Purpose: To compare the premorbid profile of participants who died within  three months after lower limb amputation (LLA) to that of survivors.

Relevance: The premorbid status of patients surviving a non-traumatic lower limb amputation in the Johannesburg metropolitan area is unknown. It may be useful to know the premorbid differences in demographic and functional characteristics between patients who survive versus those who die after a lower limb amputation in order to intervene in those patients who are at risk.

Participants: Consecutive sampling was used to recruit one hundred and fifty-four participants, with 147 being eligible for analysis.

Methods:  This was a quantitative cross- sectional comparison study.  Vascular amputees  were included if they were scheduled for first time unilateral lower limb amputation. Participants with co-morbidities that interfered with function pre-operatively were excluded . Institutional ethical clearance was obtained (Ethical clearance no. M110124) and the candidate hospitals and participants gave  informed consent. A demographic questionnaire, the Barthel Index (BI), the Participation Scale (P-Scale) and the EuroQol Quality of life measure (EQ-5D) were used to interview participants prior to the amputation. Participants were followed up to 3 months to determine survival.

Analysis: IBM SPSS 21 was employed to analyse the data. Descriptive statistics were used to reduce the data. Following data testing for statistical assumptions and distribution, premorbid differences in function, participation and quality of life were analyzed using non-parametric tests for medians of independent samples and Pearson’s Chi square was used to test categorical data.

Results: Of the 154 entrants recruited, n=147 were followed up and death was confirmed in n=29 (19%, n=29/154), n=7 were lost to follow up and n=118 were alive. There were no differences in the premorbid median scores for BI (median=20), P-Scale (median=0), EQ-5D index (median=0.264 and 0.193 for the survivors and demised respectively), EQ-5D VAS (median=75 and 70 for the survivors and demised respectively) and age (median=58 and 62 for the survivors and demised respectively) at baseline (p˃0.05). Individual items (categorical data) of the BI, P-Scale, EQ-5D also showed no differences between the two groups (p˃0.05) at baseline. Those who died drank significantly more alcohol (p=0.02) and smoked significantly more than those who survived (p=0.002).

Conclusion:  Although the groups were generally comparable in activity levels, participation levels, quality of life and age, being a smoker and drinking alcohol premorbidly seems to decrease the chance of survival following LLA.

Implications: This study helps to identify the profile of patient who most likely will not survive three months after a LLA. These findings may help strengthen preoperative programs to minimize death after a LLA especially efforts to emphasise lifestyle modifications long before disability sets in.

Funding source acknowledgement: We thank the MRC grant, Carnegie grant, Faculty Research Fund for the financial  assistance  that helped us to fund the study.

Keywords: lower limb amputation, survival, quality of life,