Number of octogenarian surgical procedures by calendar year.
Mortality rate by calendar year.
Johnson WM, Smith JM, Woods SE, Hendy MP, Hiratzka LF. Cardiac Surgery in OctogenariansDoes Age Alone Influence Outcomes?. Arch Surg. 2005;140(11):1089-1093. doi:10.1001/archsurg.140.11.1089
Outcome differences in octogenarians vs patients younger than 80 years undergoing coronary artery bypass grafting or valve surgery can be analyzed to isolate the effect of age alone on morbidity and mortality.
Eight-year hospitalization cohort study. Physicians, nurses, and perfusionists prospectively collected data on 225 variables.
A consecutive sample of 7726 patients undergoing coronary artery bypass grafting or valve surgery between October 1, 1993, and February 28, 2001.
Main Outcome Measures
There were 9 main outcomes of interest: mortality, length of hospital stay, gastrointestinal tract complications, neurologic complications, pulmonary complications, renal complications, return to intensive care unit, intraoperative complications, and reoperation to treat bleeding. We controlled for 16 potential confounding variables to isolate outcome differences according to age.
Of 7726 patients who fit the inclusion criteria, 522 were octogenarians. Compared with nonoctogenarians, octogenarians had a significantly higher New York Heart Association functional classification, higher incidence of hypertension, and underwent a greater number of coronary artery bypass grafting plus valve surgical procedures (P <.05). They also had significantly lower body surface area, fewer total number of grafts used, less history of tobacco use, and less abnormal left ventricular hypertrophy, and there were fewer nonwhite patients and fewer men. At multivariate analysis, octogenarians had a higher risk for death (relative risk [RR], 1.72; 95% confidence interval [CI], 1.52-1.83), longer hospital stay (RR, 1.03; 95% CI, 1.01-1.04), more neurologic complications (RR, 1.51; 95% CI, 1.26-1.67), and were more likely to undergo a reoperation to treat bleeding (RR, 1.49; 95% CI, 1.09-1.72). Univariate analyses revealed no difference between octogenarians and nonoctogenarians for diabetes mellitus, urgency of procedure, prior myocardial infarction, time since last myocardial infarction, cerebrovascular history, chronic obstructive pulmonary disease, or pump time.
Age alone has been shown to influence outcomes after cardiac bypass or valve surgery. Octogenarians undergoing cardiac surgery have more comorbidities and higher mortality even after controlling for 16 potential confounding variables, compared with nonoctogenarians.
Open heart surgery for coronary artery bypass grafting (CABG) or valve replacement in octogenarians has risen dramatically since 1980, increasing by 67% from 1987 to 1990.1 In 2000 the average life expectancy at birth for the US population was 76.9 years.2 According to the Administration on Aging, there were 4.2 million Americans aged 85 years or older in 2000, and this number is projected to increase to 8.9 million by 2030.3 Approximately 40% of all octogenarians have symptomatic cardiovascular disease, including 18% with ischemic heart disease.4 Increasingly, elderly patients with ischemic heart disease are being referred for coronary artery revascularization by surgical and percutaneous means. However, these strategies are being questioned because of reports of poor outcomes in the elderly.
Previous observational studies have shown that octogenarians undergoing open heart surgery for CABG or valve replacement are at higher risk for postoperative death.5,6 However, these studies have reported different predictors of mortality. For example, Akins et al5 identified chronic lung disease and congestive heart failure as independent predictors of mortality, while Edmunds et al7 found that preoperative variables predictive of early death include New York Heart Association (NYHA) functional classification IV (ie, the person is unable to carry out any physical activity without discomfort; there are symptoms of cardiac insufficiency at rest; and if any physical activity is undertaken, discomfort is increased) previous myocardial infarction (MI), and emergency operation. Until recently, no large series has determined whether age is an independent risk factor for morbidity and mortality in octogenarians undergoing open heart surgery.
This study compared the characteristics of octogenarians with nonoctogenarians undergoing open heart surgery for CABG or valve replacement. By controlling for differences between these groups, we sought to determine whether being an octogenarian is an independent risk factor for mortality as well as for 9 other measured outcomes.
