Abstract
BACKGROUND: Lower limb amputation (LLA) is a devastating complication in dialysis patients, often related to diabetes and peripheral vascular disease. National data on LLA incidence, variation and outcomes in England are limited. METHODS: A retrospective cohort study conducted using linked data from the UK Renal Registry (UKRR) and Hospital Episode Statistics for incident dialysis patients in England (2003-17). Cumulative incidence plots assessed LLA rates across three patient subgroups: primary renal disease (PRD) diabetes, diabetes as comorbidity and no diabetes. Fine-Gray competing risk models identified independent risk factors for LLA. One-year mortality following LLA was presented as Kaplan-Meier curves and analysed using age- and sex-adjusted standardized mortality rates (SMRs). RESULTS: Among 69 701 incident dialysis patients, 9.1% with PRD diabetes underwent LLA within 3 years, versus 2.6% with diabetes as comorbidity and 0.6% without diabetes. Risk factors for LLA included diabetes [subdistribution hazard ratio (sHR) 9.3 for PRD diabetes, sHR 3.6 for diabetes as comorbidity], peripheral vascular disease (sHR 2.11), prior amputation (sHR 2.19), ischaemic heart disease and White ethnicity. Major amputation rates declined over time, while minor amputation rates remained stable. Significant inter-centre variation was observed after adjustment. One-year mortality after LLA exceeded 40% in all subgroups. SMRs were highest in the PRD diabetes group compared with the general population (34.7), but highest in the diabetes as comorbidity group compared with the dialysis population (5.31). CONCLUSION: Incident dialysis patients, particularly those with PRD diabetes, face high LLA risk and very poor survival after amputation. Persistent inter-centre variation suggests scope for national audit, standardized foot care and preventive strategies.