The adverse impact of pre-operative anaemia on survival following major lower limb amputation

Introduction: Approximately 5000 major lower-limb amputations (MLLA) for PAD occur per-annum in the UK with clinical outcomes being poor for this high-risk cohort of patients. Existing evidence suggests that anaemic surgical patients have an increased 30-day mortality, but this has not been explored in the context of MLLA. Recent prioritization processes suggested that MLLAs are a target area for research into outcome improvement. This cohort study evaluates the impact of anaemia on the outcome of MLLA to understand if optimization might improve outcomes. Methods: All PAD patients undergoing MLLA during 2015–2018 at a tertiary vascular centre were reviewed. Patients were stratified into groups; non-anaemia (>12 g/dL), mild-anaemia (12-10 g/dL) and severe-anaemia (<10 g/dL) by pre-operative haemoglobin (Hb). Primary outcome was overall survival by Kaplan–Meier. Secondary outcomes included length of stay (LOS), post-operative blood-transfusion, surgical-site infection (SSI) and myocardial infarction (MI). Cox-proportional-hazard and receiver-operator characteristics (ROC) analyses were conducted. Results: 345 patients were followed up over (mean) 23 months. 105 were non-anaemic, 111 mildly anaemic and 129 severely anaemic. Patients with severe-anaemia had a higher incidence of heart and renal failure (p = 0.003) than those with non- or mild-anaemia. Overall survival worsened significantly with increasing anaemia (p = 0.001). LOS was significantly longer in mild-anaemia which is 26 (16–43) days, (p = 0.006) and severe-anaemia of 28 days (17–40), (p < 0.001) compared to non-anaemia of 18 (10–30) days. Post-operative blood-transfusion (RBC) was required more frequently in 70.5% of severely anaemic patients (p < 0.001), compared to mildly anaemic (24.3%) and non-anaemic (7.6%) patients, with those receiving RBCs having a significantly worse survival. There was no difference in MI, SSI or wound dehiscence. Anaemia was significantly associated with mortality; (HR 1.7 (1.04–2.78), p = 0.03). A minimum-Hb of 10.4 g/L (by ROC) was identified as a cutoff Hb for an increased risk of mortality. Conclusion: Pre-operative anaemia is associated with worse outcome following MLLA, with increasing severity of anaemia associated with increasing mortality and RBC transfusion being potentially detrimental. More work is required to prospectively evaluate this relationship in this complex and multi-morbid cohort of patients.