Factors associated with amputation or graft occlusion one year after lower extremity bypass in northern New England.

Background: Optimal patient selection for lower extremity bypass surgery requires surgeons to predict which patients will have durable functional outcomes following revascularization. Therefore, we examined risk factors that predict amputation or graft occlusion within the first year following lower extremity bypass.

Methods: Using our regional quality-improvement initiative in 11 hospitals in northern New England, we studied 2,306 lower extremity bypass procedures performed in 2,031 patients between January 2003 and December 2007. Sixty surgeons contributed to our database, and over 100 demographic and clinical variables were abstracted by trained researchers. Cox proportional hazards models were used to generate hazard ratios and surrounding 95% confidence intervals (CIs) for our combined outcome measure of major amputation (above-knee or below-knee) or permanent graft occlusion (loss of secondary patency) occurring within the first year postoperatively.

Results: We found that within our cohort of 2,306 bypass procedures 17% resulted in an amputation or graft occlusion within 1 year of surgery. Of the 143 amputations performed (8% of all limbs undergoing bypasses), 17% occurred in the setting of a patent graft. Similarly, of the 277 graft occlusions (12% of all bypasses), 42% resulted in a major amputation. We identified eight preoperative patient characteristics associated with amputation or graft occlusion in multivariate analysis: age <50, nonambulatory status preoperatively, dialysis dependence, diabetes, critical limb ischemia, need for venovenostomy, tarsal target, and living preoperatively in a nursing home. While patients with no risk factors had 1-year amputation/occlusion rates that were <1%, patients with three or more risk factors had a nearly 30% chance of suffering amputation or graft occlusion by 1 year postoperatively. When we compared risk-adjusted rates of amputation/occlusion across centers, we found that one center in our region performed significantly better than expected (observed/expected ratio 0.7, 95% CI 0.6-0.9, p < 0.04).

Conclusion: Preoperative risk factors allow surgeons to predict the risk of amputation or graft occlusion following lower extremity bypass and to more precisely inform patients about their operative risk and functional outcomes. Additionally, our model facilitates comparison of risk-adjusted outcomes across our region. We believe quality-improvement measures such as these will allow surgeons to identify best practices and thereby improve outcomes across centers.