Using the Society for Vascular Surgery Wound, Ischemia, and foot Infection classification to identify patients most likely to benefit from revascularization

Background: The Society of Vascular Surgery Wound Ischemia foot Infection (WIfI) classification system for chronic limb-threatening ischemia was intended to predict 1-year major lower extremity amputation (LEA) risk and to identify which patients benefit most from revascularization. We aimed to identify which WIfI presentations benefited most from revascularization to explore whether a cluster analysis could identify a more data-driven WIfI score, and to quantify which component of the WIfI score was most strongly associated with 1-year LEA after revascularization. Methods: Composite multi-institutional nested cohort data from centers who previously validated WIfI were reviewed retrospectively. We collected each patient's WIfI component grades and whether LEA was performed. To examine the benefit of revascularization, the predicted LEA rates were subtracted from observed LEA rates. We used k-means cluster analysis to model predicted vs observed LEA rates after revascularization. Multivariable linear regression analysis was performed to quantify which WIfI score component(s) best predicted LEA. Results: Data from 10 centers, accumulated between 2005 and 2015 were collated (2878 limbs at risk; 314 LEAs performed). The subset of patients who underwent revascularization comprised the study base (1654 limbs; 169 LEAs). Of 64 potential WIfI grade combinations, 15 were never reported and were excluded from the analysis. By original WIfI stages, the observed LEA rate after revascularization was: stage 1, 10.8% (14/130); stage 2, 4.9% (5/103); stage 3, 5.1% (25/487); and stage 4, 13.4% (125/934). Based on the difference between predicted and observed LEA risk for those who underwent revascularization, the WIfI scores were placed into quartiles from greatest to no benefit of revascularization. Cluster analysis identified four clusters with the following 1-year LEA rates: cluster 1, 4.4% (46/1038); cluster 2, 14.8% (66/447); cluster 3, 28.1% (36/128); and cluster 4, 51.2% (21/41). The between sum of squares/total sum of squares was 93.9%. Multiple linear regression revealed the wound grade most strongly predicted LEA (F-value, 17.25; P <.001). Ischemia (F-value, 6.51; P =.001) and infection (F-value, 5.7; P =.003) were similarly associated with LEA risk. Interaction terms between each component of the WIfI score were not statistically significant. Conclusions: WIfI can identify which patients with chronic limb-threatening ischemia are most likely to benefit from revascularization and may provide improved prognostication, risk stratification, and equitable outcome assessments. After revascularization, wound severity is most strongly associated with LEA risk. Ischemic and infectious grades confer additive, but not synergistic, risk. Future cluster analyses comparing specific WIfI presentations treated with and without revascularization will be required to further refine WIfI.