Abstract 18259: Overestimation of Major Amputation Risk in Patients Undergoing Peripheral Vascular Intervention for Peripheral Artery Disease When Not Accounting for Mortality Risk

Background: Endovascular peripheral vascular intervention (PVI) is intended to improve symptoms and functioning, and avoid limb amputation in patients with chronic limb threatening ischemia (CLTI). CLTI is associated with high mortality risk, which differs across age groups. We aimed to offset mortality vs. amputation risks by age groups in patients with CLTI using competing risk modeling.

Methods: Patients with CLTI underwent PVI in 2017-2018 in the Vascular Quality Initiative registry were included. Mortality and major amputation outcomes over 3 years were derived from Medicare data. The cumulative incidence (CIF) of both outcomes in age groups <65 and ≥65 years, was calculated by Aalen-Johansen (AJ), taking mortality into account, and compared with corresponding CIF calculated by Kaplan-Meier (KM). The major amputation risk in <65 vs ≥65 years was estimated using a Fine and Gray competing risk model (sHR) and Cox regression (HR) adjusted for patients’ baseline characteristics.

Results: A total of 10,114 patients were included: 22% <65 years; 59% male and 19% Black. The cumulative incidence of death without major amputation increased with age (32% in <65 vs. 46% ≥65 years) and was higher than that of major amputation (26% in <65 vs. 15% in ≥65 years). By ignoring competing risk, both death without major amputation and major amputation was overestimated by a relative difference of 25.4% and 14.3% in <65 years and by 13.8% and 21.2% in ≥65 years. The adjusted major amputation risk in <65 vs. ≥65 years estimated by Cox regression was HR 1.30 95%CI 1.16-1.56 vs. FG model sHR = 1.75 95%CI 1.57-1.95. (Table)

Conclusions: Longer term mortality following PVI is high, especially among elderly with CLTI. To derive accurate risk estimates for amputation across age groups, advanced methods that can accommodate the competing risk of mortality are needed to better inform clinicians and patients as they face CLTI treatment decisions.