A controlled ‘before-after’ study: impact of a clinical guidelines programme and regional cancer network organization on medical practice

A regional cancer network has been set up in the Rhône-Alpes region in France. The aim of the project is to improve the quality of care and to rationalize prescriptions in the network. In this network, we assessed the impact of the implementation of a clinical practice guidelines project by assessing the conformity of practice with the guidelines and comparing this with the conformity in an external matched control group from another French region without a regional cancer network. Four hospitals (private and public) accepted to assess the impact of the clinical practice guidelines on the management of breast and colon cancer in the experimental group and three hospitals (private and public) in the control group. In 1994 and 1996, women with non-metastatic breast cancer (282 and 346 patients in the experimental group, 194 and 172 patients in the control group, respectively) and all new patients with colon cancer (95 and 94 patients in the experimental group, and 89 and 118 patients in the control group, respectively) were selected. A controlled ‘before-after’ study, using institutional medical records of patients with breast and colon cancer. The medical decisions concerning the patients were analyzed to assess their compliance with the clinical practice guidelines. When medical decisions were judged to be non-compliant, we verified if they were based on scientific evidence in a published article, if they were not, the medical decision was classified as having ‘no convincing supporting scientific evidence’ The compliance rates were significantly higher in 1996 than in 1994 in the experimental group; 36% (126 out of 346) vs 12% (34 out of 282) and 46% (56 out of 123) vs 14% (14 out of 103) (P<0.001) for breast and colon cancer, respectively. Whereas, in the control group the compliance rates were the same for the two periods; 7% (12 out of 173) vs 6% (12 out of 194) (P=0.46) and 39% (49 out of 126) vs 32% (31 out of 96), P=0.19. In the experimental group, in 1994, 101 of the 282 medical decisions (36%) and 27 of the 103 (26%) for breast and colon cancer, respectively, were classified as having ‘no convincing scientific evidence’ compare with 72 out of 346 in 1996 (21%) for breast cancer, and 21 of the 123 (17%) for colon cancer P<0.05. Whereas in the control group these results were 106 out of 194 in 1994 (55%) and 90 out of 172 in 1996 (52%), P=0.65 for breast cancer and 28 out of 96 in 1994 (29%) and 30 out of 126 in 1996 (24%), P=0.36 for colon cancer. The development and implementation strategy of the clinical practice guidelines programme for cancer management results in significant changes in medical practice in our cancer network. These results would suggest that introducing guidelines with specific implementation strategy might also increase the compliance rate with the guideline and ‘evidence-based medicine’. British Journal of Cancer (2002) 86, 313–321. DOI: 10.1038/sj/bjc/6600057 www.bjcancer.com © 2002 The Cancer Research Campaign