A Patient-guided Measure of Functional Status Following Traumatic Brain Injury: The Bob Brown Scale


 Background: While technology for treating traumatic brain injury (TBI) continues to improve, unexplained variation in outcomes persists. A growing body of literature provides evidence for the influence of subjective experience on objective physiological processes. We worked with TBI survivors and their family/caregivers to develop a model of recovery that incorporates patients’ subjective experience with objective measures of functional status (FS) to investigate factors that foster recovery. We then developed a measure of FS that reflects the priorities and perspectives of the patient population. The purpose of this report is to present our process for establishing the reliability and validity of the FS measure.Methods: Through an iterative process we elicited from TBI survivors their priorities for evaluation and measurement of FS. We assembled their responses into an 18-item measure and administered the pilot survey to 68 people with TBI from support groups in Oregon. Final administration was a survey mailed to 837 TBI survivors. Reliability analyses were conducted to evaluate internal consistency. A Cronbach’s Alpha was calculated for each sub-construct of the scale. A Difficulty Function was calculated to compare patient responses to family responses within each construct. Content validity analysis was conducted comparing constructs and items from the measure to those of 11 established, validated TBI outcome measures. an Analysis of Variance (ANOVA) was conducted to investigate the relationship between survivors’ FS scores and their post-trauma socio-economic status.Results: There were 248 complete cases in the sample. Reliability analysis provided adequate inter-item correlations. Difficulty functions ranged from .69 to .47. Construct validity was established with 11 validated instruments used to evaluate TBI. The ANOVA revealed a significant difference between FS scores and post-trauma SES; survivors with lower FS scores experienced lower post-trauma SES.Conclusions: Our goal was to develop a measure of functional status following TBI using the advice and assistance of survivors of TBI and their families. In this paper we present our process for developing and assessing the reliability and validity of the Functional Status measure of the Bob Brown Scale. Future research will be use of the BBS in a prospective sample of TBI survivors.

correlations. Di culty functions ranged from .69 to .47. Construct validity was established with 11 validated instruments used to evaluate TBI. The ANOVA revealed a signi cant difference between FS scores and post-trauma SES; survivors with lower FS scores experienced lower post-trauma SES.
Conclusions: Our goal was to develop a measure of functional status following TBI using the advice and assistance of survivors of TBI and their families. In this paper we present our process for developing and assessing the reliability and validity of the Functional Status measure of the Bob Brown Scale. Future research will be use of the BBS in a prospective sample of TBI survivors. Background Traumatic brain injury (TBI) has been identi ed as the "most complex disease in the most complex organ" of the body. 1 Despite establishment of evidence-based treatment guidelines, 2 and considerable investment in large, multi-centered research consortia, [3][4][5][6] little progress has been made in acute care therapeutics for TBI, and clinical trials consistently fail. 7,8 From clinical practice, observation, and long-term follow-up studies, considerable variance in outcomes among those who survive TBI is observed, even after accounting for severity of injury and differences in patient and treatment characteristics. [9][10][11][12] A dynamic and multi-systemic method to evaluate recovery is required to understand these differences and to identify factors that foster recovery. 11,13−17 A growing body of literature provides evidence for the in uence of subjective experience on objective physiological processes. [18][19][20][21][22][23] "The study of the placebo effect re ects a current neuroscienti c thought that has as its central tenet the idea that 'subjective' constructs such as expectation and value have identi able physiological bases, and that these bases are powerful modulators of basic perceptual, motor, and internal homeostatic processes." 18 (p. 10,390). This evidence supports the need to include subjective factors − what a patient is experiencing, which is potentially distinct from objective reality − in models of recovery.
Research speci c to TBI identi es subjective factors that appear to in uence patients' objective functional status, including perception of autonomy, 24 "sense of self", 25 depression, 26 self-awareness, [27][28][29] subjective experience of di culties, 30 and self-reported "psychosocial status" 31 (global representation of subjective and objective function). Given this, we assert that patient and family/caregiver experience should guide the development of instruments for assessment of outcomes and post-injury functional status.
We worked with TBI survivors and their family/caregivers to develop a model of recovery that incorporates patients' subjective experience with objective measures of functional status to investigate factors that foster recovery and how they interact in a dynamic system. We used theory and instruments from the disciplines of Systems Science, Medicine, and Developmental Psychology as the foundation for the model. Figure 1 illustrates the model. The arrows indicate the direction of in uence. There is a recursive relationship between Social Context and Perception, and both affect how the individual will engage in the recovery process. Severity can have a direct in uence on all 3 domains, and the dynamic interaction among the four domains determines Functional Status. A complete account of the development of the model is published elsewhere. 32 As a part of this model, we developed a measure of functional status (FS) that re ects the priorities of the patient population. We named the instrument the Bob Brown Scale (BBS) in honor of the late Robert C.
Brown, a TBI survivor and leader of the TBI support community in Oregon. The purpose of this publication is to present our process for establishing the reliability and validity of the FS measure of the BBS.

Methods And Materials
Samples All participants were referred by the Brain Injury Support Group of Portland, Oregon (BISG), a nonpro t organization that networks and educates TBI survivors, family members, and professionals in Oregon, Southwest Washington, and Western Idaho.

