Surrogate Perceptions of Disability After Hospitalization for Traumatic Brain Injury

Abstract

Background:

The Glasgow Outcome Scale Extended (GOSE) is a measure of recovery after traumatic brain injury (TBI). Public surveys rate some GOSE states worse than death. Direct family experience caring for TBI patients may impact views of post-TBI disability.

Study Design:

We conducted a national cross-sectional computer-adaptive survey of surrogates of TBI dependents incurring injury >1y prior. Using a standard gamble approach in randomized order, surrogates evaluated preferences for post-TBI GOSE states from GOSE2 (bedridden, unaware) to GOSE8 (good recovery). We calculated median [IQR] health utilities for each post-TBI state, ranging from −1 to 1, with 0 as reference (death=GOSE1), and assessed sociodemographic associations using proportional odds logistic regression modeling.

Results:

Of 515 eligible surrogates, 298 (58%) completed scenarios. Surrogates were median aged 46 [IQR 35,60], 54% married, with Santa Clara strength of faith 14 [10,18]. TBI dependents had a median GOSE5 [3,7]. Median [IQR] health utility ratings for GOSE2, GOSE3, and GOSE4 were −0.06 [−0.50,−0.01], −0.01 [−0.30,0.45], and 0.30 [−0.01,0.80], rated worse than death by 91%, 65%, and 40%, respectively. Surrogates rated GOSE4 (daily partial help) worse than the general population. Married surrogates rated GOSE4 higher (p<0.01). Higher strength of faith was associated with higher utility scores across GOSE states (p=0.034).

Conclusion:

In this index study of surrogate perceptions about disability after TBI, poor neurologic outcomes – vegetative, needing all-day or partial daily assistance – were perceived as worse than death by at least one in three surrogates. Surrogate perceptions differed from the unexposed public. Long-term perceptions about post-TBI disability may inform earlier, tailored shared decision-making after neurotrauma.

Précis:

In this index study of surrogate perceptions about disability after TBI, poor neurologic outcomes – vegetative, needing all-day or partial daily assistance – were perceived as worse than death by at least one in three surrogates. Surrogate perceptions differed from public perceptions unexposed to hospitalized TBI.

Introduction

Traumatic brain injury (TBI) is a major cause of death and disability in the United States and beyond.^1,2 The short- and long-term socioeconomic consequences of this “silent epidemic” are considerable.³ TBI survivors face prolonged periods of treatment in intensive care units with associated physiological impact, followed by months or years of rehabilitation and dependency. In the long term, traumatic encephalopathy and/or Alzheimer’s disease may occur as sequelae of TBI.^4,5

Long-term recovery after hospitalized TBI is highly variable and challenging to predict at the time of injury with existing calculators.^6–9 Post-TBI disability is commonly assessed with the Glasgow Outcome Scale Extended (GOSE), a validated clinical trial endpoint of recovery after TBI.^10–14 The eight-tiered GOSE scale describes outcomes based on mortality, level of consciousness, work status, and return to independence.^15,16 While validated and widely used, the GOSE scale is also nonlinear and nonparametric, meaning it can be hard to interpret and communicate to patients and families alike.

Health utilities can define the quality-of-life value judgements by evaluating preferences for specific health-related outcomes (in this case, GOSE states). To date, no quality-of-life health utilities exist for disability outcomes after actual TBI. Public surveys rate some GOSE states as worse than death.¹⁷ However, a major limitation of this prior work is the hypothetical nature of the post-TBI recovery states described.¹⁸ In contrast, direct family experience caring for TBI patients may impact views of post-TBI disability. For example, individuals with close exposure to loved ones with hospitalized TBI may rate their quality of life higher than expected, as has been shown for survivors of spinal cord injury.^19–21 We aimed to describe factors predicting post-TBI health utilities for disability states after surrogate exposure to hospitalized TBI. We hypothesized TBI surrogates would interpret post-TBI GOSE health states as nonlinear and unequal, and that some post-TBI disability states would be perceived to be worse than death. We further hypothesized that TBI surrogates would view certain post-TBI disability states differently than does the general population.

