Abstract
Aim
To describe short and longer-term mortality following major lower extremity amputation (LEA) by diabetes status over two time periods.
Methods
A retrospective cohort study of patients who underwent major LEA between 2004 and 2013 was conducted based on linkage of national population-based hospital records and a register of people with diagnosed diabetes. Post-operative mortality was estimated at 30 days, one year and where available, five years. Using logistic regression models, we estimated the odds of death associated with diabetes adjusted for age, sex and socio-economic status within these time points compared to the non-diabetic population stratified by type of diabetes and five-year calendar periods.
Results
There were a total of 5436 people who received an amputation during the study period of whom approximately 40% had diabetes. Overall mortality for the 2004-8 and 2009-13 cohorts respectively was not significantly different at 7.9% and 7.3% at 30 days and 31% and 27% at one year. Almost 64% of the 2004-8 cohort were dead within five years. The only statistically significantly associations between diabetes and mortality were observed within five year follow-up of the 2004-8 cohort with odds ratios (95% CI) compared to the non-diabetic population of 1.62 (1.17, 2.26) for type 1 diabetes and 1.38 (1.14, 1.66) for type 2 diabetes.
Conclusions
An adverse association between diabetes and mortality after LEA only became apparent in longer term follow-up.