Breast cancer statistics: use and misuse

Statistics can be manipulated, by using various methods of reporting, to support almost any type of regime in breast cancer. The basis for statistical calculations include: the definition of the population of patients treated, the time used for starting calculations, the exact nature of the treatment used and any adjuvant therapy, the prognostic parameters utilized, the importance of long-term follow-ups, adequate number of cases, therapy standardization and pathological reporting, methods of evaluating survival such as observed or crude, relative and no evidence of disease, methods of calculating observed survival rates such as absolute, actuarial or life-table and Kaplan Meier or product limit, statistical evaluation for planned improvement with acceptable Type I and Type II errors, and the use of arithmetic and logarithmic scales in plotting statistics. Purveyors of innovative methods for the treatment of breast cancer aimed at preserving part or all of the breast as cosmetic alternatives to mastectomy, by limited operations with and without primary radiotherapy, have a most appealing argument to a woman with breast cancer. To determine if such procedures are justifiable alternatives to selective mastectomies and reconstructions, detailed, long-term statistical data on large numbers of cases must be available and end-results comparisons of various therapeutic modalities must be made on the basis of comparable statistical data.