Development and Validation of a Food Insecurity Risk Scale Using Item Response Theory

Introduction and Objective: In multiple trials we found that many patients did not screen eligible using the USDA tool because of stigma surrounding food insecurity. Focus groups demonstrated discomfort with the terminology of food insecurity and negative perceptions with being classified as being food insecure. To address this challenge, we developed a food insecurity risk scale using item response theory (IRT). Methods: We used literature review and focus groups to identify items most likely to correlate with food insecurity, but that will not trigger stigma. We identified 5 items and each was coded as yes or no. The 5 items were administered to 186 food insecure adults with type 2 diabetes enrolled in a food insecurity trial. Item response theory models were used to understand the association between a participants' underlying level of a particular item and the probability of that response. The two parameter logistic IRT model was used with each of the 5 items being added as a separate parameter in the model. Higher values in item discrimination indicate better ability of a specific item to differentiate participants from each other and to identify individuals most likely to have food insecurity. Results: In the two parameter logistic IRT model, difficulty score was -1.07 for item 2, -0.83 for item 3, 0.25 for item 5, 0.60 for item 1, and 0.84 for item 4. Discrimination was 0.88 for item 2, 0.98 for item 5, 1.14 for item 4, 1.39 for item 1, and 1.64 for item 3. We identified 3 ways to score the scale. First, item 3 and item 1 had the highest discrimination. Therefore, endorsement of these two items can be used to indicate food insecurity. Second, based on the model, a cut point of 3 or more of the 5 items can be used. Third, which is less stringent, is to use a cut point of 2 or more of any of the 5 items. However, endorsing the combination of items 3 and 1 is highly recommended Conclusion: We have developed a valid and reliable scale to measure the risk of food insecurity and identified two items that can quickly screen in individuals with high risk for food insecurity in clinical or research settings. Disclosure: R.J. Walker: None. J.A. Campbell: None. L.E. Egede: None. Funding: American Diabetes Association (1-19-JDF-075), National Institutes of Health (R01MD013826)