Frailty may be measured in many ways. Not every measure of frailty is appropriate for every patient because all variables are not attainable in every patient. At least a dozen frailty measures are relevant to trauma or ICU patients and hold some validity at present (Table 10.1).

Dr. Linda Fried first described frailty as a unique phenotype in 2001 [6]. Fried’s frailty phenotype is distinct from either comorbidity or disability and is characterized by three of the following five findings: fatigue, diminution of strength, weight loss (>10 pounds in last year), decreased physical activity, and exhaustion. This definition is now widely accepted, and tallying the components constitutes one of the longest-standing and simplest frailty measures: not frail (score 0 components), pre-frail (1–2 components), and frail (3–5 components) [7]. Fried’s criteria correlate with social, psychological, and physical function [8]. Fried score and other tools have been used to predict death and disability in comorbid outpatients and also major adverse cardiac and cardiovascular events in older patients undergoing transcatheter or open cardiovascular procedures [9]. Predictive value of Fried criteria in ICU patients has not been established [10].

Frailty index (FI) is the most studied of all frailty measures. Described in 2001 by Minitski et al., from Montreal, FI was developed using 92 items from the Canadian Study of Health and Aging [11] and remains an important frailty assessment and a validation benchmark for novel frailty measurement tools. FI uses a list of dichotomous and ordinal values expressed over a denominator. If only the dichotomous variables of FI are used (excluding the ordinal variables which grade debilitation on a scale), FI is easier to calculate and maintains high predictive value for mortality, as proven in several large population-based databases [12]. FI and Fried criteria are used frequently in clinical practice and have been expanded upon and modified by many with the resulting tools often being referred to as “modified frailty index” and “modified Fried criteria.”

Necessity for easily reproducible, reliable predictors of specific outcomes in specific populations has led to the development of novel frailty assessment tools. Joseph et al. have modified the FI to a shorter, 15-variable trauma-specific FI (TSFI) that reliably predicts unfavorable discharge (discharge to a subacute nursing facility) [13]. The Trauma Quality Improvement Program (TQIP) Geriatric Trauma Management Guidelines recommend a simple test such as the Short Simple Screening Test for Functional Assessment, which, if positive, can then be followed by a more detailed evaluation for functional impairment [14].

Long before Linda Fried’s description of the frailty phenotype, the timed “get up and go” test was described by Mathias et al. as an assessment of balance [15]. Also called the “timed up and go” test or TUG, this test is performed by recording the time in seconds required to rise from an arm chair, walk to a line 3 m away, return, and sit again. TUG correlates well with the Berg Balance Scale which measures fall risk and the Barthel index which measures ADL [16, 17]. The TUG has been modified (m-TUG) and expanded (ETUG) and the novel versions evaluated systematically, but these modifications may not significantly change the original TUG’s utility for the geriatric ICU/trauma population [18].

The Berg Balance Scale (BBS) is a well-studied and reliable assessment of functional status that focuses on standing from a sitting position, stationary standing, and transfers. BBS cannot be easily applied in patients with limited lower-extremity weight bearing and is therefore most valuable in monitoring progress during rehabilitation [19].

Gait speed can be assessed at a health maintenance visit, preoperatively, or in ambulatory hospitalized patients, in which case it can be used in planning transition of care. Pamoukdjian performed a systematic review of gait speed, both by itself and as a component of the TUG, and found a gait speed of less than 1 m/s over a 4-m distance to be a predictor of early death, disability, falls, and institutionalization [20].

The British Geriatric Society (BGS) has recommended that frailty be assessed in older patients whenever they encounter community health or social service providers and that simple tools like gait speed (<0.8 m/s) and the TUG test (>10 s) be used as indicators [21]. In its recommendations, the BGS suggests that if signs of frailty are identified, they should prompt discussion with patients and families about frailty, direct attention to medical illnesses that could be treated, and suggest those frailty signs serve as a basis for physical conditioning. Such an approach would be ideal prior to major elective surgery, especially in patients who are likely to require admission to the ICU postoperatively.

The Short Physical Performance Battery (SPPB) can be used in community and in acute settings. SPPB assesses three parameters of physical function: (1) static balance, (2) gait speed, and (3) standing from a chair. Each category scores 0–4 with a composite score of 12. SPPB is a valid, reliable predictor of falls and functional dependence and has been correlated with frailty index and the modified Fried criteria [22, 23].

The Vulnerable Elders Survey-13 (VES-13) predicts complications and death in geriatric trauma patients [24]. The Life Space Assessment (LSA) developed at the University of Alabama at Birmingham assesses mobility over the preceding 4 weeks including mobility away from home, frequency of mobility, and level of independence from assistance or assistive devices. Decreasing LSA coincides with falls, but the LSA’s predictive value in hospitalized patients has not been validated [25].

Handgrip strength is commonly used to assess weakness, which is an important adverse effect of hospitalization and one of the Fried criteria for frailty. Grip strength alone predicts ICU paresis and also correlates with Physical Function Intensive Care Test (PFIT) and 6-min walk test (6MWT), though the exact clinical utility of PFIT and 6MWT themselves is uncertain [26]. PFIT, grip strength, and 6MWT can, however, identify trends and thus may be used to measure effect of interventions in the ICU targeted at reducing the harmful effects of ICU stay on functional status.

Functional independence measure (FIM) is a well-validated measure of overall physical function, social and psychological function, and ADL. FIM predicts post-discharge functional status, is worse at discharge in patients with a lower admission FIM score, and correlates with frailty index [27]. The Chelsea Critical Care Assessment Tool has also been assessed in the ICU population and been found to be valid in measuring changes in performance among ICU patients, but its relevance specifically in geriatric and frail ICU patients has yet to be established [28].

