Preoperative assessment of the frail older adult

Our population is changing rapidly. From 2007 to 2014 the number of adults in BC age 65 and older rose 28.0% to represent almost 16.0% of the population. Estimates suggest that by 2050 more than 20.0% of adults in the over-65 age group will be 85 and older. In 2006 in the US, 35.3% of patients who underwent inpatient procedures and 32.1% of patients who underwent outpatient procedures were 65 and older.[1,2]

The care of frail older adults is challenging because many are medically complex patients. Assessing their suitability for surgery poses a particular challenge. A growing body of literature and experience tells us that while traditional instruments can reliably predict postoperative outcomes in robust younger adults, these tools lack the ability to predict the negative postoperative medical and functional outcomes for frail older adults. An assessment and decision-making model that considers frailty, function, goals, values, and capacity can address this challenge and allow us to make better predictions of likely functional and medical outcomes in frail patients. With this information we can help older adults make fully informed decisions about possible surgical procedures.

Preoperative assessment tools
Traditional preoperative assessment tools have limitations when applied to frail older adults because they tend to look at a single disease entity and lack the ability to predict functional outcome. By contrast, a frailty assessment tool for patients undergoing surgery has great predictive value for major adverse cardiovascular events (MACE), functional decline, mortality, and morbidity. A recent systematic review found that frailty, as assessed by multiple criteria, had a robust positive relationship with the risk of MACE after cardiac surgery (odds ratio, 4.89).[3] Moreover, preoperative frailty assessment was able to predict relevant outcomes such as functional decline and institutionalization.

Robinson and colleagues measured frailty characteristics in 110 patients undergoing major surgery and found that the presence of three or more characteristics predicted institutionalization after discharge with a sensitivity of 82% (CI 70% to 90%) and a specificity of 84% (CI 77% to 89%).[4] Four or more characteristics predicted 6-month postoperative mortality with a sensitivity of 81% and specificity of 86%. While this was a small study, many other studies have found that including the measurement of frailty characteristics adds significantly to the predictive accuracy of standardized instruments.[5]

Understanding frailty
Frailty refers to a state of vulnerability and diminished physiological reserve that impedes the body’s ability to withstand and recover from minor challenges such as surgery. In the frail individual, reserve and function are diminished across multiple physiological systems and the ability to cope with everyday or acute stressors is compromised.

Recognizing a patient’s vulnerability assists in preoperative decision making because:
•    Frailty is a predictor of perioperative risks.
•    Frailty places patients at highest risk of functional decline.
•    Frailty is an independent predictor of mortality.

The Fried criteria[6] provide a phenotypic description of frailty that includes the following elements: weak grip strength, exhaustion, slow walking speed, low physical activity, and unintentional weight loss.

In order to operationalize these criteria, many authors have developed functional parameters and scored these on a frailty index. These parameters include dependency on others for activities of daily living (ADL), dependency on others for instrumental activities of daily living (IADL), cognitive impairment, falls, slow gait speed, involuntary weight loss, and poor nutrition.

Recommended model for preoperative assessment and decision making
The model we recommend for preoperative assessment and decision making is summarized in the Figure on the following page. The process begins with an evaluation of frailty and competing comorbidities that may pose a mortality risk to the patient before the presenting disease does.

Identifying frail patients
Age alone is not a robust predictor of outcome when used in studies that evaluate frailty.[7] There are many validated tools to quantify frailty, but essentially they all describe a patient who has limitations of strength, energy level, and weight loss unrelated to acute medical illness. These elements manifest as increased dependency in both basic ADL and IADL, impaired cognition, and decreased mobility.

Some studies have used the Fried criteria in their perioperative assessment. Others have operationalized the frailty concept using a combination of scores based on functional parameters. Afilalo and colleagues[8] studied the value of testing gait speed, increasingly recognized as a helpful single characteristic for assessing frailty and vulnerability. Afilalo and colleagues compared gait speed as a predictor of perioperative outcomes to a standardized tool for predicting outcomes in patients undergoing cardiovascular surgery. They found that slow gait speed outperformed the Society of Thoracic Surgeons (STS) score when predicting mortality and major morbidity.[8] A gait speed of less than 5 m/6 s (0.83 m/s) with a high STS score was associated with a 43% increase in postoperative mortality or major morbidity for patients undergoing cardiovascular surgery.

