Evaluating if an Advance Care Planning Intervention Promotes Do-Not-Resuscitate Orders by Facilitating Accurate Prognostic Awareness

Background: Issuing do-not-resuscitate (DNR) orders has seldom been an outcome in randomized clinical trials of advance care planning (ACP) interventions. The aim of this study was to examine whether an ACP intervention facilitating accurate prognostic awareness (PA) for patients with advanced cancer was associated with earlier use of DNR orders. Patients and Methods: Participants (n=460) were randomly assigned 1:1 to the experimental and control arms, with 392 deceased participants constituting the final sample of this secondary analysis study. Participants in the intervention and control arms had each received an intervention tailored to their readiness for ACP/prognostic information and symptom-management education, respectively. Effectiveness in promoting a DNR order by facilitating accurate PA was determined by intention-to-treat analysis using multivariate logistic regression with hierarchical linear modeling. Results: At enrollment in the ACP intervention and before death, 9 (4.6%) and 8 (4.1%) participants and 168 (85.7%) and 164 (83.7%) participants in the experimental and control arms, respectively, had issued a DNR order, without significant between-arm differences. However, participants in the experimental arm with accurate PA were significantly more likely than participants in the control arm without accurate PA to have issued a DNR order before death (adjusted odds ratio, 2.264; 95% CI, 1.036-4.951; P=.041). Specifically, participants in the experimental arm who first reported accurate PA 31 to 90 days before death were significantly more likely than their counterparts in the control arm who reported accurate PA to have issued a DNR order in the next wave of assessment (adjusted odds ratio, 13.365; 95% CI, 1.989-89.786; P=.008). Both arms issued DNR orders close to death (median, 5-6 days before death). Conclusions: Our ACP intervention did not promote the overall presence of a DNR order. However, our intervention facilitated the issuance of NDR orders before death among patients with accurate PA, especially those who reported accurate PA 31 to 90 days before death, but it did not facilitate the issuance of DNR orders earlier than their counterparts in the control arm.