Methods

Study Population

This national cross-sectional study recruited adult surrogates of patients with a TBI that had occurred at least a year before. For the purposes of this study, we defined surrogates as caregivers and/or decision-makers in charge of daily care and medical decisions, who received regular updates from the medical team, and often helped with activities of daily living for the TBI survivor. Surrogates were recruited over 90 days in 2021 via multiple methods: formal email recruitment, social media advertisement, and ResearchMatch, which is a national web-based recruitment tool designed by the Vanderbilt University Clinical and Translational Science Awards and maintained as an Institutional Review Board (IRB)-approved data repository.

We used a previously validated GOSE Scenario Survey to evaluate perceptions of post-TBI disability, similar to our prior work surveying the general population prompted to consider hypothetical head injury scenarios.¹⁷ Once they accessed the online GOSE Scenario Survey, participants provided informed consent prior to answering any research questions. Participants were incentivized with lottery-drawn Amazon gift cards in the following amounts: one (1) $1,000, two (2) $250, three (3) $100, and four (4) $50. Registration of internet protocol addresses were used to ensure no duplicate completion of the survey, and cookies were not used. The Vanderbilt University IRB approved this study.

Study Procedures: GOSE Scenario Survey

We used the electronic Qualtrics survey platform (Qualtrics, Provo, UT) to develop the structured GOSE Scenario Survey, similar to our prior work.¹⁷ In the development of the survey and the reporting of our Methods and Results, we adhered to the EQUATOR (Enhancing Quality and Transparency of Health Research) guidelines specific to web-based surveys, the Checklist for Reporting Results of Internet E-Surveys (CHERRIES).²² The structured GOSE Scenario Survey²³ describes seven GOSE health states ranging GOSE 2 (bedridden and unaware, vegetative state) to GOSE 8 (upper good recovery), by outlining the defining features for each state that determine an individual’s level of disability and functional status. GOSE 1, which is death, was used as a comparator in the scenarios and thus was not included as a separate health state for participants to evaluate.

TBI surrogates were exposed to all seven of the GOSE health states (2, 3, 4, 5, 6, 7, and 8) sequentially in a random order and values were determined for each post-TBI categorical health state. These values were in the form of health utilities, which are numerical values quantifying the quality of life of a health state, classically ranging from 0 to 1.²⁴ We used the standard gamble approach to measure quality of life preferences via health utilities, which is the most commonly employed methodology.²⁵ This approach works by including a “risk element”, which simulates the uncertainty that is part of medical decision-making.²⁴ Supplemental Figure 1 includes screenshots of the GOSE Scenario Survey demonstrating how this standard gamble approach plays out. Surrogates were specifically asked to answer the questions from their own perspective (i.e., not from the perspective of the TBI survivor). Most screens had one question per page, and survey length and duration varied depending on specific participant responses.

As the Supplemental Figure 1 shows, for each scenario, participants were asked to make a choice between (A) the certainty of continued life in the given health state (e.g., GOSE 4), and (B) taking a gamble with two possible outcomes, perfect health versus immediate death. These gamble probabilities were initially set at 50% odds of immediate death and 50% odds of perfect health and were sequentially adjusted based on the participants’ responses until the point of indifference was reached between the two choices offered. The probability of perfect health at that point of indifference then corresponds to the respondent’s health utility for that specific GOSE state.

We intentionally allowed for negative health utility values based on our prior work demonstrating that certain health states are frequently considered worse than death (e.g., GOSE 2, vegetative state).¹⁷ Perfect health corresponds to a health utility score of 1, and death is fixed at a health utility score of 0. Health states with health utilities approximating 1, therefore, are perceived as close to full health. On the other hand, states with health utilities less than 0 are viewed as worse than death.