The Score Hospitalier d’Evaluation du Risque de Perte d’Autonomie (SHERPA) uses five variables: age, ADL performance, prior falls, self-rated health, and cognitive impairment, to predict functional status after hospital discharge. The SHERPA’s utility above other scoring systems remains unproven [29]. Contrary to what one might think, dementia alone may not be a reliable predictor of outcome among ICU patients [30]. However, the Assessing Dementia 8 (AD8), a commonly used screening tool for Alzheimer’s disease, correlates with driving errors so therefore is perhaps relevant to injury prevention measures on discharge among older drivers [31].

Several assessment tools have been used in the primary care, community health, and preoperative settings but because of complexity may not be applicable to the ICU or acute care environments. The Katz Activity of Daily Living Survey measures ADLs, mobility, and instrumental ADL and has been widely validated for a number of outcome measures. Originally known as the Katz-6, it has now been modified to the Katz-15 which predicts unfavorable overall health outcome [32] Green et al. combined the Katz activity of daily living survey with serum albumin, grip strength, and gait speed to derive a novel score predicting outcome after transcatheter aortic valve replacement (TAVR) [33]. The SF-36 correlates with frailty index and postoperative complications [34]. The cognitive mini exam (Mini-Cog) has also been used to predict outcome [35].

Some assessment tools for frailty can be performed without physically evaluating the patient. The Charlson comorbidity score was first reported by Charlson et al. in 1987 and uses diagnosis data from 19 categories to predict long-term outcome, including functional recovery at 1 year in elderly ICU patients [36, 37] Kim et al. developed a predictive model of poor functional status using claims data. The model offers a high level of discrimination and could be used in epidemiologic studies and in understanding resource use across a healthcare system or population [38]. Laboratory and radiology tests can also be used to assess frailty, guide resource use, and predict prognosis. In one recent study, albumin was used as one component, along with COPD, assisted status, tube thoracostomy, injury severity score, number of rib fractures, and CHF, to comprise a frailty score to predict intubation and pneumonia in elderly patients with rib fractures [39]. Albumin has long been used as a predictor of outcome in surgical patients and can predict institutionalization at discharge after surgery among geriatric patients though not as reliably as TUG and overall functional dependence [40]. CRP, IL-6, TNF-α, D-dimer, albumin, IL-1, 25(OH)D, and low cholesterol have all been correlated with frailty and functional decline, but their clinical application is not yet clear [5].

A rapidly expanding body of evidence supports sarcopenia as an ideal objective measure for frailty that can be determined by diagnostic testing. Critical illness and muscle disuse are associated with altered protein synthesis and cell signaling that exacerbates skeletal muscle loss in the acutely and critically ill. The Nutrition and Rehabilitation Investigators Consortium reported sarcopenia to be a reliable indicator of ventilation, ICU stay, and death [41]. Future studies must elucidate the differential roles of preadmission sarcopenia versus sarcopenia as a result of critical illness [42].

Frailty assessment should help guide treatment. Surrogate responses must sometimes be used to assess physical function and frailty because the patient cannot participate completely or at all [43]. Maxwell et al. reported good agreement between patient and proxy on the VES-13, the modified Barthel index, and the Life Space Assessment [43]. Providers should assure that the frailty assessment tool they are using has been validated in the population to which it is being applied and should consider whether surrogate responses are valid.

Frailty assessment may require incorporating providers with geriatric training and experience into the trauma team, but this is not a requirement. Composition of the geriatric trauma and critical care team is a matter of great interest among providers and these compositions are changing. More than half of survey respondents who were members of the American Association for the Surgery of Trauma (AAST) reported that geriatric-specific personnel and resources are rarely or never used at their institution [44]. However, almost half of these respondents felt that further research regarding the “use of gerontologist, geriatric teams, or geriatric centers in the management of trauma patients” should be a high priority.

The American College of Surgeons Trauma Quality Improvement Program Geriatric Management Guidelines and many authors recommend proactive geriatric consultation. Seventy percent of the respondents in the AAST survey report that skilled nursing facilities (SNFs) have become the most common discharge destination, but few agree that SNFs are the best disposition. Geriatric consultation can reduce the risk of discharge to a long-term acute care facility and can also reduce the number of episodes of delirium, decrease inhospital falls, shorten length of stay, and improve functional recovery [45].

Geriatric Rehabilitation Units staffed by a multidisciplinary team, usually led by a physiatrist, can coordinate care of the elderly patient’s chronic health issues as well as their environmental factors. Regardless of whether the trauma/ICU team includes a geriatrician, multidisciplinary care is essential in geriatric trauma. The British Geriatric Society recommends a holistic medical review such as a comprehensive geriatric assessment (CGA) for the management of frailty that will diagnose medical illnesses, optimize treatment, apply evidence-based medication review checklists, include discussion with older people and caregivers to include defining the impact of illness, and then create an individualized care and support plan [21]. TQIP guidelines also recommend such CGAs for geriatric trauma patients with certain risk factors. The CGA, which should include a frailty assessment, may then serve as a basis for patient/family meeting regarding care plans and prognoses. Interventions should be targeted at optimizing physiology, minimizing complications, maintaining functional status, and returning the geriatric patient to their preadmission functional level and environment as often and as quickly as possible. For example, early physical therapy/occupational therapy in the ICU conveys better functional outcome when discharged, shorter periods of delirium, and more ventilator-free days [46].

Frailty assessment can be used to influence end-of-life care and palliative care decisions. Frailty index has been used to predict survival past 30 days in a specialized geriatric ICU. Every 1 % increase in FI correlates with an 11 % increase in mortality. In that study, no patient with FI > 0.46 survived past 90 days [47].