A Pilot Randomized Clinical Trial of a Goals-of-Care Decision Aid for Surrogates of Severe Acute Brain Injury Patients

Abstract

Background and objectives: Breakdowns in clinician-family communication in neurologic intensive care units (neuroICUs) are common, particularly for goals-of-care decisions to continue or withdraw life-sustaining treatments while considering long-term prognoses. Shared decision-making interventions (decision aids [DAs]) may prevent this problem and increase patient-centered care, yet none are currently available. We assessed the feasibility, acceptability, and perceived usefulness of a DA for goals-of-care communication with surrogate decision-makers for critically ill severe acute brain injury (SABI) patients after hemispheric acute ischemic stroke, intracerebral hemorrhage, or traumatic brain injury.

Methods: We conducted a parallel-arm, unblinded, patient-level randomized, controlled pilot trial at two tertiary-care U.S. neuroICUs and randomized surrogate participants 1:1 to a tailored paper-based DA provided to surrogates prior to clinician-family goals-of-care meetings or usual care (no intervention prior to clinician-family meetings). The primary outcomes were feasibility of deploying the DA (recruitment, participation, retention), acceptability, and perceived usefulness of the DA among surrogates. Exploratory outcomes included outcome of surrogate goals-of-care decision, code-status changes during admission, patients' 3-month functional outcome, and surrogates' 3-month validated psychological outcomes.

Results: We approached 83 surrogates of 58 patients and enrolled 66 surrogates of 41 patients (80% consent rate). Of 66 surrogates, 45 remained in the study at 3 months (68% retention). Of the 33 surrogates randomized to intervention, 27 were able to receive the DA, and 25 subsequently read the DA (93% participation). 82% rated the DA's acceptability as good or excellent (median Acceptability score 2 [IQR 2;3]); 96% found it useful for goals-of-care decision-making. In the DA group, there was a trend towards fewer comfort-care decisions (27% vs. 56%, p=0.1) and fewer code-status changes (no change, 73% vs. 44%, p=0.02). At 3 months, fewer patients in the DA group had died (33% vs. 69%, p=0.05; median GOS 3 vs.1, p=0.05). Regardless of intervention, 3-month psychological outcomes were significantly worse among surrogates who had chosen continuation-of-care.

Discussion: A goals-of-care DA to support ICU shared decision-making for patients with SABI is feasible to deploy and well-perceived by surrogates. A larger trial is feasible to conduct, although surrogates who select continuation-of-care deserve additional psychosocial support.

Clinical trials registration: Clinicaltrials.gov NCT03833375 CLASSIFICATION OF EVIDENCE: This study provides Class IV evidence that the use of a DA explaining the goals-of-care decision and the treatment options is acceptable and useful to surrogates of incapacitated critically ill patients with ischemic stroke, intracerebral hemorrhage, or traumatic brain injury.

© 2022 American Academy of Neurology.

Figures

Figure 1. CONSORT Diagram

Shown are the number of screened, recruited, and analyzed study participants and reasons for exclusion.

Figure 2. Acceptability Data

(A) The multicolor bar represents a composite of 9 questions related to the acceptability of the content of the DA as rated by the surrogates randomized to intervention. (B) The light blue bars represent favorable responses related to additional specific questions about the DA. Favorable was defined as just right for the length of the DA and amount of information; balanced for the presentation of the decision options; yes on DA usefulness for making goals-of-care decisions; easy to understand the different potential outcomes on the way prognosis was explained; easier on how the tester thought the DA helped with making the decision; and yes on enough information for surrogates to decide on goals of care for the patient. DA = decision aid.