Pre-operative variables including fitness associated with complications after oesophagectomy

Oesophagectomy is a technically-demanding operation associated with a high level of morbidity. We analysed the association of pre-operative variables, including those from cardiopulmonary exercise testing, with complications (logistic regression) and survival and length of stay (Cox regression) after scheduled transthoracic oesophagectomy in 273 adults, in isolation and on multivariate testing (maximum Akaike information criterion). On multivariate analysis, any postoperative complication was associated with ventilatory equivalents for carbon dioxide, odds ratio (95%CI) 1.088 (1.02-1.17), p = 0.018. Cardiorespiratory complications were associated with FEV₁ and pre-operative background survival (in an analogous group without cancer), odds ratios (95%CI) 0.55 (0.37-0.80), p = 0.002 and 0.89 (0.82-0.96), p = 0.004, respectively. Survival was associated with the ratio of expected-to-observed ventilatory equivalents for carbon dioxide and predicted postoperative survival, hazard ratios (95%CI) 0.17 (0.03-0.91), p = 0.039 and 0.96 (0.90-1.01), p = 0.076. Length of hospital stay was associated with FVC, hazard ratio (95%CI) 1.38 (1.17-1.63), p < 0.0001.