Complications after discharge and delays in adjuvant chemotherapy following colonic resection: a cohort study of linked primary and secondary care data

C S Arhi,E M Burns,G Bouras,P Aylin,P Ziprin,A Darzi

DOI: https://doi.org/10.1111/codi.14525

Abstract:Aim: By understanding the reasons for delays in adjuvant chemotherapy (AC) after colonic resection, there is the potential to improve patient outcome. The aim of this study is to determine the extent and impact of complications after hospital discharge on delays to AC. Method: The study cohort included patients from Hospital Episode Statistics (HES) who had a colorectal cancer resection; linkage to primary care data was provided by the Clinical Practice Research Datalink (CPRD). Complications during the index hospital stay (from HES) and after discharge (from CPRD) were compared. The risk of late AC treatment (8 weeks or later) following a complication, stoma at the index procedure or emergency admission was described after accounting for age and Charlson score. A Cox hazards model determined the association of these factors with overall survival (OS). Results: A total of 1266 patients underwent AC following colon cancer resection, of whom 598 (47.2%) received treatment within 8 weeks. Patients receiving late AC had a significantly higher proportion of re-operations (7.0% vs 3.3% P < 0.005) and wound infections (5.5% vs 3.7% P = 0.042), with 96% of the latter only being noted in CPRD. In multivariate analysis, the risk of AC delay significantly increased following a complication (OR 1.53, 95% CI 1.16-2.03, P = 0.003) or a stoma at the index operation. AC delay was associated with worse OS [hazard ratio (HR) 1.44, 95% CI 1.16-1.79, P = 0.001], as was an emergency admission (HR 1.59, 95% CI 1.21-1.98, P < 0.0005). However, the presence of a complication did not independently reduce OS (HR 1.15, 95%CI 0.89-1.48, P = 0.295). Conclusion: The true extent and impact of complications following colonic resection is underestimated when only secondary care data are used.