Discrepancies in Study of Vouchers for Healthy Foods and Diabetes—Reply

In Reply We are grateful for the comments from Shaik and Rajaduraivelpandian. In addition to the presented mean change in HbA1c and proportion of patients changing categories (eg, elevated to normal HbA1c), we can additionally report that the number of patients with an important (≥0.5%) reduction in HbA1c (48 [25%] with voucher vs 43 [22%]) did not differ substantially between groups (P = .55). The number of patients with any reduction in HbA1c was higher among those who received a voucher (97 [51%] with voucher vs 73 [38%]; P = .01).