The impact of comorbidity on the overall survival and the cause of death in patients after colorectal cancer resection.

Background: Retrospective investigation to identify associations between certain patient characteristics and survival in 531 patients with resected colorectal cancer (CRC). Special reference is given to a standardized comorbidity.
Methods: To compare different levels of exposure we determined hazard ratios (HR) in Cox proportional hazards models for survival times and odds ratios (OR) in logistic regression models.
Results: Overall survival was associated with tumor stages (III+IV vs. I+II; HR 7.48), tumor differentiation (low vs. high; HR 1.84), blood transfusions (>2 vs. < or =2; HR 1.88), and comorbidity (Charlson Comorbidity Index >2 vs. < or =2; HR 1.77). Low tumor stage (I+II vs. III+IV; OR 11.1), elevated Charlson Comorbidity Index (>2 vs. < or =2; OR 3.83), and longer ICU stay (>2 days vs. < or =2 days; OR 3.40) more frequently lead to non-cancer-related death than to cancer-related death.
Conclusion: Standardized comorbidity should be considered as a factor in survival studies of CRC.