Evaluating screening to assess endorsement of food insecurity in the inpatient setting

OBJECTIVE: Rates of food insecurity (FI) from screening in the inpatient setting is often not reflective of community prevalence, indicating that screening likely misses families with FI. We aimed to determine the combination of FI screening questions and methods that would result in identifying a percentage of FI families that matched or exceeded our area prevalence (approximately 20%). 

METHODS: Research staff approached eligible English- and Spanish-speaking families across 4 inpatient units once weekly and screened for FI using a randomly selected method (face-to-face, phone, paper, and tablet). We asked questions from the 6-Item USDA Survey, Hunger Vital Sign screener, and questions utilized by our social workers. 

RESULTS: We screened 361 families; 19.4% (N 5 70) endorsed FI. Differences in rates were not significant by method. Differences in FI rates based on screening questions were: 17.7% for the 6-item USDA survey, 16.0% for Hunger Vital Sign, and 3.1% for the social work questions. When considering method and screening questions together, the 6-Item USDA on paper had the highest positivity rate of 20.9%. A higher percentage of Spanish-speaking families endorsed FI (61.1%) compared to 17.2% of English-speaking families (P < .01). Positivity also varied significantly by self-identified race (P < .01). Caregivers that identified as Hispanic or Latino were significantly more likely to endorse FI than those that did not (P < .01). 

CONCLUSIONS: The positivity rate for FI while screening inpatient families using the 6-Item screening questions on paper matched our community prevalence of FI (approximately 20%).