To provide goal-concordant care means to provide medical care that honors a patient's individual goals and values, and to align medical treatments with those goals-of care (1, 2). Recent studies suggest that prior documentation of preferences to limit life-sustaining treatment may reduce the likelihood of being admitted to an ICU in the last 6 months of life (3). Studies of the past decade have also shown that such values and goals can be difficult to assess prior to an illness and to translate into relevant goals-of-care (4, 5).

When patients are admitted to the hospital with severe acute brain injury (SABI), which includes stroke, traumatic brain injury and hypoxic-ischemic encephalopathy after cardiac arrest, they typically lack decisional capacity and rarely have had their goals-of-care previously documented. Consequently, their family members or other surrogate decision-makers are tasked to work with clinicians to make treatment decisions based on the patients' presumed goals (6). Treatment decisions in the acute setting of SABI often concern high-stakes decisions around whether to focus medical care on survival, including the use of life-sustaining treatment (LST) such as mechanical ventilation, artificial nutrition or hydration, or to focus on comfort, which may mean limiting LST and allowing the patient to die a more natural death (7, 8). Given the substantial uncertainty regarding both the patients' prognosis and their presumed goals-of-care, LST is often administered as a time-limited trial in order to gain a better understanding of the patient's trajectory, prognosis, and likely goals over a defined period of time (9, 10). Consequently, at the end of a period of a time-limited trial, the continued use of LST has to be re-evaluated (11). Prognostic uncertainty typically persists for months after SABI and can challenge ongoing decisions in the acute care and post-acute care setting (12).

Using a cohort of patients with SABI, the objective of this study was therefore to assess (1) patients' presumed goals-of-care as assessed by family members in the acute setting; (2) the frequency of family-assessed goal-concordant care and the patient and family characteristics associated with family-assessed goal-concordant care; and (3) whether and how these goals have changed 6 months later.

Of the 222 patients enrolled in our SuPPOrTT study, families answered the goal concordance questions for 214 in the acute setting at a mean of 5.1 (SD 2.9) days after admission (Table 1). The majority of these 214 patients were non-Hispanic White (n = 147, 69%) and male (n = 118, 55%), with a mean age of 58 years (SD 18.9). Most patients had suffered a stroke (n = 129, 60%), with 30% suffering TBI (n = 65), and 9% HIE (n = 20). Family members included spouses (n = 66, 31%), adult children (n = 75, 35%), parents (n = 31, 14%) or siblings and others (n = 42, 19%), and a majority of them were white (n = 142, 66%) and female (n = 138, 64%) with a mean age of 50.9 years (SD 16.1).

In our cohort of 214 patients in the first week after severe acute brain injury (SABI), just over half of families felt that their loved one would prioritize extending life as much as possible. However, those family-assessed goals matched the family-reported care received for only 65% of patients. Life-sustaining treatment (LST) is often the default in hospital-level acute care unless a patient has specifically requested otherwise (13, 16). In the acute setting of SABI that is characterized by a high degree of prognostic uncertainty, national US guidelines even recommend “aggressive” therapy for those without advance directives to the contrary (17). Most families in our study report receiving care focused on extending life, regardless of goals, and goal concordance was accordingly more likely for patients who prioritized extending life. A presumed priority of comfort was more likely when patients were clinically sicker (by APACHE score), which may be related to their higher risk of mortality, although we do not know what information the family was given on the patient's chance of survival or recovery.

Our observation that non-Hispanic whites were more likely to prioritize comfort needs to be interpreted with caution as it was only significant for the comparison with “extending life” but not with “unsure.” We also did not collect detailed socio-economic characteristics, religious or cultural beliefs which might further confound this association. The observation is consistent with the literature that suggests higher prevalence of prioritizing aggressive care at the end-of-life in non-white compared to white patients and requires further research (18, 19).

One in four patients in our study may have been receiving unwanted aggressive care in the acute setting. It is possible that LST was provided with the mutual understanding of a time-limited trial, whereby family and clinicians have agreed on a period of aggressive interventions to see if the patient improves according to outcomes consistent with the patient's presumed goals-of-care (10, 20). In that case, this relatively high prevalence of aggressive care may be ethically justified as long as the prognostic uncertainty persists and as long as the time-limited trial is brought to a conclusion.

Future studies are needed to better understand the framework of goal concordance specifically in the setting of a time-limited trial after SABI. The possible associations of age, disease severity, and disease category also require further investigation. Goal concordance in older patients may be because they are more likely to have voiced their own goals prior to SABI, as described for the US population (21). Lower severity of SABI may account for more prognostic uncertainty leading to a trial of LST even in a patient who eventually might prioritize comfort. The trend toward a higher likelihood of goal concordance in patients with stroke could be related to the disease category itself, but also to the subspecialty of their medical providers (i.e., stroke neurologists vs. neurosurgeons or intensivists).

Six months after the acute event, a large proportion of family members were unsure of the patient's priorities. Given that SABI survivors are at high risk of re-hospitalization, these findings suggest important missed opportunities for improved communication between SABI survivors, their families and clinicians, even long after the event. Periodic re-evaluation of patient-centered goals and intentional conclusions to time-limited trials should be a routine part of post-SABI clinic visits (12, 22).

Our findings need to be considered in the setting of several important limitations. First, the single center design may limit generalizability of results. However, our center is the only comprehensive stroke and level 1 trauma center for a five-state region and is an academic county hospital serving wide variety of patients which may mitigate this limitation. Second, most of the patients were non-Hispanic white, and small numbers of non-white patients preclude analyses of separate minority races. Third, because patients were unable to communicate their own wishes or perspectives, we relied on families to provide substituted judgment. This does, however, reflect clinical practice where, if patients are unable to participate in decisions, goals-of-care are determined by family surrogates. Of note, only 22 (10%) of our patients had some type of pre-SABI advance directives documented in the EHR of which only half (n = 9) indicated any treatment preferences.

The observed high prevalence of patients potentially receiving unwanted aggressive care after SABI may be justified in the acute setting as long as prognostic uncertainty exists and provided it is in the context of a well-implemented time-limited trial. The high prevalence of families who are unsure of their loved one's goals of care 6 months after SABI suggest missed opportunities in communication between clinicians, families and patients as well as missed opportunities for completion of time-limited trials in the post-acute setting. More research is needed to better understand goal concordance in both the acute and post-acute care setting in the context of a time-limited trial of life-sustaining treatment.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation available to sincere requests.

The protocol was approved by the ethical review board of the University of Washington (STUDY 00003393). The ethics committee waived the requirement of written informed consent for participation.

RR and CC: obtained funding and drafting of the manuscript. CC, RJ, and JC: supervison. JB: statistical analysis. RR, JB, JC, and CC: acquisition, analysis, or interpretation of data. RR, JC, RJ, and CC: concept and design. All authors critically revised the manuscript for important intellectual content.

RR was supported by a grant of the Swiss National Science Foundation (P400PM_186732). CC received funding from a career development award from the National Institute of Neurological Disorders and Stroke (K23 NS099421). The researchers' funders had no role in the design, data analysis, interpretation or writing of the manuscript. All co-authors had full access to all the data in the study and had final responsibility for the decision to submit the publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer, KD, declared a shared affiliation, though no collaboration, with three of the authors, RR, GB, and RJ to the handling editor.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.