Instrument Development
We interviewed a convenience sample of 12 people (5 TBI survivors, 4 relatives, and 3 professionals) and, using a free-list method, 33 elicited their priorities for evaluation and measurement of TBI outcome. We condensed their responses, placed them into 70 sets of triads in a survey form. 33 We subsequently administered the triad comparison survey during a local social club meeting to 30 people (21 TBI survivors and 9 family or caregivers). We categorized their responses into dimensions.
Concurrently we reviewed research to identify instruments that were being used to evaluate brain injury (Table 1). From them we identi ed 14 dimensions of measurement (executive function, cognition, personality/emotionality, social, satisfaction, psychological, occupational, activities of daily life, family, nancial, independence, physiological, adaptation, and descriptive). We matched the survivor-identi ed dimensions with those from the existing instruments and condensed them into ve key dimensions for measuring functional status after TBI: Memory, Mobility/Independence, Organization/Productivity, Inappropriate Behavior, and Physical Limitations. Using these dimensions, we constructed a draft survey, and administered it to a focus group of 12 people; 9 with TBI and 3 family members. The family input allowed for generation of a "family version" of the survey.
We developed a second draft of the pilot survey and administered it to 68 people with TBI in social club and support group meetings in Portland, Bend, Corvallis, and Eugene, Oregon. From this input, we developed a nal version. Survey packets for the nal administration were mailed to 837 people on the membership list of the BISG. Appendix A is a complete list of constructs and variables for the BBS FS.

FS Measure
The BBS FS is an 18-item instrument (spanning the ve key dimensions identi ed above) with a total score formed by summing items (see Appendix A). Response options for individual items range from 1 (worst) to 4 (best); thus, total scores range from 18 (worst) to 72 (best). We collected BBS surveys for each case from both the patient and a family member or care giver and assessed concordance.
We also assessed Functional Status by measuring socio-economic status (SES), which was a composite of education, occupation, and income. The post-trauma SES score was subtracted from the pre-trauma SES score to calculate the difference (SES CHANGE), and categorized as "decreased," "no change," or

Results
Of 837 mailed surveys, 402 were returned; 210 from survivors and 192 from family members. Onehundred fty-four cases have a response from both a survivor and a family member. Fifty-six cases have only a survivor response, and 38 have only a family response. There are 248 cases in the sample. Table 2 contains demographic characteristics of the respondents. Means and standard deviations for the BBS FS are presented in Table 3.

Discussion
Our goal was to develop a measure of functional status following TBI using the advice and assistance of survivors of TBI and their families. In this paper we present our process for developing and assessing the reliability and validity of the Functional Status measure of the Bob Brown Scale. The Cronbach's α's for the patient and family Bob Brown Scales indicate good internal consistency. The α coe cient for the subconstruct of Organization/Productivity was lowest for both patient and family reports. Revisions of items in this sub-construct may be considered if similar results are found in subsequent studies. The high interitem correlation for the Mobility/Independence sub-construct when reported by patients or by family members validates anecdotal information gathered in focus groups that independence is an important and easily observed aspect of functional status.
The correlations representing the Di culty Function indicate considerable disagreement between patients and family on responses. However, the function was calculated based on identical answers, and does not account for degrees of difference. The low correlations might be the results of errors or bias present when gathering retrospective data. Reporting errors might even be greater when relying on retrospective reports from survivors of TBI. The lowest correlation of patient and family responses was observed in the subconstruct of Organization/Productivity. On average, patients rated themselves higher than family on these items. The items in this sub-construct (refer to Appendix A) may require more abstract thought than is possible for many persons with TBI. This result suggests the need for a critical evaluation and possible revision of items for this sub-construct.
Content validity of the BBS was established comparing constructs and items of the BBS FS with those of 11 validated instruments used to assess functional status in patients with TBI.
Group comparisons showed that functional status was lower among TBI survivors who had reported a decrease in socioeconomic status from pre-to post-trauma. However, it is unclear from this design whether lower functional status is caused by circumstances such as loss of livelihood and/or change in independence, or whether the lower functional status after trauma precipitates the negative circumstances. Regardless of the direction of the causal pathway, it is clear that SES and functional status do co-vary.
Limitations. Most respondents in this sample were survivors or family members who had some association with the Brain Injury Support Group. Their participation ranged from active to peripheral, but the fact that they had at least one contact with the support organization de nes the parameters of the sample. It does not represent persons with mild injuries who may not seek support, or who choose not to be associated with a support group. Similarly, the lowest functioning survivors and the socially isolated are not well represented in this sample. In addition, all but 24 patients underwent some form of rehabilitation. Our survey was created and responded to by the subset of TBI survivors in the mid-range of function and can be generalized only to that population. The instrument was developed with a convenience sample utilizing a retrospective design. The next steps will be to revise the instrument and implement it in a prospective study.

Conclusions
As discussed in the Introduction, in collaboration with TBI survivors and their family/caregivers, we developed a model of recovery 32 that incorporates patients' subjective experience with objective measures of social context, severity of injury, and functional status, to investigate factors that foster recovery and how they interact in a dynamic system (see Fig. 1). The BBS FS is the functional status component for the overall model. With establishing the reliability and validity of the BBS FS, we can move forward to use the entire model to assess the dynamic process of recovery from TBI in a prospective study. Use of such a model may provide access to key factors that in uence recovery, which could inform the development of more effective interventions.