We conducted a cohort study, using cases and control subjects from an 8-year hospitalization cohort. Inclusion criteria for the cohort were CABG or valve replacement between October 1, 1993, and February 28, 2001, and age older than 18 years. Physicians, nurses, and perfusionists collected data on 225 variables concurrently with admission (Table 1). Data were grouped into demographic, medical history, postoperative, perfusion, and procedure sections. Using a series of cross-checking questions, 2 individuals audited all data forms for completeness and consistency. In addition, a physician audited a random 10% of patient forms for accuracy and consistency on an ongoing basis. Data were entered into an interactive multi-institutional database (Patient Analysis and Tracking System; Axis Clinical Systems, Portland, Ore).
Patients were grouped according to age. Patients (n = 522) were aged 80 to 89 years (octogenarians), and control subjects (n = 7204) were 18 to 79 (nonoctogenarian). We controlled for the following 16 potential confounding variables: race, sex, diabetes mellitus, body surface area, urgency of surgical procedure, hypertension, NYHA functional classification, previous MI, time since last MI, cerebrovascular history, tobacco history, chronic obstructive pulmonary disease, pump time, left ventricular hypertrophy, total number of grafts used, and surgery type. We were interested in the following 9 outcomes: mortality, length of hospital stay, gastrointestinal tract complications, neurologic complications, pulmonary complications, renal complications, return to the intensive care unit, intraoperative complications, and reoperation to treat bleeding.
We performed univariate analysis using χ2 and t tests to compare cases and controls with each of the 16 potential confounding variables (Table 2). To generate the unadjusted risks of each outcome, we performed χ2 and t tests comparing cases and controls with each of the 9 outcomes of interest (Table 3). Additional analyses were conducted to compare cases and controls with the total number of grafts used, both venous and arterial, and specifically the number of arterial grafts used. Using logistic regression analysis for dichotomous variables and linear regression for continuous variables, we then determined the adjusted risk between case patients and control subjects with each of the 9 outcomes of interest while controlling for the 16 confounding variables (Table 4). Analysis was performed with Stata statistical software (Stata Corp, College Station, Tex).
Overall, 7726 patients met the inclusion criteria. Of these, 5011 (65%) were men and 2715 (35%) were women. Among the 522 octogenarians, 266 (51%) were men and 256 (49%) were women. Among the 7204 nonoctogenarians, 4745 (66%) were men and 2459 (34%) were women; their mean ± SD age was 63.3 ± 10.5 years. Figure 1 shows the number of octogenarians in this cohort undergoing surgery in each calendar year.
There was no significant difference between cases and controls for diabetes mellitus, urgency of surgical procedure, previous MI, time since last MI, cerebrovascular history, chronic obstructive pulmonary disease, or pump time. Octogenarians had significantly higher NYHA functional class, incidence of hypertension, and greater number of CABGs plus valve surgical procedures (P<.05); this group also had significantly fewer nonwhite patients and fewer men (P<.05). Furthermore, octogenarians had significantly less body surface area and tobacco use history, fewer instances of abnormal left ventricular hypertrophy, and fewer total number of grafts used (P<.05). Specifically, octogenarians averaged significantly fewer (mean ± SD, 0.7 ± 0.7) arterial grafts used than did nonoctogenarians (mean ± SD, 0.8 ± 0.7; P<.001).
Multivariate analysis revealed no significant difference between octogenarians and nonoctogenarians for gastrointestinal tract complications, pulmonary complications, renal complications, return to the intensive care unit, or intraoperative complications. However, octogenarians were found to have a higher risk for death (relative risk [RR], 1.72; 95% confidence interval [CI], 1.52-1.83), longer length of hospitalization (RR, 1.03; 95% CI, 1.01-1.04), and a higher risk for neurologic complications (RR, 1.51; 95% CI, 1.26-1.67); they were also more likely to undergo reoperation to treat bleeding (RR, 1.49; 95% CI, 1.09-1.72; Table 4). Figure 2 shows the mortality each year in octogenarians and nonoctogenarians.
