What influences clinicians' operative preferences for women with breast cancer? An application of the discrete choice experiment.

INTRODUCTION: Little is known regarding cancer clinicians' treatment preferences. AIM: Determine the impact of pre-operative variables over specialist breast clinicians' operative preferences using discrete choice experiment methodology. METHODS: Cross-sectional survey of operative preferences to hypothetical scenarios based on: patient age, bra cup size, cancer size, site and focality. RESULTS: 73% response rate (68/93). Multinomial logistic regression across scenarios (n=1695) with allowance for response clustering, comparing equal preference for mastectomy and breast conservation surgery (BCS) with preference for mastectomy or BCS. Increasing patient age, cancer size, central site, multi-focality and reducing cup size, all associated with preference for mastectomy, over equal preference, over BCS (p<0.001). Doctors preferred specific treatments, females and nurses avoided mastectomy (p=0.015 and p<0.001 respectively). CONCLUSIONS: Clinician preferences were predominantly treatment guideline congruent, but significantly influenced by patient age, clinician gender and occupation. This methodology is capable of elucidating treatment preferences and could be applied elsewhere where treatment options and practice variability exist.