Abstract
OBJECTIVE: The purpose of this study is to investigate variation in great saphenous vein (GSV) use among the various centers participating in the Vascular Quality Initiative infrainguinal bypass modules. Further, differences in outcomes in femoral-popliteal artery bypass with single segment GSV conduit vs prosthetic conduit will be documented. Center GSV use rate impact on outcomes will be investigated. METHODS: Primary exclusions were patients undergoing redo bypass, urgent or emergent bypass, and those in whom prosthetic graft was used while having undergone prior coronary artery bypass grafting. The distribution of GSV use across the 260 centers participating in Vascular Quality Initiative infrainguinal bypass module was placed into histogram and variance in mean GSV use evaluated with analysis of variance analysis. Centers that used GSV in >50% of bypasses were categorized as high use centers and centers that used the GSV in <30% of cases were categorized as low use centers. Baseline differences in patient characteristics and comorbidities in those undergoing bypass with GSV vs prosthetic conduit were analyzed with χ(2) testing and the Student t test, as were those undergoing treatment in high vs low use centers. Multivariable time-dependent Cox regression analyses were then performed for the primary outcomes of major amputation ipsilateral to the operative side and mortality in long-term follow-up. High vs low use center was a dichotomous variable in these regressions. Secondary outcomes of freedom from graft infection and freedom from loss of primary patency were performed with Kaplan-Meier analysis. RESULTS: Among centers with >50 patients meeting inclusion criteria for this study, GSV use ranged from 15% to 93% (analysis of variance P < .001). When considering all centers irrespective of number of patients, the range was 0% to 100%. On Kaplan-Meier analysis, GSV conduit use was associated with improved freedom from loss of primary or primary assisted patency, improved freedom from major amputation after index hospitalization, improved freedom from graft infection after the index hospitalization, and improved freedom from mortality in long-term follow-up (log-rank P < .001 for all four outcomes). Both low use center (hazard ratio, 1.35; P < .001) and prosthetic graft use (hazard ratio, 1.24; P < .001) achieved multivariable significance as risks for mortality in long-term follow-up. Other variables with a multivariable mortality association are presented in the manuscript. Low use center and prosthetic bypass were significant univariable but not multivariable risks for major amputation after index hospitalization. CONCLUSIONS: There is remarkably wide variation in GSV use for femoral popliteal artery bypass among various medical centers. GSV use is associated with enhanced long-term survival as well as freedom from loss of bypass patency and graft infection. The data herein indicate institutional patterns of prosthetic conduit choice, which has the potential to be altered to enhance outcomes.