Developing an accurate and cost-effective food environment validation method

Background: A major concern in food environment research is the lack of accuracy in secondary commercial business listings of food stores, which are convenient and commonly used.  Such accuracy concerns may be particularly pronounced in rural food environments. Ground-truthing (on-site verification) has been deemed by a number of researchers to be the necessary standard to validate business listings, but it is also costly and time-consuming. This study simulated a modified ground-truthing validation method aimed at increasing efficiency without comprising accuracy.

Methods: Traditional ground-truthing was conducted in three areas of rural Minnesota (564 miles total). Geographic coordinates of food stores and the route driven were recorded by a GPS device paired with a tablet. The ground-truthed list of stores was compared with secondary data business listings, and the positive predictive value (PPV) and sensitivity were calculated. New stores identified were mapped to see where they fell in relation to the central commercial clusters (the 1/8 mile buffer around a “cluster” of ≥ 2 stores) in the study area.

Results: The PPV ranged from 0.45 to 0.63 across the three sites. Sensitivity ranged from 0.60 to 0.71. Of the 51 new stores identified by ground-truthing, 96% were located within central commercial clusters. Driving only the street networks in these clusters would have resulted in a 90% reduction in the total number of roads to validate (54.5 miles).

Conclusions: On-site validation is indeed necessary for accurate food environment exposure assessment, but more efficient ground-truthing methods could be used and warrant further evaluation.