Participants were unable to “go back” and change their responses to a previous scenario once they advanced to the next scenario. We designed the GOSE health states to be presented in random order in order to control for any carryover effects, and we included the random scenario order as a model covariate. We excluded participant responses if they were incomplete, if all GOSE health states were valued as the same, or if there were at least two logical inconsistencies in quality of life valuations of GOSE states 2 through 8 (excluding GOSE 1, death), similar to in our prior work.¹⁷ We defined a ‘logical inconsistency’ as valuing a more severe disability state higher than a less severe disability state, such as ranking GOSE 4 (needing daily partial help) above GOSE 6 (can partially resume work).

Study Procedures: Demographics and TBI Exposures

After completing the GOSE Scenario Survey for all GOSE states 2–8, participants answered a series of demographic questions about themselves, including age, sex, ethnicity, race, education level, marital status, geographic region, employment status, religious strength of faith as measured by the validated Abbreviated Santa Clara Strength of Religious Faith Questionnaire,²⁶ current health status as measured by the EQ-5D-3L and the EQ-5D Visual Analogue Scale,^27,28 and prior exposure to traumatic life events using the validated Brief Trauma Questionnaire (BTQ).²⁹ By design, we asked these demographic questions after the GOSE Scenario Survey in order to minimize stereotype threat.³⁰

Furthermore, we collected data on the degree of participants’ prior exposure to TBI. Surrogates were asked to define their loved one’s current post-TBI categorical GOSE state, by outlining the defining features for each state that determine an individual’s level of disability and functional status. Using non-medical terminology, participants were also specifically asked whether they witnessed their loved one going through the following measures of TBI disease severity: craniectomy or craniotomy, intracranial pressure (ICP) monitor placement, admission to the intensive care unit, ventilator support, tracheostomy, feeding tube placement (excluding temporary nasoenteric tubes), and discharge to a long-term acute care hospital, skilled nursing facility, or inpatient rehabilitation center.

Statistical Analysis

We calculated median health utilities [interquartile range, IQR] for each GOSE state, with values ranging from −1 (worse than death) to 1 (full health), with 0 set as reference (death, GOSE 1). GOSE health utilities as determined by the surrogates in this study were then visually compared with those from our prior survey in the general population naïve to hospitalized TBI.¹⁷ We assessed the association between risk factors and the GOSE utility scores using a proportional odds logistic regression model with GOSE utility scores as the dependent variable (outcome). Independent variables were the GOSE states, respondent age, sex, marital status, geographic region, years of education (modeled continuously), EQ-5D-3L index score, religious strength of faith, number of past traumatic life events, history of exposure to tracheostomy and feeding tube placement for their loved one with TBI, and an integer representing the random order in which the GOSE Scenario was presented to the participant.

Religious strength of faith was determined according to the Santa Clara scale (scale: 5–20, with higher equating to more strength of faith). The number of past traumatic events was assessed using the Brief Trauma Questionnaire (BTQ), ranging from 0–10. Interactions between GOSE states and all other factors were included in the model. All statistical analyses were done using R version 4.3.1.

Results

Demographics and TBI Exposures

Of 515 eligible surrogates, 298 (58%) consented and completed scenarios. Figure 1 contains a flow chart of the breakdown of individuals who accessed the survey, were determined to meet a priori eligibility criteria, consented to proceed, and then completed the GOSE Scenario Survey.

Figure 1. Population Flow Chart.

Figure 1 shows the flow chart of potential participants who accessed and were assessed for eligibility for the GOSE Scenario Survey to assess perceptions of post-traumatic brain injury (TBI) disability states.

Table 1 demonstrates the surrogate respondent characteristics. Surrogates were a median aged 46 years [IQR 35, 60] and 78% female. Slightly more than half were married. They had a median 15 [14,18] years education and 30% worked in professional employment. Surrogates were most commonly from the South (48%) and the remaining 52% were from the Northeast, Midwest, and West regions. They had a median Santa Clara strength of faith scale of 14 [IQR 10, 18] and an overall low exposure to traumatic life events as measured by the Brief Trauma Questionnaire.

Table 1.

