Practice setting and physician influences on judgments of colon cancer treatment by community physicians.

OBJECTIVE: This article compares judgments about the treatment of Dukes' B2 and C colon cancer made by general surgeons to those of internists and family practitioners. Physician and practice variables were specialty, affiliation with a Community Clinical Oncology Program (CCOP) hospital, time in practice, professional centrality (level of participation in cancer information networks), solo practice, and number of colon cancer patients. DATA COLLECTION METHODS: Data are combined from national probability samples of CCOP- and non-CCOP-affiliated physicians. This study focused on 1,138 internists, family physicians, and general surgeons who participated in decision making for patients diagnosed with Dukes' B2 or C stage colon cancer. Judgments were elicited using brief vignettes. METHODS OF ANALYSIS: Judgments of adjuvant therapy are classified as (a) consistent with the National Institutes of Health Consensus Conference recommendations (experimental for Dukes' B2, accepted for Dukes' C); (b) accepted treatment for both stages; or (c) experimental for both stages. Multinomial logit analyses were used to examine the association of practice setting and physician characteristics to judgments of treatment. RESULTS: Surgeons and CCOP-affiliated physicians were more likely to endorse the NIH consensus conference position. Surgeons, younger physicians, and those in group practice were more likely to approve of chemotherapy for both cancer stages. The most common position (chemotherapy experimental) was more likely from nonsurgeons, solo practitioners, and non-CCOP physicians. CONCLUSION: Physician and practice setting characteristics, including organized structures such as the CCOP, are possible mediating structures that can facilitate dissemination of standards of treatment