

Research continues to support that in certain populations access to a grocery store and healthful food “is associated with lower prevalence of overweight and obesity, improved fruit and vegetable consumption and better diet quality”. (Literature Review) Food access may also play a role in the four major chronic conditions where diet is a major factor. (source). Thus, “one hypothesis is that economically and socially disadvantaged neighborhoods have inadequate access to healthy food, thus negatively affecting dietary quality and health.”(source)

Where do we get food?

An essential input to these models is a categorization of food outlets that can be difficult to achieve at scale. In the United States, the regulation of food outlets often takes place at a local municipal level which can comprehensively catalogue and license restaurants, convenience stores, and supermarkets. While one city may have comprehensive data source for food outlets, another city may not.

Understanding the Output

Algorex Health uses a methodology similar to the Tufts paper cited above.

Where is Food Available:

To find food outlets, Algorex Health utilizes a proprietary database of businesses classified by the types of food that they offer. At a regional level, specific outlets must be removed or re-categorized based on consultation with Algorex Health customers or known characteristics of the businesses.

The outlets are then weighted with supermarkets, grocery stores, warehouse stores, and markets receiving the highest rating (ratings are based on business size).

Geospatial Processing

For each food outlet, a geospatial polygon is calculated in four zones. Two zones are for walkability (14 mile distance and 12 mile distance) and the other for drivability (5 mile or 20 mile) based on spatial distance.

Individual patients are given an access score based on the highest retail outlet weight they have access to. Patients outside of these polygons are given a score of zero.

Algorex Health customers can use this as a weighting factor to direct interventions such as: providing access to fresh food sources, enrolling members in mail-order nutrition kits, or organizing rides to food stores. All these things help combat food-related disease burden and costs.

About this series

You can find all of the blog posts in our SDOH Deep Dive series at SDOH Deep Dive Series