In a prospective multicentric cohort of patients intervened for colorectal cancer with 5 years follow-up, characteristics and clinical results will be compared, as quality of life and use of health services depending on their detection mode. In adjusted analysis variability among hospitals will be taken into account through hierarchical models and generalized linear mixed models will be applied for continuous outcomes (quality of life, use of health services) and dichotomous outcomes (reinterventions, complications).