The effect of pre-operative optimization on post-operative outcome in Crohn's disease resections.

BACKGROUND: The timing of surgical intervention in Crohn's disease (CD) may depend on pre-operative optimization (PO) which includes different interventions to decrease the risk for unfavourable post-operative outcome. The objective of this study was to investigate the effect of multi-model PO on the post-operative outcome in CD.
METHOD: This is a multicentre retrospective cohort study. The primary outcome was 30-day post-operative complications. Secondary outcomes were intra-abdominal septic complications, surgical site infection (SSI), re-operation, length of post-operative stay in a hospital and re-admission. PO included nutritional support, discontinuation of medications, pre-operative antibiotic course and thrombosis prophylaxis.
RESULTS: Two hundred and thirty-seven CD elective bowel resections were included. Mean age was 39.9 years SD 14.25, 144 (60.8 %) were female and 129 (54.4 %) had one or more types of medical treatment pre-operatively. Seventy-seven patients (32.5 %) optimized by at least nutritional support or change in pre-operative medications. PO patients were more likely to have penetrating disease phenotype (p = 0.034), lower albumin (p = 0.015) and haemoglobin (p = 0.021) compared to the non-optimized. Multivariate analyses showed that treatment with anti-TNF alpha agents OR 2.058 CI [1.043-4.4.064] and low haemoglobin OR 0.741 CI [0.572-0.0.961] increased the risk of overall post-operative complications. Co-morbidity increased the risk of SSI OR 2.567 CI [1.182-5.576] while low haemoglobin was a risk factor for re-admission OR 0.613 CI [0.405-0.926]. Low pre-operative albumin correlated with longer stay in hospital.
CONCLUSIONS: PO did not change post-operative outcome most likely due to selection bias. Anti-TNF alpha agents, low haemoglobin, low albumin and co-morbidity were associated with unfavourable outcome.