This population-based study examined predictors of mastectomy for women with breast cancer in the greater western region of Sydney (GWRS) in New South Wales (NSW), Australia in 1992. Patients with a first diagnosis of breast cancer in 1992, the year prior to population-based mammographic screening in the region, were identified through the NSW state cancer registry. Data on stage, treatment, and demographic and health service characteristics were obtained from hospital records for patients treated within the region. The 282 patients who received surgical treatment for operable disease were considered in this analysis. Logistic regression was used to determine the odds ratio (OR) for mastectomy in relation to predictor variables with adjustment for confounding where appropriate. Age was included in all regressions even though it was not statistically significant. The tumor characteristics that were significant predictors (age adjusted) of mastectomy were stage [I (referent); IIA: OR = 1.5 (ns); IIB/IIIA-B: OR = 6.3, p < 0.05] and tumor size [T0-1 (referent); T2: OR = 1.8, p < 0.05; T3: OR = 12.9, p < 0.05]. There was a significant linear trend (p < 0.05) for lower ORs for mastectomy in women treated by surgeons with a higher breast cancer caseload (adjusted for age and stage). Women from municipalities with high socioeconomic status had lower mastectomy rates than others (OR = 0.5, p < 0.05), adjusted for age, stage, and surgeon activity level. Distance of residence from the main referral hospital and radiotherapy unit did not affect mastectomy rates. This study documents factors, in addition to stage at diagnosis, that play an important part in decisions about surgery for women with operable breast cancer. The experience of the surgeon, measured by the number of breast cancer patients, had an effect, as did the socioeconomic status of the woman. The latter may act through educational characteristics and participation of women in decisions concerning surgery.
- Journal Article