In a prospective cohort study, Afilalo and colleagues compared different frailty scales that included gait speed scores. They found that only the 5-m gait speed score was statistically significant. Gait speed demonstrated an odds ratio of 2.63 for predicting postoperative mortality or major morbidity and demonstrated superior predictive ability when compared with other frailty scales.[9]

The Canadian Study of Health and Aging frailty index (CSHA-FI) measures accumulated deficits, while the National Surgical Quality Improvement Program (NSQIP) modified frailty index (mFI) relies on a simplified version of the CSHA-FI using 11 variables. Farhat and colleagues found that the mFI is highly predictive of postoperative outcome, including mortality and wound infection.[10]

To date there is no consensus regarding the frailty assessment tool with the best predictive value for postoperative outcomes. Buigues and colleagues tried to answer this question in a systematic review and found that despite the presence of strong evidence supporting the use of frailty assessment to predict postoperative outcomes, no single frailty scale was superior.[11]

In a prospective study Makary and colleagues applied the Fried criteria and gait speed to determine frailty. They found a strong association between frailty and postoperative complication, discharge to nursing home, and length of hospital stay with reasonable predictive ability.[12] Frail patients had a 7.7-day hospital stay, a 45.0% postoperative complication rate, and a 43.0% chance of disposition to a skilled nursing facility. Patients not assessed as frail had a 4.2-day hospital stay, 19.5% postoperative complication rate, and a 2.9% chance of disposition to a skilled nursing facility. It is important that we consider these outcomes and inform our patients of the possible consequences of surgery so that they can arrive at the best decisions for themselves as individuals.

Assessing the impact of disease and comorbidities
Frailty can mask or mimic medical illness and it is important to determine whether the limitation the older adult is experiencing is due to frailty or disease. For example, is the patient’s lack of mobility due to frailty or progressive aortic stenosis?

Older adults have multiple comorbidities and the likely outcome and impact of these on morbidity and mortality should be incorporated into the decision-making process. For instance, will the patient’s moderate to severe dementia limit function long before a competing cancer affects health status and survival?

Assessing geriatric risks 
Geriatric risks include delirium, decreased mobility, functional loss, and disposition to a skilled nursing facility. The evidence we have to quantify the risk for functional loss is relatively weak, but a number of studies provide some guidance when looking at the other geriatric risks.

Delirium. Goldenberg and colleagues developed a tool for predicting delirium using age, cognition, number of medications, and values for albumin and hematocrit. If the patient met the criteria for all of these, the risk for delirium was 100.0%. If the patient did not meet the criteria for any of these, risk was 3.0% or less.[13] Witlox and colleagues[14] found that if an individual experienced a postoperative delirium the risk of mortality at 2 years was 38.0% (HR 1.95), the risk of dementia in the subsequent 4 years was 62.5% (HR 12.52), and the risk of institutionalization at 14.6 months was 33.0% (HR 2.41). 

Decreased mobility. The loss of independent mobility is a dreaded outcome for older adults. Although we could not identify any studies that use this as a specific outcome, Schoenenberger and colleagues investigated deterioration of independent function after transcatheter aortic valve implantation (TAVI) and found that frailty was associated with decline in at least one basic ADL in 31.0% of patients.[15] 

Disposition. Makary and colleagues found frailty predicted postoperative disposition to a skilled nursing facility in 42.0% of patients.[12] This geriatric risk and the others discussed above will concern every older adult undergoing surgery and should be communicated as part of the informed consent process.

Assessing capacity to consent
Cognitive assessment to determine capacity to consent is essential. If a patient is cognitively impaired, we need to ensure that consent to any intervention is based on a full understanding of the complex risks involved, especially the risk of delirium. A cognitive assessment is also needed to provide a baseline for future assessment and comparison.

Discussing goals and values
While longevity may be important to care providers and sometimes family members, many patients do not hold this as their highest value and may value independent functioning and quality of life more. A frank discussion regarding values is fundamental to decision making. Evidence from Mallery and Moorhouse suggests that these discussions can shift patient decisions significantly.[16] If patients have moderate dementia and the prospect of progressive dependency, they may not wish for life-prolonging cancer surgery or valve surgery. Also, if the surgery could prolong life but threaten independent living, they may choose not to undergo surgery.

In addition, a realistic discussion of the patient’s goals and expectations is important. For example, if the patient and family hope that a TAVI will improve cognition and cognitive improvement is the goal, they need to understand there is no strong evidence it will do this and that delirium and worsening cognition may result.

Mitigating risk
Ongoing studies are looking at preoperative exercise programs for frail patients undergoing coronary artery bypass graft (CABG) and valvular surgery to assess improvement in functional capacity postoperatively.[17] While completed studies suggest that nutritional supplementation has a positive impact on frail elderly patients,[18] it remains to be seen whether preoperative supplementation can improve frailty and in turn postoperative outcomes. In all cases a thorough review of medications is key, especially given that medications can predispose a patient to delirium.

Summary
Determining perioperative risk can be challenging when a patient is frail. There is a growing body of evidence that provides guidance in preoperative assessment of the vulnerable frail older adult and helps us inform patients of the functional, mobility, and cognitive risks they face so they can arrive at the best decision about treatment based on their goals and values. A model that incorporates a comprehensive geriatric assessment and uses adapted predictive tools based on the most current evidence can improve our ability to provide appropriate care to frail patients.

Competing interests
None declared.

This article has been peer reviewed.

Dr Hill is a clinical associate professor in the Division of Geriatric Medicine at the University of British Columbia and head of the Division of Geriatric Medicine, Providence Health Care. Dr Alkeridy is a subspecialty resident in the Division of Geriatric Medicine at the University of British Columbia.