Preventive Services for Food Insecurity: Evidence Report and Systematic Review for the US Preventive Services Task...

An estimated 12.8% of US households experienced food insecurity in 2022.

To review the evidence on benefits and harms of screening and interventions for food insecurity in health care settings.

MEDLINE, CINAHL, and the Cochrane Central Registry of Controlled Trials through October 11, 2022; references of existing reviews; surveillance through January 24, 2025.

English-language randomized clinical trials (RCTs), nonrandomized studies of interventions, and pre-post studies conducted in US health care settings that examine the impact of screening or interventions for food insecurity; instrument accuracy studies of brief screening tools.

Single extraction, verified by a second reviewer. Results were narratively summarized.

Food insecurity; dietary, physiologic, quality of life, health outcomes; sensitivity and specificity of screening tools.

One RCT (n = 789) examined the impact of screening for food insecurity and found no difference in food insecurity after 6 months (29.6% in the intervention group vs 29.8% with usual care). Ten accuracy studies (n = 123 886) compared 1-, 2-, and 6-item subsets of the US Department of Agriculture Household Food Security Survey (HFSS) with the full HFSS. Sensitivity was typically above 95% and specificity above 82%, although most studies did not administer the screener separately from the reference standard, potentially overstating the accuracy. Twenty-nine studies (n = 74 292) examined interventions to address food insecurity, but 27 were rated as poor quality for the outcomes of interest for this review. Of the 2 fair-quality studies, 1 randomized crossover study (n = 44) found that home delivery of medically tailored meals was associated with reduced food insecurity (41.9% while "on meals" vs 61.5% while "off meals," P = .05). The other fair-quality propensity-matched cohort study found a smaller increase in body mass index among children whose families participated in a mobile food bank than those who did not after 6 months (mean difference in change, -0.68 [95% CI, -1.2 to -0.2]). The remaining studies found wide-ranging effect sizes for change in food security status but generally trended in the direction of benefit.

Brief screening tools likely have sufficient sensitivity to identify people with food insecurity in health care settings, but most studies of interventions to improve food insecurity had high risk of bias, limiting the ability to draw firm conclusions.