Introduction: Food insecurity (FI) is associated with poor health outcomes, including obesity, diabetes, hypertension, and mental illness. While food voucher programs help address FI, their clinical impact is understudied. At Prisma Health, where 1 in 5 patients face barriers to healthy eating, we implemented a voucher program at a Family Medicine clinic to improve food access and assess health outcomes.

Materials and Methods: Between February 1 and November 1, 2024, nurses and program coordinators recruited patients during routine visits if they met ≥1 of the following criteria: (1) positive screen on a validated FI questionnaire; (2) homelessness; (3) on Medicaid; (4) uncontrolled DM (A1c ≥8.0%), or (5) HTN (≥130/90 mmHg). Participants received one to three $15 vouchers per visit, redeemable for produce at a nearby market. Voucher use was tracked at the individual level, and clinical data were obtained through manual chart review. Analyses included descriptive statistics via SAS, negative binomial regression for redemption rates, and paired t-tests or Wilcoxon tests for clinical outcomes.

Results: We distributed 4,552 vouchers to 943 participants, with a 54.6% redemption rate. After exclusions, 877 were analyzed: 78% female, 85% Black, and 78% extremely low income. Diagnoses included DM (22%), HTN (52%), depression (19%), anxiety (23%), and overweight/obesity (61%; among 710 patients with BMI data). In multivariate analysis, HTN (IRR 1.26, p = 0.0018), age ≥45 (IRR 1.42, p = 0.0002), Black vs. White race (IRR 1.37, p = 0.0009), and Latino vs. Black race (IRR 1.39, p = 0.0339) were independently associated with higher redemption. In a pre-post analysis of 535 overweight/obese patients, BMI significantly decreased (mean difference -0.44, p = 0.0006). Among 136 patients with HTN who had a pre- systolic blood pressure ≥ 140, SBP decreased on average by 11.2 mmHg. In patients with uncontrolled DM, (pre- A1c ≥ 8, n=68) there was a 1.17 decrease in A1c. Patients with a PHQ-9 ≥ 10 experienced an 8.8-point decrease (p<0.0001) in their score. No significant changes were seen in GAD-7.

Conclusion: This study identified patients likely to engage in a food access program, including those with HTN, age ≥45, and Black or Latino race. BMI improved among overweight/obese participants; other clinical outcomes remained stable. As the first study of its kind in SC – one of seven states with FI prevalence above national average – future research is indicated to evaluate whether longer-term or higher-intensity programs improve clinical outcomes.