The outcome of diabetes mellitus-associated major lower limb amputations in a tertiary care hospital in Sri Lanka: morbidity, mortality, and rehabilitation access

Introduction: Developing countries such as Sri Lanka are seeing an increase in the prevalence of diabetes mellitus and related complications, including major lower limb amputations. This leads to high morbidity and mortality.
Objectives: To evaluate the outcomes of diabetes mellitus-associated major lower limb amputations at a specialised tertiary care centre in Sri Lanka
Methodology: Patients who underwent major lower limb amputations due to diabetes mellitus-related complications from June 2022 to July 2023 were voluntarily recruited. Patients were followed up for a year at clinics and via telephone calls. Student's t-test, chi-squared test, regression models and Kaplan-Meier survival tables were used for results analysis.
Results: Of the 149 patients, the mean age was 63.7 years (SD 10.7). 30.9% (n=46) were females, and 83.9% (n=125) of the patients had undergone below-knee amputations (BKA). After one year, 28.2% (n=42) of patients were deceased and the commonest cause of death was myocardial infarction (59.6%, n=25). Of postoperative complications, surgical site infection was the commonest immediate complication (14.8%, n=22); phantom limb sensation was the commonest late-onset local complication (49.7%, n=74) and myocardial infarction was the most common late-onset systemic complication (37.6%, n=56). In multivariate analysis, binary logistic regression identified smoking (current or previous) as a significant independent predictor of one-year mortality (95% CI: 1.5–4.5, p=0.018). According to Cox regression, both smoking (95% CI: 1.8–6.0, p=0.015) and fasting blood glucose levels (95% CI: 1.0–1.3, p=0.017) significantly reduced survival time. Other risk factors were not statistically significant.
Conclusion: Diabetes-related major lower limb amputations are associated with substantial one-year mortality, largely attributable to myocardial infarction, alongside a high burden of postoperative complications. Both immediate complications, particularly surgical site infections, and late complications such as phantom limb phenomena and cardiovascular events were common. Long-term functional recovery was limited. Smoking and elevated fasting blood glucose emerged as significant predictors of reduced survival. Elevated fasting plasma glucose was significantly associated with reduced survival time in the Cox regression model, but it was not a significant predictor of one-year mortality in the binary logistic regression analysis, indicating an effect on the timing rather than the occurrence of death.