Comorbidity is defined as the simultaneous presence of two or more diseases in the same patient, an occurrence that becomes more common and changes phenotype with increasing age, because in the elderly, comorbidities are usually chronic, as opposed to acute in the young [37]. From the epidemiological point of view, significant comorbidity in the elderly is associated with increased risk of death [38], rehospitalization, disability, and poor quality of life [39]. Comorbidity influences the diagnostic and therapeutic process, as the coexistence of several diseases hinders the interpretation of clinical signs and symptoms. A classic example is represented by breathlessness in an elderly suffering from heart failure, chronic obstructive airways disease, and anemia.

The high prevalence of comorbidity in the elderly plays a central role in their exclusion from clinical trials [40]. In most geriatric cardiology studies, comorbidity is simply assessed in terms of disease count, rather than weighted on type of disease. However, this approach shows significant limitations, particularly when the outcome is not limited to mortality and morbidity, but includes the risk of becoming disabled. Marchionni et al. [41] clearly demonstrated that in elderly individuals with heart failure, the risk of becoming disabled is related to their disease index but increases differently in patients with COPD or cerebrovascular disease: in the first instance risk progresses arithmetically, in the second exponentially. This study demonstrated that the interplay between two chronic conditions is much more complex than the simple sum of risks, confirming the fallacy of a disease-count approach.

Finally, evaluation of comorbidity in the elderly must necessarily include an assessment of the severity of individual diseases. Among many studies, the Index of Coexisting Disease (ICED), Greenfield et al. [42], and the Geriatric Index of Comorbidity (GIC) [43] are worth mentioning. The ICED predicts risk of disability in patients suffering from chronic disease and consists of two subscales: the first subscale comprises a list of 14 chronic conditions with clinical severity increasing from 0 to 4, with the possibility to classify also diseases that are not on the list, and the second subscale includes 12 domains in which the functional disability is graded from 0 to 2. The ICED was validated in eight populations including patients with myocardial infarction or coronary artery bypass grafting [44]. The GIC consists of a list of 15 diseases with severity grading from 0 to 4, with the possibility to build four classes of comorbidity basing on the presence of one or more associated diseases [43]. While these scores are certainly able to characterize the level and severity of comorbidity, it is important to emphasize how – in many chronic conditions of the elderly – symptom severity may show a remarkably different association with risk of death as opposed to risk of disability. This geriatric paradox is best exemplified by severe osteoarthritis, in which the risk of becoming disabled is high when the symptoms are poorly controlled by drugs, but the risk of death appears low or even negligible.

The assessment of surgical risk in elderly patients cannot be considered comprehensive unless it includes evaluation of global functional status, defined as the capability of the physical and cognitive abilities to perform normal activities of daily living and to maintain sufficient social network. This concept is quantified in geriatric medicine by the basic activities of daily living (BADL) [45] and instrumental activities of daily living (IADL) [46] scales. The predictive value of these scales is supported by a wealth of solid literature. Fukuse and coll. [47] showed that in older patients undergoing thoracic surgery, a reduction of autonomy, measured by the number of lost BADL points, is able to predict major perioperative complications, independent of other risk factors.

Predicting the likelihood of remaining disabled following surgery is challenging, due to the complexity of capturing preoperative levels of physiological homeostatic reserves and vulnerability to stressful events, i.e., features that constitute the concept of frailty [47]. The conceptual definition of frailty is much discussed in geriatric medicine [48]. Campbell defines frailty as “a condition or syndrome which results from a multi-system reduction in reserve capacity to the extent that a number of physiological systems are close to, or past, the threshold of symptomatic clinical failure” [49]. Given this definition, there are two models of frailty suitable for the stratification of surgical risk of the elderly. They are summarized in the model called “frailty phenotype” arising from the Cardiovascular Health Study [50] and in the “deficit accumulation model of frailty” created by the Canadian Study of Health and Aging [51]. The first substantially recognizes a set of five domains: unintentional weight loss, muscle strength measured by handgrip, the feeling of self-reported exhaustion, walking speed, and the amount of physical activity habitual self-reported [50]. The Canadian model builds a frailty index from a list of over 70 both functional and clinical items, exploring the physical, cognitive, and disability in daily life, scoring from 0 to 7 for increasing fragility [51]. Both models have been tested in the stratification of surgical risk of elderly patients, proving adequate in predicting long-term adverse outcomes [48] but showing limitations related to the complexity of their implementation in clinical routine.