Respondent Characteristics

Characteristic	Median [IQR] or n (%)
Age	46 [35, 60]
Male sex	64 (22%)
Hispanic ethnicity	17 (6%)
Caucasian	255 (86%)
Black	11 (4%)
Asian	10 (3%)
Other	17 (6%)
Education in years	15 [14, 18]
Married	160 (54%)
Divorced or separated	42 (14%)
Widow(er)	13 (4%)
Never been married	79 (27%)
Professional employment	89 (30%)
Skilled, non-manual employment	38 (13%)
Skilled, manual employment	34 (11%)
Manual employment	15 (5%)
Unemployed or not applicable	119 (40%)
Northeast region	22 (7%)
Midwest region	53 (18%)
South region	143 (48%)
West region	49 (16%)
EQ-5D visual analogue scale (0–100)	80 [65, 90]
Santa Clara strength of faith scale (5–20)	14 [10, 18]
Brief Trauma Questionnaire scale (0–10)	2 [0, 3]

TBI, traumatic brain injury

GOSE, Extended Glasgow Outcome Scale

EQ-5D, five-dimension scale measuring quality of life, range 0–100

IQR, interquartile range

Table 1 includes demographic details and other characteristics of TBI surrogates who completed the GOSE scenario surveys. EQ-5D is a five-dimension scale measuring quality of life and ranges from 0 (worst) to 100 (best). The Santa Clara scale of strength of faith measures religious strength of faith and ranges from 5 (low) to 20 (high). The Brief Trauma Questionnaire quantifies the number of past traumatic lifetime events and ranges from 0 (none) to 10 (high).

Surrogates rated their own quality of life at a median 80 [IQR 65, 90] on a 0–100 scale. TBI survivors had a current median GOSE 5 [3, 7], and respondents judged the survivors’ current quality of life at a median 60 [IQR 30.3, 78.0]. Regarding past exposure to severity of TBI treatments, 53% had family members who underwent craniectomy or craniotomy, 48% were intubated, 41% had an ICP monitor, 35% had a feeding tube, and 24% had a tracheostomy. Almost three quarters (74%) had been admitted to an intensive care unit, and the majority were discharged to a facility, either inpatient rehabilitation (56%) or a higher-level skilled facility (30%).

GOSE Health Utility Ratings

In general, TBI surrogates rated lower GOSE states (i.e., those corresponding to more severe post-TBI disability) as having lower utility, although the relationship was non-linear and intervals were unequal between states (Table 1). GOSE 7 and GOSE 8 were both rated highly, but there was a notable drop-off in healthy utility ratings from GOSE 6 to GOSE 5 and then again to GOSE 4. Median [IQR] health utility ratings for GOSE 2, GOSE 3, and GOSE 4 were −0.06 [−0.50, −0.01], −0.01 [−0.30, 0.45], and 0.30 [−0.01, 0.80], and these states were rated worse than death by 91%, 65%, and 40% of respondents, respectively. Surrogates rated GOSE 4 (daily partial help) worse than did the general population unexposed to hospitalized TBI, represented in data available from our prior publication and reproduced as part of Figure 2.¹⁷ Visually, GOSE 4 health utilities from the TBI surrogates had a bimodal distribution compared to the GOSE 4 health utilities from the general population (Figure 2B).

Figure 2. Extended Box Plots of Utility Scores after TBI across GOSE levels.

A. The extended box plots display the unadjusted health utility valuations by Glasgow Outcome Scale Extended (GOSE) scores 2 through 8 after being presented in random order and rated by 298 traumatic brain injury (TBI) surrogates across the United States (green boxes) versus our previous survey in the general public (orange boxes). The vertical dashed red line represents death, or GOSE 1, which served as the reference state for evaluating the other GOSE health states. A quality of life of 1 is interpreted as perfect health and a quality of life of 0 is interpreted as death. Health states rated as worse than death by TBI surrogates were assigned a negative value. The center black dot depicts the mean health utility score for that GOSE state, and the centermost box covers the interquartile range for health utility scores. GOSE 4 (needing daily partial help) was rated significantly worse by TBI surrogates than by the general public. GOSE 2 (i.e., vegetative condition after TBI, bed-ridden and unable to communicate) was rated as worse than death by the mean, median, and 75th percentile of the TBI surrogates.

