Is colon cancer management in a comprehensive cancer centre evidence-based?

Between 1993 and 1994, the 80 physicians in the French comprehensive cancer centre, Léon Bérard, developed and implemented Clinical Practice Guidelines (CPGs) based on an analysis of the literature and a␣consensus of intra-institutional experts. This study reports the impact of CPGs on the management of colon cancer and the identification of patients' characteristics linked with CPGs' compliance rates.

A 'before-after' study, using records of colon cancer patients, was performed. Records for all patients newly referred with colon cancer in 1993 and 1995 (n = 77 and n = 81, respectively) were selected in 1993 and 1995. Medical decisions on these records were analysed to assess their compliance with the CPGs. Univariate and multivariate analyses of compliance factors for the CPGs were performed on the cohort of 128 patients. A systematic search of the literature was performed to determine the scientific evidence for non-compliant decisions.

The compliance rate for colon cancer was significantly higher in 1995 than in 1993, 70% (62/88) (95% confidence interval (CI), 60-80%) vs 50% (34/68) (95% CI, 38-62%) (P = 0.009). Using logistic regression, year of management 1995 (OR 2.8; 95% CI, 1.12-6.9), metastatic stage (OR 4.12; 95% CI, 1.43-11.8), and Karnofsky scale > 70 (OR 11.5; 95% CI, 1.14-41.4) were identified as independent factors for compliance with the CPGs. In 1993 and 1995, 71% (48/68) (95% CI, 60-81%) and 81% (71/88) (95% CI, 73-89%), respectively, of medical decisions for colon cancer, conformed to the CPGs or were judged to be based on 'scientific evidence'.

Compliance rates were significantly higher in 1995. The development and implementation of CPGs for cancer management seemed to result in significant changes in medical practice. Moreover, the patients' characteristics appeared to influence compliance with the guidelines.