In this perspective, the assessment of surgical risk of elderly (including risk of disability) should be based on two key principles: the first is that clinical tools need to be simple and reproducible, because of the complexity of the surgical setting, and the second is that risk scores must have the ability to stratify perioperative risk as well as identify modifiable factors that may improve outcome [48]. Thus, in the preoperative setting, it appears very reasonable to rely on surrogate, but reliable, measures of frailty such as physical performance tests. The strength of the upper limbs measured with the dynamometer (handgrip test) or the gait speed performed in a short corridor is tools that have proved their worth in the field of surgical stratification of elderly patients [52, 53]. The attention toward measures of physical performance in the field of geriatric cardiology and cardiac surgery has significantly increased in recent years. Di Bari et al. have shown [54] that the Short Physical Performance Battery [55] is an independent predictor of prognosis in elderly patients with heart failure as powerful as the New York Heart Association (NYHA) class. Furthermore, at Duke University, Afilalo et al. [56] have demonstrated in 131 elderly undergoing cardiac surgery that reduced gait speed identifies a subpopulation at higher independent risk of mortality and morbidity. Of note, risk is 2–3 times higher for each level of the STS score [57] used for surgical preoperative screening by cardiac surgeons.

The rapid increase in the number of elderly people, due to extended life expectancy, is accompanied by an increased demand for cardiac surgery, to an extent that one of the most important topics in cardiology and geriatrics today is represented by the difficult choice of when and who to refer for cardiac surgery in this age group. As discussed, current risk prediction models have poor performance and are overall inaccurate in elderly patients, generally overestimating operative risk [58]. Nevertheless, the perception that older patients have lower functional reserves and more extensive comorbidity leads cardiologists and cardiac surgeons to hesitate in proposing cardiac surgery to the elderly. Nevertheless, despite a number of adverse features including increased incidence of diabetes, dyslipidemia, hypertension, and left main disease, rates of surgical mortality reported in the literature in octogenarians have fallen significantly, due to a reduction in the incidence of postoperative complications such as stroke and acute pump failure [59]. Contemporary cardiac surgical procedures can thus be performed safely and with therapeutic benefit in adequately selected nonagenarians [60, 61]. A recent study [62] analyzing the variations in the surgical population and type of intervention between 2001 and 2010 has shown a reduction in isolated CABG procedures with a concomitant increase in interventions on the valves (largely comprised aortic valve replacements) and thoracic aorta. With regard to age, there has been a significant reduction in patients <70 years and an increase in those > 80 years of age, with no change in the 70–79 year group. As a result, the average age at intervention has increased of about 10 years, with an increase in the proportion of frail patients in all age groups. However, particularly among patients >80 years, although only the oldest and most frail patients showed a significant increase in total hospital stay and intensive care unit stay. Based on the evidence available, it thus appears unjustified to deny access to cardiac surgery to any patient solely based on age, as the elderly receive comparable benefits compared to younger in terms of symptoms, functional recovery, and quality of life [63, 64].

In order to maximize the benefit of surgery for the elderly, however, it is critical to improve selection of optimal candidates. The Comprehensive Geriatric Assessment (CGA), developed in 1991, is a “multidimensional interdisciplinary diagnostic process focused on determining a frail elderly person’s medical, psychological and functional capability in order to develop a coordinated and integrated plan for treatment and long term follow up” [65] (Fig. 11.1). This tool includes assessments of medical, psychiatric, functional, and socioeconomical domains followed by a tailored management plan including rehabilitation; the team includes a minimum experienced medical, nursing, and therapy staff. The evaluation consists in the administration of standardized assessment tools to gather information in a semi-structured way. For example, this might include basic and instrumental activities of daily living for functional evaluation [45], Geriatric Depression Scale for psychological screening [66], or the Mini Mental State Examination (MMSE) for cognition [67]. Only few studies incorporated disabilities in activities of daily living as a predictor of outcomes after cardiac surgery [68, 69].

