As food insecurity interventions are incorporated into hospitals' population health initiatives, addressing the needs of hospitals' patients and communities through the same interventions may be ineffective if the groups vary and have different needs. This study examined whether food insecurity predictors were different in the general population compared to individuals with hospital discharges, and also whether food-insecure hospital patients differed from food-insecure community members. National data were extracted from the 2016 Medical Expenditures Panel Survey. Summary statistics were compared to test for differences between food security status groups. Logistic regressions were estimated for the general population and for individuals with hospital discharges to identify associations between food insecurity and demographic, socioeconomic, and health characteristics. Food-insecure individuals with and without hospital discharges differed statistically across multiple variables, including 15 of 16 health-related variables. However, compared to food-secure individuals with hospital discharges, food-insecure individuals with hospital discharges differed on only half of the health variables. Food insecurity predictors also differed among the general population and hospital discharge samples; for instance, age and race were only associated with higher likelihoods of food insecurity in the population sample. Furthermore, 9 health-related variables were associated with food insecurity in the population sample relative to only 2 in the hospital discharge sample. Food insecurity predictors differed between the general population and individuals with hospital discharges; food-insecure individuals with and without hospital discharges also differed statistically. Therefore, hospitals should carefully consider their target populations when constructing population health initiatives.
Keywords: clinical screening; community investment; food insecurity; population health; social determinants of health; value-based payment.