Open heart surgery in octogenarians has risen steadily since the 1980s. A number of factors justify this increase. Cane et al8 have shown actuarial survival for octogenarians undergoing CABG that is comparable to that of the age-matched population. They concluded that octogenarians should be offered the opportunity for CABG “with the expectation of reasonable results and late survival that parallels their demographic group.”8(p1037) Other investigators have noted that octogenarians enjoy a higher quality of life after undergoing CABG or valve surgery. For example, in their review of 68 octogenarians undergoing CABG or valve surgery, Kumar et al9 found that approximately 85% of patients reported that, in retrospect, they definitely would have made the decision to undergo open heart surgery. This population also had postoperative improvement in NYHA functional class.
In a trial comparing invasive vs medical therapy in elderly patients with chronic symptomatic coronary artery disease, investigators found that those older than 75 years benefit more from revascularization than from medical therapy.10 Their findings showed a significant reduction in major adverse cardiac events with revascularization; although patients had an immediate higher mortality, they subsequently had improved short-term survival. Similarly, Graham et al11 found revascularization to be superior to optimal medical treatment: patients aged 80 years and older undergoing CABG had a 4-year survival rate of 77.4% compared with 60.3% in patients treated medically. In particular, these researchers noted improved 1-year survival in the subset of patients with left main coronary artery disease undergoing revascularization.
Because of ongoing evidence supporting revascularization in patients aged 80 years or older with coronary artery disease, it was our intent to evaluate the differences between octogenarians and nonoctogenarians undergoing CABG or valve replacement. In our experience, octogenarians had significantly higher NYHA functional class, prevalence of hypertension, and less body surface area. Octogenarians also had fewer instances of abnormal left ventricular hypertrophy and averaged fewer total number of grafts used than did nonoctogenarians. Although similar predictors of postoperative morbidity and mortality have been reported by some centers, no physiologic variables have been found to be universally predictive. Like other investigators, we found that octogenarians were at higher risk for postoperative death.5,6,12 However, our results show that after controlling for differences between octogenarians and nonoctogenarians, age is an independent risk factor for morbidity and mortality.
Our finding raises some interesting questions. As previously reported, chronologic and physiologic age may not always match.12 However, it is logical to assume that a person’s overall health is a summation of his or her physiologic status. It would be reasonable to suggest that in our study we may have failed to recognize a physiologic variable predictive of postoperative outcome. Although Akins et al5 did not show statistical significance, they noted improved survival in patients undergoing CABG when at least 1 internal mammary artery graft was used rather than only venous grafts. In our analysis, we did not include the type of graft used in calculating outcome differences. Certainly other physiologic variables that have not become manifest could explain outcome differences between the 2 age groups; alternatively, intangible characteristics might enable an octogenarian to do well after open heart surgery. Nevertheless, after controlling for known physiologic differences between the 2 groups in our study, age alone is clearly a predictor of postoperative morbidity and mortality.
As the younger population ages and life expectancy rises, an estimated 12 million Americans will swell the ranks of octogenarians by the year 2010.13 An increasing number of octogenarians undergoing coronary revascularization or valve surgery are certain to strain an already burdened health care system. It is, therefore, incumbent on researchers to develop more refined algorithms to predict postoperative outcomes. As this study indicates, age should be considered a component of any such algorithm.
Correspondence: Mary Pat Hendy, BS, E. Kenneth Hatton Institute for Research and Education, Room 11-J, Good Samaritan Hospital, 375 Dixmyth Ave, Cincinnati, OH 45220 (firstname.lastname@example.org).
Accepted for Publication: December 22, 2004.
Previous Presentation: Presented at the 68th Annual International Scientific Assembly of the American College of Chest Physicians; November 5, 2002; San Diego, Calif.
Acknowledgment: We are grateful to Debra K. Hiratzka, clinical database manager, and Cardiovascular and Thoracic Surgeons, Inc, for the use of their data set for this study. We also thank Two Herons Consulting, Oxford, Ohio, for editorial review of the manuscript.