'Prognostic risk modelling for patients undergoing major lower limb amputation: an analysis of the UK National Vascular Registry' was published in the European Journal of Vascular and Endovascular Surgery (EJVES) in 20201. The observational study examined risk factors for in-hospital mortality and morbidity following 9549 major lower limb amputations, and generated a risk model with online calculator named UKAmpRisk1. A subsequent systematic review and narrative synthesis of risk prediction tools used to estimate mortality, morbidity and other outcomes following major lower limb amputation was published in the EJVES in July 20212. It observed that although several risk prediction tools demonstrated acceptable to outstanding discrimination for objectively predicting post-operative complications, the majority of tools lack external validation and could not be recommended to support clinical activity. UKAmpRisk was found to have an acceptable discriminatory performance to predict in-hospital cardiac and renal morbidity, but borderline poor discrimination for respiratory complications2.

As an exercise in external validation of UKAmpRisk, we designed a small local study to see whether the outcomes from the large data-set of the UK National Vascular Registry translated to our busy vascular unit in East Kent. We obtained the details of all patients who underwent major lower limb amputation over a three-year period (Jan 2017 to Dec 2020) from our hospital's clinical coding department. All amputations were done after completion of the original study's data collection period to avoid replication. 155 amputations were found of which 154 were included in the analysis. We calculated the UKAmpRisk score for each patient and compared the presence or mean for each criteria with the original article alongside that study's outcomes of mortality and type of inpatient post-operative complication (Table 1).

Pre-operatively, our cohort of patients possessed a significantly higher mean serum creatinine (135 vs 81 μmol/L) and lower albumin (24 vs 30 g/L) values, whereas other criteria for the UKAmpRisk scoring system were similar (Table 1). In our cohort of patients, the UKAmpRisk calculator predicted a mean score of 8.3% mortality for all patients undergoing major lower limb amputation as either an emergency or planned procedure. This compared with the observed mortality of 5.8% (Table 1). Ambler et al reported the UK National Vascular Registry data of inpatient mortality of patients undergoing major lower limb amputation to be 9.1%. Our whole-cohort inpatient mortality was lower (5.8%), however post-operative morbidity was similar, with the exception of renal failure which was higher in our cohort (29% vs 4.3%) (Table 1). The most significant correlation with the lower mortality rate in our local study was our lower rate of unplanned admissions compared to the national dataset; highlighting the importance of early recognition of the deteriorating foot, as emphasised in the original study1.

As part of a multi-site NHS Trust, our unit is co-located with a tertiary renal centre, and is geographically separated from the general medical and surgical (including orthopaedic) teams. This could explain the significant differences in pre- and post-operative renal failure between the two cohorts, and also the numbers of unplanned traumatic amputations (with associated comorbidity) which may be undertaken by a local orthopaedic team when vascular surgery is not on-site. We were interested in the significant but potentially modifiable risk factors of pre-operative serum albumin, weight and WCC in planning non-urgent amputation and intend to be cognisant of this when planning surgery; indeed the importance of reducing these values may be a useful area of investigation in the future. When compared with our smaller cohort, UKAmpRisk appears to overestimate the mortality risk of major lower limb amputation (8.3 vs 5.8%) however, when ascertaining risk, and in the subsequent communication of likelihood of post-operative mortality with patients and their families, we believe an approximation of "5-10% chance of death" is an acceptable and useful metric for surgeons undertaking surgery. We therefore advocate the use of the UKAmpRisk scoring system and have implemented its use in our practice.

1. Ambler GK et al., Prognostic Risk Modelling for Patients Undergoing Major Lower Limb Amputation: An Analysis of the UK National Vascular Registry, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.12.006 2. Preece et al (2021). A Systematic Review and Narrative Synthesis of Risk Prediction Tools Used to Estimate Mortality, Morbidity, and Other Outcomes Following Major Lower Limb Amputation. European Journal of Vascular and Endovascular Surgery, 62(1), 127-135. https://doi.org/10.1016/j.ejvs.2021.02.038