B. Comparison of GOSE 4 health utility scores between general public and TBI surrogates.

Predictors of GOSE Health Utility Ratings

Demographics and other characteristics that were predictive of respondents rating high mean GOSE utility score are shown in Figure 3. As expected, higher GOSE states were very likely to be given higher health utility scores, meaning that, for example, surrogates were likely to rate GOSE 5 higher than GOSE 4 and GOSE 4 higher than GOSE 3 (p<0.0001 for the combined GOSE main effect and interaction effects with other variables).

Figure 3. Effect Plot of Factors associated with GOSE Health Utility scores among TBI Surrogates.

Figure 3 shows the association of demographic, TBI exposure, and other factors with GOSE health utility scores of TBI surrogates. “GOSE” refers to the GOSE level queried in the survey scenario (i.e., GOSE 7 was more likely to be rated higher than GOSE 6). “Actual GOSE” refers to the TBI family member’s current GOSE functional level as reported by the surrogate. The Santa Clara scale of strength of faith measures religious strength of faith, ranging from 5 (low) to 20 (high).

Marital status did have a significant impact on health utility ratings. Married surrogates rated GOSE 4 higher (p<0.01). Higher strength of faith was associated with higher utility scores across GOSE states (p=0.03). There was also variation by geographical area in the predicted health utilities. Compared to the Northeast region, the other regions (Midwest, South, and West) had higher expected health utilities. Surrogate age, sex, education level, current quality of life, and past traumatic experiences did not demonstrate any significant association with GOSE health utility ratings. The randomized order in which the GOSE scenarios were presented to the respondents did not affect the GOSE utility score (p=0.7461 for the combined order main effect and interaction effect).

Discussion

In this study of nearly 300 surrogates of patients with hospitalized TBI, 91% judged being in a vegetative state (GOSE 2) as worse than death. Furthermore, 65% judged being housebound and dependent on all-day assistance (GOSE 3) as worse than death, and 40% judged being partially dependent (GOSE 4) as worse than death. The relationship between different GOSE states valuations was non-linear and intervals were unequal. While TBI surrogates viewed post-TBI disability states similar to the general population in many ways, GOSE 4 (needing daily partial help) was rated significantly worse by TBI surrogates than by the general public in our prior study.¹⁷

We found marital status, strength of faith, and geographic region to be significant predictors of GOSE health utility ratings in this population of TBI surrogates. In our prior investigation of public perceptions of hypothetical post-TBI disability states, marital status was similarly demonstrated to be an important predictor of health utility ratings.¹⁷ Surrogate marital status is one of many factors thought to impact variation in medical decision making, processes of care, and outcomes across centers.^31,32

We asked about strength of faith in this study because this factor is often volunteered by family members during goals of care and prognostic discussions in the acute phase of TBI care. Indeed, higher self-reported strength of faith on the validated Santa Clara scale was associated with higher utility ratings across all disability states. This merits future study. In patients with cancer, higher religious strength of faith has been associated with more intensive medical care; data are not available for quality of life valuations in trauma patients.³³

Regional variation has also been demonstrated in end-of-life care for patients with cancer and other chronic conditions; again, no specific data are available for regional preference variations in the setting of acute trauma.^34,35 There was limited evidence for an association with other nuances in demographics and TBI exposures, including variability in exposure to tracheostomy and surgical feeding tube placement, potentially due to small sample size in this study.

