The Impact of Surgeons on the Likelihood of Mastectomy in Breast Cancer.

Abstract

This study evaluates the impact of individual surgeons and institutions on the use of mastectomy or breast conserving surgery (BCS) among elderly women with breast cancer. Current literature characterizes patient clinical and demographic factors that increase likelihood of mastectomy use. However, the impact of the individual provider or institution is not well understood, and could provide key insights to biases in the decision-making process. A retrospective cohort study of 29,358 women 65 years or older derived from the SEER-Medicare linked database with localized breast cancer diagnosed from 2000 to 2009. Multilevel, multivariable logistic models were employed, with odds ratios (ORs) used to describe the impact of demographic or clinical covariates, and the median OR (MOR) used to describe the relative impact of the surgeon and institution. Six thousand five hundred ninety-four women (22.4%) underwent mastectomy. Unadjusted rates of mastectomy ranged from 0% in the bottom quintile of surgeons to 58.0% in the top quintile. On multivariable analysis, the individual surgeon (MOR 1.97) had a greater impact on mastectomy than did the institution (MOR 1.71) or all other clinical and demographic variables except tumor size (OR 3.06) and nodal status (OR 2.95). Surgeons with more years in practice, or those with a lower case volume were more likely to perform mastectomy (P < 0.05). The individual surgeon influences the likelihood of mastectomy for the treatment of localized breast cancer. Further research should focus on physician-related biases that influence this decision to ensure patient autonomy.