Factors influencing mortality and morbidity following colorectal resection in France. Analysis of a national database (2009-2011).

Aim: Correlation between outcome and hospital volume regarding colorectal resection (CRR) has been described, but it suggests that provider variability may have an impact. Our aim was to analyse the influence of institutional characteristics and the impact of volume [high volume (HV) or low volume (LV)] on mortality and morbidity after CRR at a national level.

Method: Data from 2009-2012, including patient demographics, diagnosis, procedure, mode of admission and discharge and hospital type, were obtained. Each hospital admission was classified as one of four levels of severity.

Results: Of 176,444 patients included, 5408 (3.06%) died and 41,240 (23.37%) had a complication. Multivariate analysis showed that factors influencing morbidity were age over 80 years, severity level, pathology other than diverticular disease, male gender, demanding surgery, open surgery and surgery in an HV institution. Factors influencing mortality were the same except for the impact of volume. In HV centres, surgery was significantly more demanding (54.66% vs 47.17%, P < 0.0001), morbidity more frequent (26.59% vs 22.07%, P < 0.0001), but mortality was lower (2.17% vs 3.43%, P < 0.0001). In total, 6038 (3.4%) patients were transferred after surgery. Transfer rate and mortality after transfer were significantly higher in LV institutions (respectively: 4.3% vs 2.5%, P < 0.0001; and 12% vs 10.3%, P < 0.0001).

Conclusion: High volume centres have higher morbidity, but lower mortality. Six per cent of patients in LV centres required transfer. A national mortality rate after CRR of 3.5% can be expected. Transfer rate and mortality after transfer should be included in the evaluation of institutional mortality. Volume of institution, regardless of type, influences mortality after CRR.