Effects of a Communication Intervention Randomized Controlled Trial to Enable Goals-of-Care Discussions

Abstract

PURPOSE: Patients with advanced cancer often have a poor understanding of cancer incurability, which correlates with more aggressive treatment near the end of life (EOL). We sought to determine whether training oncologists to elicit patient values for goals-of-care (GoC) discussions will increase and improve these discussions. We explored its impact on use of aggressive care at EOL. METHODS: We enrolled and used block randomization to assign 92% of solid tumor oncologists to 2-hour communication skills training and four coaching sessions. We surveyed 265 patient with newly diagnosed advanced cancer with < 2-year life expectancy at baseline and 6 months. We assessed prevalence and quality of GoC communication, change in communication skills, and use of aggressive care in the last month of life. RESULTS: Intervention (INT) oncologists’ (n = 11) skill to elicit patient values increased (27%-55%), while usual care (UC) oncologists’ (n = 11) skill did not (9%-0%; P = .01). Forty-eight percent (n = 74) INT v 51% (n = 56) UC patients reported a GoC discussion ( P = .61). There was no difference in the prevalence or quality of GoC communication between groups (global odds ratio, 0.84; 95% CI, 0.57 to 1.23). Within 6 months, there was no difference in deaths (18 INT v 16 UC; P = .51), mean hospitalizations (0.47 INT v 0.42 UC; P = .63), intensive care unit admissions (5% INT v 9% UC; P = .65), or chemotherapy (26% INT v 16% UC; P = .39). CONCLUSION: Use of a coaching model focused on teaching oncologists to elicit patient values improved that skill but did not increase prevalence or quality of GoC discussions among patients with advanced cancer. There was no impact on high care utilization at EOL.