Trade-offs between overall survival and side effects in the treatment of metastatic breast cancer: eliciting preferences of patients with primary and metastatic breast cancer using a discrete choice experiment

Abstract

Objectives There has been a recent proliferation in treatment options for patients with metastatic breast cancer. Such treatments often involve trade-offs between overall survival and side effects. Our study aims to estimate the trade-offs that could be used to inform decision-making at the individual and policy level.Design We designed a discrete choice experiment (DCE) to look at preferences for avoiding severity levels of side effects when choosing treatment for metastatic breast cancer. Treatment attributes were: fatigue, nausea, diarrhoea, other side effects (peripheral neuropathy, hand–foot syndrome and mucositis) and urgent hospital admission and overall survival. Responses were analysed using an error component logit model. We estimated the relative importance of attributes and minimum acceptable survival for improvements in side effects.Setting The DCE was completed online by UK residents with self-reported diagnoses of breast cancer.Participants 105 respondents participated, of which 72 patients had metastatic breast cancer and 33 patients had primary breast cancer.Results Overall survival had the largest relative importance, followed by other side effects, diarrhoea, nausea and fatigue. The risk of urgent hospital admission was not significant. While overall survival was the most important attribute, respondents were willing to forgo some absolute probability of overall survival for reductions in all Grade 2 side effects (12.02% for hand–foot syndrome, 11.01% for mucositis, 10.42% for peripheral neuropathy, 6.33% for diarrhoea and 3.62% for nausea). Grade 1 side effects were not significant, suggesting respondents have a general tolerance for them.Conclusions Patients are willing to forgo overall survival to avoid particular severity levels of side effects. Our results have implications for data collected in research studies and can help inform person-centred care and shared decision-making.