Factors predicting one-year post-surgical mortality amongst older Asian patients undergoing moderate to major non-cardiac surgery - a retrospective cohort study.

BACKGROUND: While short-term perioperative outcomes have been well studied in Western surgical populations, the aim of this study is to look at the one-year perioperative mortality and its associated factors in an Asian surgical population after non-cardiac surgery. METHODS: A retrospective cohort study of 2163 patients aged above 45 undergoing non-cardiac surgery in a university-affiliated tertiary hospital from January to July 2015 was performed. Relevant demographic, clinical and surgical data were analysed to elicit their relationship to mortality at one year after surgery. A univariate analysis was first performed to identify significant variables with p-values ≤ 0.2, which were then analysed using Firth multiple logistic regression to calculate the adjusted odds ratio. RESULTS: The one-year mortality in our surgical population was 5.9%. The significant factors that increased one-year mortality include smoking (adjusted OR 2.17 (1.02-4.45), p = 0.044), anaemia (adjusted OR 1.32 (1.16-1.47), p < 0.001, for every 1 g/dL drop in haemoglobin level), lower BMI (adjusted OR 0.93 (0.87-0.98), p = 0.005, for every 1 point increase in BMI), Malay and Indian ethnicity (adjusted OR 2.68 (1.53-4.65), p = 0.001), peripheral vascular disease (adjusted OR 4.21 (1.62-10.38), p = 0.004), advanced age (adjusted OR 1.04 (1.01-1.06), p = 0.004, for every one year increase in age), emergency surgery (adjusted OR 2.26 (1.29-3.15), p = 0.005) and malignancy (adjusted OR 3.20 (1.85-5.52), p < 0.001). CONCLUSIONS: Our study shows that modifiable risk factors such as malnutrition, anaemia and smoking which affect short term mortality extend beyond the immediate perioperative period into longer term outcomes. Identification and optimization of this subset of patients are therefore vital. Further similar large studies should be done to develop a risk scoring system for post-operative long-term outcomes. This would aid clinicians in risk stratification, counselling and surgical planning, which will help in patients' decision making and care planning.