Risk factors contributing to 30-day and 1-year mortality event scores following major lower extremity amputation for limb ischemia.

Abstract

The purpose of this study was to create a risk score for 30-day and 1-year mortality after major lower extremity amputation to facilitate clinical expectations and the identification of patients in need of heightened vigilance in longitudinal care. In the Vascular Quality Initiative, 25,150 patients were identified who underwent lower extremity amputation. Two primary outcomes were investigated: 30-day mortality after major lower extremity amputation and 1-year mortality after amputation. Univariable analysis for the 30-day and 1-year mortality analysis was conducted with χ² analysis. Significant (P < .05) univariable factors were then included in binary logistic regression analysis to perform multivariable investigation toward the outcomes. Variables that achieved multivariable significance (P < .05) for the outcomes were then used in the respective risk scores with the regression beta coefficient being used to weigh the variables. Overall, 7.2% of patients experienced 30-day mortality and 22.4% suffered mortality within 1 year. Variables with a significant multivariable association (P < .05) with 1-year mortality were female sex, advancing age, body mass index less than 20 kg/m², coronary artery disease, history of coronary revascularization, congestive heart failure (CHF), chronic obstructive pulmonary disease, dialysis status at presentation, baseline renal insufficiency, anemia, lack of statin medication at presentation, being on anticoagulation at presentation, need for above-knee amputation, and need for emergent amputation. Variables that were protective vs mortality on multivariable analysis (P < .05) were body mass index greater than 30 kg/m², history of ipsilateral amputation at a lower level, and history of ipsilateral infrainguinal bypass. Pertinent negatives included all socio-demographic variables including rural living status, insurance status, and area deprivation index home area. Variables with a statistically significant (P < .05) multivariable association with 30-day mortality were female sex, advancing age, history of coronary artery disease, history of prior coronary revascularization, history of CHF, class 3 or 4 CHF, chronic obstructive pulmonary artery disease, dialysis requirement at presentation, baseline renal insufficiency, lack of antiplatelet medication at time of presentation, lack of statin medication at time of presentation, need for above-knee amputation, acute ischemia indication, and need for emergent amputation. Variables that were protective (P < .05) vs 30-day mortality on multivariable analysis were diabetes, prior ipsilateral amputation at a lower level, prior infrainguinal bypass ipsilateral to amputation, prior ipsilateral infrainguinal endovascular revascularization (endovascular or bypass), and prior ipsilateral inflow arterial bypass. For receiver operating curve analysis, the 30-day risk score had an area under the curve of 0.715 and the 1-year risk score analysis 0.722. Hosmer-Lemeshow investigation of the multivariable regressions resulted in an overall accuracy of 92.8% for the 30-day mortality investigation (99% accurate for survival) and 78% overall accuracy on the 1-year mortality risk score (96.8% in predicting survival accurately). This indicates that these models perform far better in determining which patients will survive rather than precisely determining who will experience 1-year mortality. Risk scores for mortality at 30 days and 1 year after major lower extremity amputation have been created that have good accuracy and steep escalation with advancing comorbidity. Variables with the most potent deleterious effect on survival were renal insufficiency, dialysis requirement, CHF, and need for emergent amputation. Patients who have already been receiving care from vascular specialists at the time of amputation have improved survival vs those without prior arterial interventions or a lower level of amputation. Social determinants of health do not impact survival among patients undergoing major lower extremity amputation at Vascular Quality Initiative centers.