Relationship between pre-operative glycated haemoglobin and surgical site infection in patients undergoing elective colon cancer surgery.

Surgical site infections remain a significant cause of morbidity following colon cancer surgery. Although diabetes has been recognised as a risk factor, patients with asymptomatic diabetes are likely underdiagnosed. The aim of the present study was to determine the relationship between preoperative glycated haemoglobin (HbA1C), clinicopathological characteristics and the influence on surgical site infection in a cohort of patients undergoing potentially curative colon cancer surgery. Patients who underwent elective, potentially curative colon cancer surgery between January 2011 and December 2014 were assessed for HbA1C levels (mmol/mol) measured within 3 months preoperatively. Clinicopathological data were recorded in a maintained database. A multivariate binary logistic regression model was used to assess the relationship between HbA1C, clinicopathological characteristics and surgical site infections. A total of 362 patients had HbA1C levels preoperatively recorded. HbA1C was significantly associated with body mass index (BMI), diabetes, smoking status, visceral fat area and skeletal muscle index. As determined by multivariate analysis, preoperative HbA1C levels remained independently associated with an increased risk of surgical site infections (OR 1.69, 95% CI 1.05-2.7; P=0.031) together with BMI (OR 1.91, 95% CI 1.36-2.67; P<0.001). Notably, in the present study, tumour-based factors, such as tumour location and TNM status, were not associated with infective complications. By contrast, host factors, such as BMI and pre-operative HbA1C were associated with surgical site infections suggesting that these factors were of more importance in determining short-term outcomes. In conclusion, objective measurements of BMI and HbA1C effectively stratified the risk of developing surgical site infection from 8 to 59%; therefore, HbA1C levels should be determined to allow for preoperative optimisation.