OBJECTIVES: Although clinical trial data are often used in economic models, they may not reflect real-world utilization and outcomes of therapy. To determine real-world resource use and outcomes of diabetic foot ulcer (DFU) care using Graftskin (a living bi-layered skin substitute) and to supplement clinical data for a DFU treatment model, a medical record abstraction study was conducted at 21 U.S. treatment centers that practice good wound care (GWC). METHODS: Two physician samples were recruited: a Graftskin sample (physicians who frequently use and receive reimbursement for Graftskin in the treatment of DFUs), and a non-Graftskin sample (physicians who treat DFU using GWC and who do not receive reimbursement for using Graftskin). Separate abstraction forms were developed for each sample, sent by mail to the treatment centers, completed by center staff, and returned via fax. The center received a small honorarium per completed form. RESULTS: Twenty-one centers and 166 cases (83 Graftskin and 83 GWC) were included in the analysis. Graftskin patients had more severe DFUs (21 cm2) than those in the Graftskin pivotal trial (2.97 cm2), which compared GWC plus Graftskin to GWC alone. In addition, Graftskin patients had more severe DFUs than the GWC patients in this study (11 cm2). The mean number of Graftskin applications was much lower in this study (1.27 applications) than in the Graftskin pivotal trial (3.9 applications). Moreover, in this study, the incidence of severe adverse events was significantly lower with Graftskin (37%) than with GWC (52%), even when controlling for ulcer severity (p=0.006). CONCLUSION: In actual practice, the number of Graftskin applications was considerably lower and DFU severity was substantially higher than in the Graftskin pivotal clinical trial. These data were applied to the DFU treatment model, which yielded improved estimates of the real-world impact of utilizing Graftskin versus GWC.