Traumatic brain injury in older adults: do we need a different approach?

Matthew E Peters*,1 & Raquel C Gardner2,3 1Department of Psychiatry & Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, MD, 21205, USA 2Department of Neurology, University of California San Francisco, CA, 94143, USA 3San Francisco Veterans Affairs Medical Center, San Francisco, CA, 94121, USA *Authors for correspondence: Tel.: +1 410 550 6337; Fax: +1 410 550 0564; matthew.peters@jhmi.edu

trauma-field triage criteria to optimally identify older adults with TBI who require emergent transfer to a trauma center [13], and neurorehabilitation practices specific to older adults with a focus on removing 'excess disability' [14]. In addition, neurocritical care teams increasingly include a geriatrician and the next edition of the Resources for Optimal Care of the Injured Patient includes accreditation standards for geriatric trauma care [17].
Despite these successes, many challenges remain. Widely used measures for determining initial TBI severity and estimating prognosis in the acute clinical setting are problematic in older adults. As an example, the GCS is a reliable predictor of morbidity and mortality in younger, but not older, adults who may have an abnormal GCS at baseline or an intact GCS despite accumulating intracranial hemorrhage [3,18,19]. Prognostic models for outcome prediction after TBI (e.g., Corticosteroid Randomization After Significant Head injury CT (CRASH CT), Immediate Post-concussion Assessment and Cognitive Testing [IMPACT]) show poor prognostic performance in older adults with TBI [3]; perhaps because these models do not include key geriatric outcome predictors such as comorbidities, polypharmacy, baseline function and frailty [20]. Recently, the US FDA approved two blood-based biomarkers, GFAP and UCHL1, to aid in the evaluation of mild TBI [21]. However, a small study demonstrated that this GFAP assay is significantly less accurate for identifying CT evidence of intracranial trauma in older versus younger adults with mild TBI [22]. With diagnostic biomarkers rapidly making their way into clinical practice, much more research is needed to guide age-appropriate use.
Lastly, most in-hospital deaths in older adults with TBI occur after elective withdrawal of care [23]. Withdrawal of care decisions are frequently made within 72 h of injury, despite evidence that lack of neurological improvement within 72 h in older adults with severe TBI does not predict long-term recovery among survivors [9]. There are dramatic variations in mortality rates after geriatric TBI between centers [3], suggesting a lack of reliable prognostic indicators to guide acute care decisions in this population. Thus, despite recent and ongoing efforts to improve ageappropriate management of geriatric TBI in the acute setting, there remains much work to be done to implement current evidence into widespread practice and to further develop and refine evidence-based prognostic indicators and care guidelines in order to optimize medical decision making and outcomes in this population.

The need for a geriatric approach to TBI clinical research
In an effort to study 'pure TBI' many prior studies have implemented upper age limits or excluded patients with pre-existing conditions, a practice which preferentially excludes older participants and limits generalizability of findings to this population. Although broad inclusion of older adults in TBI research comes with challenges, we argue that these challenges must, and can, be surmounted if we are to advance care and improve outcomes in the large and rapidly growing population of older adults with TBI. Here, we review some of these challenges and discuss possible solutions.
• TBI diagnosis is challenging in an older adult with head trauma in whom loss or alteration of consciousness may be attributed to TBI, concurrent medical conditions (e.g., stroke, syncope, seizure, dehydration and dementia), or both and therefore may preclude a TBI diagnosis using standard clinical criteria [24,25]. In these ambiguous cases of head trauma, we have chosen to rely on evidence of acute intracranial trauma on neuroimaging studies, such as head CT, for the diagnosis of TBI. Ambiguous cases of head trauma without neuroimaging evidence of intracranial trauma remain a diagnostic challenge and deserve additional dedicated study. Ultimately, a battery of neuroimaging-and blood-based biomarkers may be needed to definitively make the diagnosis in the most challenging or ambiguous cases and more research is needed to identify these optimal diagnostic biomarkers. • TBI outcome assessment is challenging in an older adult with multiple pre-existing conditions as it may be impossible to isolate the effect of TBI. This is not a new problem in TBI research and has many parallels to the study of TBI in the context of polytrauma. We propose that this challenge can be addressed by systematically measuring and studying the impact of, rather than excluding for, pre-existing conditions and disability. The American College of Surgery Trauma Quality Improvement Program Geriatric Trauma Management Guidelines (all-cause; not TBI-specific) [26] recommends measuring pre-existing comorbidities, functional status and physical frailty in all geriatric trauma patients as these have been proven outcome predictors in geriatric trauma surgery and geriatric medicine inpatient populations [27,28]. Reports of pre-injury disability are already routinely elicited in the context of the most widely used TBI outcomes assessment interview (the Glasgow Outcome Scale Extended [29]) in which patients or their informants are asked to compare current disability to pre-injury disability. Systematically quantifying and recording pre-injury disability, either via patient or informant report, is therefore only a small departure from current practice. Efforts are currently underway to develop improved TBI end points [30]. It is therefore an ideal time to build expert consensus around a set of geriatric-specific TBI end points and common data elements that will facilitate broader inclusion of older adults in TBI research and inform design of clinical trials. • Frail older adults may be unable to complete outcome assessments. Greater use of proxy informants and study partners, as are commonly used in geriatrics and dementia research, may facilitate inclusion, retention and outcome assessment of the more frail or disabled older patients. Innovative follow-up methods, such as home visits and telemedicine assessments, have the potential to increase follow-up rates and generalizability of results, but will require increased staffing and financial resources. Because older adults with pre-existing conditions may be more likely to have implanted medical devices, MRI clearance protocols (e.g., collection of model and manufacturer information on implanted devices) will be needed to reduce unnecessary exclusions from MRI protocols.
By combining methods commonly used in geriatric research with those already used in TBI studies, the challenges of inclusion of older patients in TBI research can and must be overcome. Only then can the research community achieve generalizability to real-world older adults with TBI, develop better diagnostic and prognostic tools to guide care, design inclusive trials and optimize outcomes.

Conclusion
This article's title poses the question "Do we need a different approach to TBI in older adults?". The answer is 'yes', but we are not starting from scratch. Clinical endeavors, such as comprehensive, multidisciplinary fall and TBI clinics, are increasing. Geriatric research on conditions relevant to the TBI population (e.g., Alzheimer's disease and related dementias) stand to teach us much about how to design studies focused specifically on older adults with TBI. Older adults with TBI deserve the same advocacy, and focused study, as sports-and military-related TBI. The path forward is increasingly illuminated and the number of us carrying the torch is growing.

Financial & competing interests disclosure
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.

Open access
This work is licensed under the Creative Commons Attribution 4.0 License. To view a copy of this license, visit http://creativecomm ons.org/licenses/by/4.0/