An effective test for the identification of frail elderly patients is the Short Physical Performance Battery (SPPB). The SPPB consists of three timed physical performance tests (walking speed, rising from a chair, and balance), each with a score from 0 to 4: a total score of 5 out of 12 identifies frailty. The SPPB is a powerful predictor of mortality in the elderly population, even after adjusting for age, sex, levels of multimorbidity, and cognitive status [55, 70].

In contrast to scales as the SPPB, which are composed of multiple items, gait speed and the measurement of the upper limb strength by a dynamometer are measurable indicators of frailty which can be employed as single items [71–73]. Gait speed has shown excellent intra- and inter-investigator reproducibility. The most important predictor in the test is not the total distance walked, which can vary in the different tests from 3 to 10 m, as the time needed to cover each meter [74]. At present, 5 m is the most commonly used distance, as it takes into account the possible occurrence of cardiopulmonary symptoms for longer distances. Of note, a pooled analysis of nine cohort studies showed a reduction of approximately 12 % in risk of death at 5 and 10 years follow-up for each 0.1 m/s increment of gait speed [75]. A patient with a walking speed lower than 0.8 m/s is at greater risk of loss of autonomy, declining health status, institutionalization, and death [76, 77].

Until recently, there were no studies specifically focusing on the use of gait speed as a predictor of postoperative mortality and morbidity in elderly cardiac surgery patients. A pioneering study in this field was conducted by Afilalo et al. [56], in 131 patients undergoing cardiac surgery at 70 years of age or more (mean age of 75.8 ± 4.4 years). Overall, 46 % were defined as slow walkers (time to walk 5 m ≥6 s) before cardiac surgery. The primary end point (a composite of in-hospital postoperative mortality or major morbidity) occurred in 23 %. Slow gait speed independently predicted the composite end point after adjusting for the STS risk score (odds ratio: 3.05; 95 % CI: 1.23–7.54). A subsequent study by the same group confirmed these results: slow walking and the presence of high-level disability (at least three impairments in Nagi’s scale) were associated with a significant increase in adjusted risk. Among the risk scores evaluated, the Parsonnet score and STS-PROMM demonstrated the best discriminative ability for mortality or major morbidity in elderly patients [78].

Another important study [69] defined frailty as any impairment in activities of daily living, ambulation, or a documented history of dementia. Of 3826 elderly patients, 157 (4.1 %) were frail. Frail patients were older, were more likely to be female, and had risk factors for adverse surgical outcomes. By logistic regression, frailty was an independent predictor of in-hospital mortality (odds ratio 1.8; 95 % CI: 1.1–3.0), as well as institutional discharge (odds ratio 6.3; 95 % CI: 4.2–9.4) and of reduced midterm survival (hazard ratio 1.5; 95 % CI: 1.1–2.2). Sundermann and coll. [79] tested a more complex evaluation tool, called Comprehensive Assessment of Frailty (CAF), including unintentional weight loss, weakness (handgrip at dynamometer), self-reported exhaustion (through a specific questionnaire), slowness of gait speed (time to walk 4 m), low activity (IADL), evaluation of standing balance, a test to assess body control, laboratory tests (serum albumin, creatinine, and brain natriuretic peptide), and measure of forced expiratory volume in 1 s (FEV1). This protocol was applied to 400 patients ≥74 years. Median frailty score was 11; median of logistic EuroSCORE was 8.5 % and of STS score was 3.3 %. There was a significant correlation between frailty score and observed 30-day mortality (P < 0.05). The evident limit of this study was the complexity of the evaluation, which makes its application in the real world unrealistic. Fortunately, the same study group simplified the assessment and created the “Frailty predicts death One yeaR after Elective Cardiac Surgery Test (FORECAST)” [80], limited to the more predictive variables. These were chair rise (the patient gets up and down from a chair three times, and time is measured); weakness (in the last 2 weeks); stair (patient climbs as many stairs as he is able to); Clinical Frailty Scale; and creatinine. At a 1-year follow-up, mortality rate was 12.2 %. Patients who died within 1 year had a median frailty score of 16 compared to 11 in survivors (P = 0.001).

