Region Better Predicts Amputation Risk than Many Patient-Level Factors after Common Lower Extremity Procedures
Thomas FX O'Donnell, MD1, Chun Li, MD1, Nicholas J. Swerdlow, MD1, Peter A. Soden, MD1, Ageliki G. Vouyouka, MD2, Joseph L. Mills, MD3, John W. Hallett, Jr., MD4, Marc L. Schermerhorn, MD1.
1Beth Israel Deaconess Medical Center, Boston, MA, USA, 2Mount Sinai, New York, NY, USA, 3Baylor College Of Medicine, Houston, TX, USA, 4Medical University of South Carolina, Charleston, SC, USA.
OBJECTIVES: Prior studies identified significant regional variation in amputation rates. However, they included heterogeneous cohorts, and failed to quantify the impact of region on subsequent amputation risk. To that end, we studied amputation rates after four common lower extremity procedures, and quantified the regional effects.
METHODS: We identified all patients with occlusive disease undergoing first-time femoropopliteal bypass, tibial/pedal bypass, isolated superficial femoral artery (SFA) intervention, or isolated tibial intervention between 2009-2016 in the 18 de-identified regions of the VQI. We excluded regions with <100 procedures. We utilized multilevel logistic regression, clustering at the region and center level, to calculate risk-adjusted 30-day amputation rates following each procedure, and quantify the degree to which regional and center-level variation contributes to outcomes using the median odds ratio (MOR). The MOR is the median of the set of odds ratios for patients with similar comorbidities between the region with the highest risk and the region with the lowest risk.
RESULTS: We identified 24,966 patients: 6,936 first-time femoropopliteal bypasses; 6,521 tibial/pedal bypasses; 8,300 isolated superficial femoral artery intervention; and 2,591 isolated tibial interventions. The overall adjusted 30-day amputation rate after first time femoropopliteal bypass was 0.6%, but rates varied from 0-2.3% across the regions. Overall adjusted 30-day amputation rate following tibial/pedal bypass was 1.7%, with regional rates varying from 0.2-4.6%. Adjusted 30-day amputation rates following SFA intervention was 0.9% in the overall cohort, with regional rates varying from 0.3-2.7%. Adjusted 30-day amputation rates following tibial interventions was 4.5% overall, with regional rates varying from 2.9-11.8%. The region in which a patient underwent their procedure was one of the most significant predictors of their subsequent risk of amputation after each procedure except tibial interventions (femoropopliteal bypass: MOR 2.15 [80% IOR 1.38-2.92], tibial bypass: MOR 1.48 [1.15-1.81], SFA: MOR 1.33 [1.07-1.59]), Table 1. Following tibial interventions, center, but not regional variation was associated with increased risk of amputation (MOR 1.54 [1.18 - 1.91]). CONCLUSIONS: The region in which patients undergo procedures better predicts amputation risk than many patient level factors. Further efforts should aim at determining best practices to decrease unwarranted variation and improve patient-level outcomes. 
Back to 2018 Program
