ABSTRACT
BACKGROUND: This study evaluated the outcomes of the first 5 years of sentinel node biopsy (SNB) in British Columbia (BC), Canada, 1996 to 2001. METHODS: There were 547 SNB procedures for breast cancer performed by 29 surgeons at 12 hospitals in BC between October 1996 and July 2001. Identification, accuracy, and false-negative rates were determined and correlated to patient, tumor, and surgical factors with the chi-square test. RESULTS: SNB mapping was performed using blue dye alone (15%), radiopharmaceutical alone (6%), or both (79%). A completion axillary dissection was performed in 93%. A median of 2 (range 1 to 16) sentinel nodes was biopsied. The overall identification rate was 88%, accuracy was 92%, and false-negative rate was 22%. All rates were improved in younger (age <50 years) compared with older women. A positive lymphoscintiscan and the mapping agent used were associated with higher identification rates but not accuracy or false negative rates. Increasing surgeon experience was not significantly associated with improvements in identification or false-negative rates. CONCLUSIONS: The potential of SNB was not fully translated into surgical practice in BC by 2001.
