As the population ages, an increasing number of elderly patients become candidates for major surgery. According to the National Center for Health Statistics, the rate of hospital discharges for hip replacement in elderly people rose from 2.5 per 10 000 in 1970 to 72.5 per 10 000 in 2004 (http://www.cdc.gov/nchs/agingact.htm). Based on current rates of surgery and census projections, the number of elderly patients undergoing oncologic procedures is expected to increase by up to 51% by the year 2020.

Surgical risks in general are declining over time, and many assume that surgery for elderly patients is getting safer. Case series suggest that surgery can be performed with low morbidity and mortality in elderly people. For lung, esophageal, and pancreatic resection, single-center studies report operative mortality rates between 3% and 4% in very elderly patients. Case series are advocating bariatric surgery in obese patients older than 65 years, touting low morbidity and mortality rates.

Results from clinical series, however, may lead to unrealistic expectations about the safety of surgery in the elderly population, as many published studies have selection bias. Reports of operative mortality tend to represent experiences of high-volume, tertiary academic centers. It is well documented that for high-risk cancer surgery, high-volume surgeons and hospitals have superior outcomes. In addition, results from case series are more likely to be submitted and published if the observed mortality is low, resulting in a publication bias toward lower operative mortality. As a result, existing data generally yield unrealistic risk estimates.

This chapter examines operative risk in very elderly patients using “real-world” results from large databases. In addition, evidence will be reviewed about predictors of surgical outcomes in elderly patients—fatal surgical outcomes as well as functional status and independence after major surgery.

Age

In general, increasing age is associated with increased rates of morbidity, functional dependence, and mortality after surgery. Countless studies for several decades have documented that age predicts adverse outcomes after surgery. In a comprehensive analysis of more than 50 0000 patients undergoing noncardiac surgery in the Veterans Affairs National Surgical Quality Improvement Project (NSQIP) database, Hamel et al. found that age was an independent predictor of operative mortality, after adjusting for comorbidities. For all operations in aggregate, 30-day mortality was 2.8% for patients younger than 80 years compared with 8.2% in older patients. For colectomy, mortality was 6.4% in younger patients versus 11.9% in older patients. Similarly, younger patients undergoing elective hip replacement had much lower operative risk (1.3%) when compared to older patients (6.8%). In a single-center study, patients aged 80 years and older had a fourfold increase in risk of both operative mortality and discharge to nursing home after cardiac surgery. This finding persisted after adjustment for patient characteristics. Age is also a known independent predictor of deterioration in cognitive function after coronary artery bypass graft (CABG) surgery.

To assess the impact of age on mortality for a wide range of procedures frequently performed in the elderly, national mortality rates were examined using the Nationwide Inpatient Sample. The Nationwide Inpatient Sample is a large national database containing hospital discharge data for all payers. It is a sample of all U.S. nonfederal hospitals and contains data from approximately 1000 hospitals in 22 states. Selected hospitals represent five strata of hospital characteristics: ownership control, bed size, teaching status, rural–urban location, and geographic region. Based on analysis of all discharges for CABG, mitral and aortic valve surgery, hip replacement, colectomy, and cystectomy in 2003 using ICD-9 procedure codes, operative mortality increased substantially with age for all six procedures (Figure 37-1). Operative mortality was highest for mitral valve replacement, ranging from 7% for patients younger than 65 years to 21% among patients aged 85 to 99 years. For CABG, aortic valve replacement, colon resection, and cystectomy, operative mortality rates among patients older than 85 years ranged from 7% to 11%. Operative mortality was low after hip replacement (0.2%–3%).

Comorbidity

The presence of medical comorbidities increases operative risk in all age groups. In general, with age, patients accumulate medical conditions. It is unclear whether age—as a risk factor for adverse outcomes after surgery—is primarily a proxy for increased comorbidity burden. The relative importance of age and medical comorbidities has been examined carefully in vascular surgery. While age is an independent predictor of mortality in vascular surgery, in multivariate analysis, cardiac, pulmonary, and renal dysfunction have a greater impact on mortality than age for abdominal aortic aneurysm repair, lower extremity bypass, and amputation. In an analysis of noncardiac procedures, Hamel found that risks associated with selected comorbidities were the same in patients younger or older than 80 years. While it is well known that both age and comorbidity impact operative risk, the interaction between age and comorbidity in surgical outcomes has not been fully elucidated.

Functional Status and Cognitive Function

There is limited evidence that functional status is an important independent predictor of morbidity and mortality after surgery. In a comprehensive analysis of surgical outcomes after abdominal surgery in 24 patients older than 80 years, including physiologic, biochemical, and functional characteristics as well as surgical process measures, age, and activities of daily living score were two of three independent predictors of operative mortality. For gastrectomy, patients with partially dependent functional status are at higher risk of operative mortality than fully independent patients. Poor preoperative functional status has also been shown to predict neurologic and cardiac complications after major surgery in elderly patients.

Although cognitive impairment is common in older adults, the influence of cognitive function on clinical and functional outcomes of surgery in this population is largely unexplored. In one study, surgical patients with cognitive impairment were at higher risk of mortality than other general surgery patients (9.6% vs. 6.3%).

Frailty

Frailty is a common clinical syndrome in elderly patients characterized by decreased physiological reserve and vulnerability to stressors (see Chapter 52). The phenotype includes unintentional weight loss, self-reported exhaustion, grip weakness, slow walking speed, and low physical energy. Frailty appears to be an independent predictor of surgical complications in elderly patients. In one study, frailty was highly associated with postsurgical complications, while ASA score and Lee’s Revised Cardiac Index score were not.

Functional Status Outcomes

For frail older adults undergoing surgery, functional status is an essential outcome measure. Small changes in functional status after surgery are likely to have an important impact on the ability of such patients to care for themselves and live independently. In a study of 672 elderly patients admitted to the intensive care unit after general, vascular, or trauma surgery, activities of daily living scores decreased significantly at late follow-up, and 13% of patients lost their ability to live independently. For cardiac surgery, functional declines are even more dramatic. In a study of 191 patients aged 80 years and older, undergoing elective cardiac surgery, only 64% of late survivors were fully autonomous. Only the occurrence of a postoperative complication was a significant predictor of impaired postoperative autonomy.

Postoperative functional impairment is common and often protracted in elderly patients. In one study, elderly patients undergoing major open abdominal surgery had comprehensive postoperative functional assessments at 1, 2, and 6 weeks and 3 and 6 months. Mean functional recovery time ranged from 3 weeks for Folstein mini-mental state examination to 6 months for instrumental activities of daily living (Table 37-1). A substantial proportion of elderly patients did not return to baseline function after 6 months for instrumental activities of daily living, Medical Outcomes Study Short Form-36 Physical Component Scale, timed walk, and functional reach. In this study, preoperative physical conditioning and depression were important predictors of functional recovery after surgery.