The drop-off in perceived utility ratings from GOSE 5 (unemployed, independent) to GOSE 4 (needing daily partial assistance) merits particular note. Prior TBI studies have overwhelmingly used the GOSE endpoint in a dichotomized form as a primary study endpoint, with favorable functional status corresponding to a GOSE of 5 or higher (5–8) and unfavorable functional status to a GOSE 4 or less (1–4).^36,37 This dichotomized GOSE metric is the TBI clinical trial endpoint recommended by the National Institutes of Health and the Food and Drug Administration,^38–40 despite a paucity of data demonstrating this as a validated cut-off. Of the experts who participated in the recent Seattle International Severe traumatic Brain Injury Consensus Conference (SIBICC), more than 80% judged certainty of a persistent vegetative state (GOSE 2) to be grounds for withdrawal of life-sustaining care and more than 60% said the same for lower severe disability (GOSE 3).⁷ However, only 15% of the provider experts felt that upper severe disability (GOSE 4) justified withdrawal of life-sustaining care, an interesting contrast to the 40% of surrogates in the present study who rated GOSE 4 as worse than death.

The current study does provide more patient-centered data that GOSE 4 is indeed an unfavorable outcome through the lens of a large proportion of TBI surrogates who have direct experience with a loved one after a hospitalized TBI, reinforcing current practices of dichotomizing the GOSE scale at 4 and below. Unfortunately, at present, GOSE predictions are often inaccurate and never definitive for such nuances in the acute setting (e.g., predicting GOSE 4 versus GOSE 5). The vast majority of the same SIBICC group cited above rarely or never use prognostic calculators for severe TBI patients, because the existing calculators are not useful for individual patient prognostication.⁷ A better understanding of actual stakeholders’ views of the varied and often unpredictable long-term outcomes after TBI is also critical for better shared clinical decision-making in the acute setting. Prognostication and shared decision-making with surrogates in neurocritical care settings (including, but not limited to, TBI) is an area of active debate and research, with a growing understanding of the nuances and caveats of surrogate decision-making.^18,41–44

To our knowledge, this study is the first attempt to assess TBI surrogates’ views of health utilities of disability states after hospitalized TBI. Nevertheless, our study does have several limitations, including selection bias, recruitment bias, and survivorship bias. Relatively small sample size may have resulted in type 2 errors. While we sought to collect data on respondents’ exposure to the severity of hospitalized TBI care, we did not collect data on caregiver burden metrics, or on whether the surrogate cohabitated with the TBI survivor during the recovery process.⁴⁵ We intentionally minimized respondent burden by only including a limited number of demographic and TBI exposure questions after the GOSE Scenario Survey to avoid stereotype threat.³⁰ We used multiple solicitation methods to cast a broad net for potential surrogate respondents, but in doing so, sacrificed control over this large sample population. Relative to our cross-sectional study design, a prospective cohort study design would have allowed better definition of TBI survivors’ trajectories over time and the impact of these trajectories on the surrogates’ determination of health utilities. Finally, the present study was not designed to survey TBI survivors directly. Future work may better capture the experiences of TBI survivors at GOSE 3, 4, and 5 levels of functional status in particular, where voices are lacking with important implications for clinical practice and clinical trials. In the multicenter CENTER-TBI observational study of predominantly mild TBI patients, some patients with post-TBI disability had higher quality of life than expected based on their GOSE alone.⁴⁶

Conclusions

In this index study of surrogate perceptions about disability after TBI, poor neurologic outcomes – vegetative, needing all-day or partial daily assistance – were perceived as worse than death by at least one in three surrogates. Surrogate perceptions differed in important ways vis-à-vis provider perceptions and public perceptions unexposed to the realities of hospitalized TBI and the recovery trajectory. These long-term perceptions about post-TBI disability may inform earlier, tailored shared decision-making after neurotrauma and improve patient-centered outcomes for future TBI clinical trials.

Abbreviations

BTQ

Brief Trauma Questionnaire

CHERRIES

Checklist for Reporting Results of Internet E-Surveys

EQUATOR

Enhancing Quality and Transparency of Health Research

GOSE

Glasgow Outcome Scale Extended

ICP

intracranial pressure

IQR

interquartile range

IRB

Institutional Review Board

SIBICC

Seattle International Severe Traumatic Brain Injury Consensus Conference

TBI

traumatic brain injury