Common Data Elements for Disorders of Consciousness: Recommendations from the Working Group on Goals-of-care and Family/Surrogate Decision-Maker Data

Introduction: In order to facilitate comparative research, it is essential for the fields of neurocritical care and rehabilitation to establish common data elements (CDE) for disorders of consciousness (DoC). Our objective was to identify CDEs related to goals-of-care decisions and family/surrogate decision-making for patients with DoC. Methods: To achieve this, we formed nine CDE working groups as part of the Neurocritical Care Society’s Curing Coma Campaign. Our working group focused on goals-of-care decisions and family/surrogate decision-makers created five subgroups: (1) clinical variables of surrogates, (2) psychological distress of surrogates, (3) decision-making quality, (4) quality of communication, and (5) quality of end-of-life care. Each subgroup searched for existing relevant CDEs in the NIH/CDE catalog and conducted an extensive literature search for additional relevant study instruments to be recommended. We classified each CDE according to the standard definitions of “core,” “basic,” “exploratory,” or “supplemental,” as well as their utility for studying the acute or chronic phase of DoC, or both. Results: We identified 32 relevant pre-existing NIH CDEs across all subgroups. A total of 34 new instruments were added across all subgroups. Only one CDE was recommended as disease core, the “mode of death” of the patient from the clinical variables subgroup. Conclusions: Our findings provide valuable CDEs specific to goals-of-care decisions and family/surrogate decision-making for patients with DoC that can be used to standardize studies to generate high-quality and reproducible research in this area.


Introduction
In 2019, the Neurocritical Care Society launched the Curing Coma Campaign to bring together clinician and nonclinician scientists in a common scienti c goal: to identify the gaps in our understanding of disorders of consciousness (DoC) and develop new research across the full spectrum of care. 1 Three pillars were identi ed as the fundamental components needed to bring the science and care of patients with DoC forward. These pillars are (1) endotyping to better characterize coma and its clinical trajectory, (2) discovering biomarkers of coma and coma recovery, and (3) conducting proof-of-concept clinical trials to inform and advance the design of future trials. From this work, it became clear that the development of well-de ned common data elements (CDE) would be paramount to the success of future scienti c work in this area. 2 CDEs are standardized data points and instruments that enhance reproducibility and data comparability for research trials to facilitate disease-speci c scienti c communication and progress 3 . The National Institute of Neurological Disorders and Stroke (NINDS) has led the development of CDEs for multiple neurological diseases and disorders since 2005 with the goal of improving data consistency among NINDS-funded projects.
For the Curing Coma Campaign's DoC CDE project, nine working groups were created, including one focused goals-of-care, family/surrogate decision-maker data (F/SDM) and related outcomes. This focus is novel for a CDE project, as there are no prior CDE projects addressing data elements related to surrogate medical decision making, or surrogate psychological outcomes or well-being. With increasing emphasis in critical care on the best ways to communicate with surrogates of patients with severe acute brain injury and facilitate high-quality decision-making, it is imperative that the neurocritical care eld develop consistent and uniform instruments that may facilitate research on these aspects of clinical care. 4,5 The goal of this working group was to develop CDEs for goals-of-care and F/SDM data. CDEs like those proposed by our working group will enable the community to compare family and surrogate interventions, as we as a medical community continue to explore this understudied aspect of neurocritical care.

