COX regression analysis of colorectal cancer prognostic factors

Objective To study prognostic factors of colorectal cancer using COX regression analysis. Methods Serum tumor markers, tumor location, Ducks stage and other 11 related physiological indicators of 598 cases of colorectal patients were performed by univariate Kaplan-Meier survival analysis and multivariate COX proportional hazards model analysis by SAS 9.2 software. Results 598 cases of colorectal cancer 1, 3, 5-year survival rates were 91.79%, 56.14%, 21.89% respectively; Ducks stage, age, pathological types, serum CEA, CA242, CA19-9, tumor scales and lymphatic metastasis were influential factors of colorectal cancer (P0.05). Multivariate analysis revealed that the Ducks stage, age, pathological types, serum CEA and CA242 were influential factors of colorectal cancer. Conclusions To determine clinical protocols and prognosis for colorectal cancer patients, it should fully consider Ducks stage, age, pathological types, serum CEA and CA242.