Comparing Care Coordination and Food Assistance Referrals to Improve Diabetes Outcomes: A Cohort Study

Top Things to Know

The results support food-based interventions as a viable clinical strategy for addressing food insecurity and cardiometabolic health.

The study indicates that healthcare organizations can choose food-based interventions based on local infrastructure, community partnerships, reimbursement opportunities, and patient preferences.

The findings suggest that health systems do not necessarily need a single "best" model; both direct food assistance and broader social-needs care coordination can be incorporated to support patients with diabetes.

Summary of Conclusion/Findings

This retrospective cohort study compared diabetes outcomes among 2,108 adults with type 2 diabetes who received either a food assistance referral (n=1,148) or a care coordination referral (n=960) in community health centers across the United States. Using electronic health record data and advanced causal inference methods, the investigators evaluated changes in HbA1c, blood pressure, and LDL cholesterol over 6 to 24 months following referral. The primary finding was that there were no clinically meaningful differences between the two referral approaches for any measured outcome. At 6 months, the difference in HbA1c between groups was only 0.01 percentage points (95% CI −0.15 to 0.18), with similarly negligible differences in systolic blood pressure, diastolic blood pressure, and LDL cholesterol. Results remained consistent across longer follow-up periods, among patients with documented food insecurity, and before versus after the COVID-19 pandemic. The authors concluded that food assistance referrals and care coordination referrals appear to produce comparable diabetes-related outcomes, suggesting that health systems can select either approach based on feasibility, available resources, and patient preferences rather than expected differences in clinical effectiveness.