Methods
The Steering Committee for the DoC CDE project identi ed leaders with expertise in each of the identi ed task groups: clinical/behavioral phenotyping, hospital course/confounders, imaging, electrophysiology, biospecimens, physiologic data ("big data"), therapeutic interventions, outcomes, and endpoints, and goals-ofcare and F/SDM data. The chair of each working group subsequently identi ed and invited members from across the Neurocritical Care Society, including trainees in fellowship, junior faculty members, non-physician clinicians (e.g. nurses, pharmacists, advanced practice providers), senior physician content experts, and at least one family advocate. Each working group then drafted a charter to describe its mission and scope, including the process overview, reporting plan, and deliverables.
Work groups were tasked with identifying and/or developing CDEs applicable to their mission and scope. Our group's work process is described in Fig. 1. First, the working group performed an in-depth review of the existing CDEs published at https://www.commondataelements.ninds.nih.gov for other neurological diseases pertinent to the overall scope of goals-of-care decisions and the impact of decision-making on F/SDM of patients with DoC.
Early meetings of the working group focused on the development of several key domains important to the care of surrogate decision makers of patients with DoC. These domains included (1) clinical variables of F/SDM (i.e., family/surrogates' relevant demographic information), (2) psychological well-being of F/SDM, (3) decision making quality, (4) quality of communication, and (5) quality of end-of-life care. For each subgroup, we identi ed 2-3 co-authors who were tasked with conducting a literature review to identify any known study instruments relevant to their domain. Subgroups also determined through a literature search whether or not instruments had been previously validated in populations of patients with DoC or other relevant neurological disorders. All questionnaires requiring patient participation were excluded, as the working group was intended to focus on goals-of-care and F/SDM outcomes and as this patient population is typically not able to participate in surveys and similar instruments due to their DoC. Following review by individual subgroups, the full working group met to discuss each CDE and scale to determine its relevance to DoC and the working group's scope. Over several meetings, the list of potential relevant CDEs was narrowed, and subsequent meetings focused on classifying each study instrument as described below.
To maintain consistency with previously published CDEs, each working group was responsible for classifying each of the recommended CDEs and instruments as "Core", "Basic," "Supplemental," or "Exploratory," according to the following standard de nitions: 6 1. Core CDEs-elements which should be consistently collected across all studies of DoC and which are required for collection in all studies of DoC regardless of neurological disease and study focus; 2. Basic CDEs-elements that are essential based on certain conditions or study focus and that are strongly recommended for studies related to these conditions or study foci in DoC; 3. Supplemental CDEs-elements that are commonly recommended for speci c DoC studies, depending on the context and goals of the speci c study; 4. Exploratory CDEs-elements which are reasonable to use, but whose validity is limited due to insu cient availability of data; thus, they require further validation.
We further classi ed CDEs according to their utility in the acute phase of DoC (the initial admission for the inciting neurological event through hospital discharge), the chronic phase (post-discharge DoC, including patients in rehabilitation centers, nursing facilities, or home), or both. A nal list of CDEs was submitted to the Steering Committee. Work group leaders met to nominate and agree upon the Disease Core CDEs. Case report forms (CRF) for each CDE were developed by each working group and made available for public comment prior to nalizing the CRFs with inclusion of public feedback. All nalized CRFs are available to the scienti c community focused on studying patients with DoC at https://zenodo.org/record/7210236#.ZCXwb-zML0p.

Results
Our working group identi ed ve domains.

Clinical variables
We identi ed 22 relevant pre-existing NIH CDEs for this subgroup and ten additional items through literature review (Table 1). We also identi ed "mode of death" (death from withdrawal of life sustaining treatments or brain death or respiratory/cardiac arrest) of the patient as an important clinical variable related to goals-of-care, as surrogates decide on withdrawal of life sustaining treatments during their goals-of-care decision-making process. "Mode of death" of the patient/subject, was identi ed as a Disease Core CDE. An additional ve elements were identi ed as basic; the remainder were supplemental and exploratory.   We identi ed 10 relevant pre-existing NIH CDEs (Table 5). All were deemed supplemental. Two CDEs (the Center for Epidemiologic Studies Depression Scale (CEDS) and Zarit Caregiver Burden Scale) apply only to chronic DoC; we recommend the remainder for use in both acute and chronic DoC. Eight additional instruments were added in this category, all of which are exploratory (Table 5)  In neurocritical care speci cally, shared decision-making tools are being explored to address the challenges of decision-making in the setting of uncertain prognosis and to help families arrive at patient goal-concordant decisions in which they feel con dent and reduce decisional con ict and decision regret [86][87][88][89][90].
There are no speci c surveys or other instruments currently available that have been speci cally developed for evaluation of F/SDM outcomes following surrogate decision-making for patients with DoC. The inconsistent use of speci c instruments to evaluate the experiences of F/SDM is among the challenges in establishing e cacy and reproducibility of results pertaining to decision tools, family interventions, and serious illness communication techniques. While limited by the lack of extensive work in our speci c disease population, our group has collated instruments that are well-validated in comparable populations. The consistent, widespread use of these instruments could promote careful study of communication, decision-making, and end of life management in our study population, and thus improve the care of patients with DoC and their caregivers. This CDE for goals-of-care and F/SDM is among the rst necessary steps to harmonize all future research in this special population of caregivers and decision-makers for patients with DoC. Additional next steps include the validation of high-value study instruments speci cally in F/SDM of patients with DoC, as well as adapting extant instruments to better address concerns and outcomes pertinent to this population.

Declarations
Authorship Con rmation and Compliance with Journal Policies: This manuscript has not been published or under consideration elsewhere. This manuscript complies with all instructions to authors. All authors have met requirements for authorship, as outlined below, and approved the nal manuscript.
Author Contributions: MNJ, HLK and SM wrote the initial draft of the manuscript. All co-authors edited the manuscript and approved the nal content. All co-authors contributed equally to the Case Report Forms released with the manuscript.