A further clinical implication of multidimensional geriatric evaluation is the ability to identify patients at high risk of developing delirium during hospitalization. Rudolph et al. [81] recently validated a scoring system to predict delirium after cardiac surgery, as defined by the Confusion Assessment Method [82]. They enrolled 122 elderly cardiac surgery patients that underwent a delirium assessment pre- and postoperatively beginning on postoperative day 2. Delirium occurred in 52 % of the patients, and multivariate analysis identified four independent variables associated with delirium: previous stroke, MMSE scores, abnormal serum albumin, and the Geriatric Depression Scale Scores.

Solid evidence from randomized controlled trials demonstrates that a geriatric intervention guided by CGA has positive effects on health, functional status, and mortality in hospitalized elderly patients [83]. In the last years, an increasing number of studies have assessed whether the CGA was able to increase our accuracy in predicting the prognosis of elderly subjects candidate to intervention. In many fields of surgery and medicine, trials have been performed to estimate whether malnutrition, cognitive impairment, disability, or poor social resources impact on postoperative mortality, length of stay, and institutionalization. A Japanese study [47], involving 120 patients with mean age of 70.3 undergoing thoracic surgery, showed that patients with a MMSE ≤24 developed postoperative complications, including delirium, four times more often than patients >24, also. Functional dependency estimated by the Barthel Index also correlated strongly with postoperative complications. Furthermore, patients with malnutrition (defined as low levels of albumin, transferrin, and peripheral lymphocyte) were more likely to suffer postoperative complications. The authors concluded that intensive care of elderly patients after thoracic surgery might be improved by CGA.

Of particular interest is a study by a French group focusing on lung oncology, showing that a preoperative multidimensional geriatric assessment, targeted to measure global functional status and performed by a team of geriatricians in collaboration with anesthesiologists and surgeons, was able to reduce perioperative mortality as well as risk of institutionalization at discharge [84]. Likewise, in elderly patients undergoing elective surgery for colorectal cancer [85], a standardized CGA was able to independently predict severe complications (potentially life-threatening without or with lasting disability or fatal). Based on the preoperative CGA assessment, patients were categorized into three groups: “fit,” “intermediate,” and “frail”; frail were those patients who were dependent in personal activities of daily living or had a pathological score at some test (MNA < 17; CIRS >4; GDS > 13; MMSE < 24) or took more than seven medications a day. Extermann and Hurria reviewed studies of CGA in older adults and concluded that CGA predicted survival, chemotherapy toxicity, postoperative morbidity, and mortality in cancer patients [86].

An attractive study extended the concept of CGA applied to oncological patients, creating the Preoperative Assessment of Cancer in the Elderly (PACE) [87]. This includes, besides CGA, the evaluation of fatigue with Brief Fatigue Inventory (BFI), physical performance with Eastern Cooperative Oncology Group Performance Status (ECOG-PS), and a measure of the degree of sickness or physical state prior to anesthesia and surgery according to ASA grade. This international multicenter prospective study, based on a cohort of 460 patients with a mean age of 76.8 years candidate to breast, genitourinary, or gastro-intestinal elective surgery, demonstrated that moderate/severe fatigue, dependence in IADLs, and impaired PS were the most important independent predictors of postoperative complications [88]. Furthermore, dependence in ADLs and IADLs and impaired PS was associated with increased length of stay [88].

A study by Richter et al. [89] is partly in disagreement with these results. In a cohort of 62 patients >60 years of age, scheduled to undergo pelvic floor surgery, 32 received additional preoperative assessment of function (BADL/IADL); time up and go; clock-drawing test, MNA, and GDS; and social support scale, compared with 30 patients receiving usual care. There were no differences in primary outcome (scores of the Physical Component Summary and Mental Component Summary of the Medical Outcomes Study Short Form 36 Health Survey), but almost all the patients were “fit”: 87.5 % had no deficiencies in BADL/IADL and 87.5 % performed the Get Up and Go test in less than 20 s; almost had no malnutrition, completed the Clock Test correctly and scored perfectly on the social support scale. Thus, the cohort enrolled in the study appears scarcely representative of the real world.