LAUSR

High quality advanced cancer care includes goals of care (GOC) discussions and should be tailored according to clinical diagnosis, patient characteristics, and in concordance with patient’s goals. Metastatic breast cancer (MBC) and treatment has heterogeneity according to subtype which makes the timing of initiating and continuing GOC discussions challenging. With an ever-increasing array of therapy, women with advanced stage disease are unique survivors in that they receive relatively aggressive cancer care to not only palliative symptoms but extend survival time. It is the purpose of this paper to explore the prevalence, pattern, and likelihood of having a GOC discussion according to key tumor, demographic, social, and clinical factors. A single-institution retrospective chart review of women deceased from MBC between November 2016 and November 2019 with double verification chart review was completed. Data were analyzed with descriptive, correlative, and comparative statistics. Total sample was N=167 women, with 14.4% (n=24) Black and 85.6% (n=143) White. Mean age was 55.3 years (SD 11.73). Overall, MBC survival was 3.12 years (SD 3.31), White women 41.2 (3.4 years) months, and Black women 19 (1.6 years) months. A total of 87.4% (n=146) participated in one or more GOC discussions. GOC discussions were more likely among those who were referred to palliative care (p<0.001) and social work (p=0.035) services. Of the GOC discussions, only 29.3% were conducted by the primary oncologist. Black women were more likely to have GOC discussion near death. Advanced stage cancers are treated, at times relatively aggressively, to extend survival time instead of merely offering palliation. This new paradigm of survivorship requires thoughtful integration of GOC conversations. Describing the current status of GOC discussions among a cohort of women deceased from MBC highlights the patients most vulnerable to having a GOC discussion avoided or delayed. These identified vulnerabilities will indicate where targeted interventions can be implemented in the future.