Randomized Controlled Trial of a Decision Support Intervention About Cardiopulmonary Resuscitation for Hospitalized Patients Who Have a High Risk of Death.

Abstract

Many seriously ill hospitalized patients have cardiopulmonary resuscitation (CPR) as part of their care plan, but CPR is unlikely to achieve the goals of many seriously ill hospitalized patients. To determine if a multicomponent decision support intervention changes documented orders for CPR in the medical record, compared to usual care. Open-label randomized controlled trial. Patients on internal medicine and neurology wards at two tertiary care teaching hospitals who had a 1-year mortality greater than 10% as predicted with a validated model and whose care plan included CPR, if needed. Both the control and intervention groups received usual communication about CPR at the discretion of their care team. The intervention group participated in a values clarification exercise and watched a CPR video decision aid. The primary outcome was the proportion of patients who had a no-CPR order at 14days after enrollment. We recruited 200 patients between October 2017 and October 2018. Mean age was 77years. There was no difference between the groups in no-CPR orders 14days after enrollment (17/100 (17%) intervention vs 17/99 (17%) control, risk difference, - 0.2%) (95% confidence interval - 11 to 10%; p= 0.98). In addition, there were no differences between groups in decisional conflict summary score or satisfaction with decision-making. Patients in the intervention group had less conflict about understanding treatment options (decisional conflict knowledge subscale score mean (SD), 17.5 (26.5) intervention arm vs 40.4 (38.1) control; scale range 0-100 with lower scores reflecting less conflict). Among seriously ill hospitalized patients who had CPR as part of their care plan, this decision support intervention did not increase the likelihood of no-CPR orders compared to usual care. Canadian Frailty Network, The Ottawa Hospital Academic Medical Organization.