Click here to read this article as a CME activity
Ann Thorac Surg 2008;85:1974-1979. doi:10.1016/j.athoracsur.2008.01.042
© 2008 The Society of Thoracic Surgeons
Original Articles: Adult Cardiac
Propensity Matched Comparison of Outcomes in Older and Younger Patients After Coronary Artery Bypass Graft Surgery
Negmeldeen F. Mamoun, MDa,
Meng Xu, MSb,
Daniel I. Sessler, MDc,
Joseph F. Sabik, MDd,
C. Allen Bashour, MDa,c,*
a Department of Cardiothoracic Anesthesia, Cleveland Clinic, Cleveland, Ohio
c Department of Outcomes Research, Division of Anesthesiology, Critical Care Medicine and Comprehensive Pain Management, Cleveland Clinic, Cleveland, Ohio
b Department of Quantitative Health Sciences, Lerner Research Institute, Cleveland Clinic, Cleveland, Ohio
d Department of Cardiothoracic Surgery, Cleveland Clinic, Cleveland, Ohio
Accepted for publication January 11, 2008.
* Address correspondence to Dr Bashour, Cleveland Clinic, 9500 Euclid Ave, Mail Code G58, Cleveland, OH 44195 (Email: bashoua{at}ccf.org).
| Cardiothoracic anesthesiology:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract
|
|---|
Background: By 2050, the number of people in the United States aged 85 years and older is expected to increase from 4.2 to 20.8 million. We therefore compared outcomes after isolated coronary artery bypass grafting (CABG) in patients aged 85 years and older with those 55 to 65 years old.
Methods: Propensity matching and multivariable analysis were used to compare morbidity and mortality outcomes in 132 patients aged 85 years and older and in 5243 patients aged 55 to 65 years who underwent CABG between 1993 and 2004. Variables of interest were compared using Student t, Wilcoxon rank sum, 
2, or Fischer exact tests, as appropriate.
Results: After propensity matching, each group had 122 patients who were matched on all baseline and intraoperative variables except number of intraoperative red blood cell (RBC) transfusions, which was greater in the older group. Atrial arrhythmias occurred more frequently and intensive care unit (ICU) and hospital lengths of stay (LOS) were longer in the older group. Multiple logistic regression analysis demonstrated that atrial arrhythmias and catheter-related sepsis were more common in the older group.
Conclusions: Older patients received more RBC transfusions, presumably reflecting a lower threshold to transfuse older patients. Atrial arrhythmias were more common in the older group, this can account for their longer ICU and hospital LOS, which can also explain their greater incidence of catheter sepsis. The two age groups had similar mortality and morbidity. Advanced age is not associated with substantively worse outcome after CABG compared with a propensity-matched younger age group.
By the year 2050, people 85 years and older will comprise 5% of the total domestic population compared with 1.5% in 2000 [1]. United States residents who now live to be 85 years old have an overall additional life expectancy of 6.6 years (5.9 years in men and 7 years in women), and their life expectancy will likely continue to increase [2]. The average annual rate of first major cardiovascular event increases almost tenfold (from 7 to 68 events per 1000 men) as their age range increases from 35 to 44 to 85 to 94 years. Comparable rates occur in women 10 years older, and the gender gap narrows with advancing age [3].
Although rapid development of catheter-based treatments for cardiac and vascular diseases will likely reduce the need for a cardiac operation at any age, a substantial number of older patients will still undergo cardiac operations, even if only after less invasive procedures are no longer an option. It thus seems likely that the average age of patients undergoing cardiac operations will increase as the population ages.
Several investigations have evaluated outcomes after coronary artery bypass graft (CABG) procedures in older patients. Most report older patients have increased morbidity or mortality, or both but that CABG procedures can be performed in these patients with acceptable results [4–9]. Age-outcome studies cannot be randomized, they depend on collected data, and are typically observational without any interventions implemented by randomized assignment rules. Standard methods of multivariable analysis such as linear or logistic regression are limited because they cannot completely eliminate selection factors or provide an accurate assessment of the effect of the variable of interest being compared after properly adjusting for patient differences [10]. Propensity score matching is a more reliable method of evaluating large observational data sets because outcome differences between patients who receive different treatments but have similar balancing scores provides an unbiased estimate of the effect attributable to the variable of interest being compared [10, 11]. Propensity matching thus changes "apples to oranges" comparisons, as with multivariable analysis, to "apples to apples" comparisons [10].
We used propensity matching in this investigation to test the hypothesis that patients aged 85 years and older who undergo CABG have greater morbidity and mortality than those 55 to 65 years old.
|
Material and Methods
|
|---|
After approval from the Institutional Review Board and waiver of the need for patient consent, information was obtained from the Department of Cardiothoracic Anesthesia Patient Registry for patients who underwent isolated CABG between January 1, 1993, and June 30, 2004, at our institution. During this period, 132 patients 85 years and older, and 5262 patients aged 55 to 65 underwent isolated CABG. The database contained two CABG procedures for 19 patients from the younger age group and the second record was excluded. The 132 patients who underwent isolated CABG in the older group and the 5243 in the younger group were included, and no other exclusion criteria were applied.
Perioperative and postoperative variables in the two groups of patients were compared, and a propensity score was calculated for each patient from a logistic model that included preoperative and intraoperative variables (Tables 1 and 2).
Patients were matched one-to-one on propensity scores with greedy matching techniques [12]. Outcomes of interest between the matched groups were compared using the Student t or Wilcoxon rank sum tests for continuous variables, and 2 or Fisher exact tests for categoric variables. Results are presented as median (25%, 75% percentile) or frequency and percentage; a value of p < 0.05 was considered statistically significant. Multiple logistic regression analysis with backward variable selection was also used to assess the effect of age on outcome after adjusting for other significant factors. All results were analyzed with SAS 8.2 software (SAS Institute Inc, Cary, NC).
The incidence of postoperative atrial arrhythmias was evaluated using information from both the Department of Cardiothoracic Anesthesia Patient Registry, which tracks postoperative atrial arrhythmias in the intensive care unit (ICU) that required treatment (atrial fibrillation, atrial flutter, supraventricular tachycardia, and premature atrial contractions), and the Department of Cardiothoracic Surgery Cardiovascular Information Registry, which tracks postoperative atrial arrhythmias (atrial fibrillation and atrial flutter) during the entire postoperative hospital stay from ICU admission to hospital discharge.
Because this investigation included patients from the period that exceeded 11.5 years, the year of operation may have affected outcome because surgical techniques and perioperative care presumably improve over time. We did not determine the effect of operation year on outcome; however, the distribution of patients per operation year for both groups was analyzed.
|
Results
|
|---|
Each age group had 122 patients after propensity matching, thus successfully matching 92% of the smaller older group. The median ages of the two groups were 60 and 87 years. The patients were matched on all preoperative and intraoperative variables except intraoperative transfused red blood cells (RBC), with a median (25%, 75% percentile) RBC administered of 1.5 (0.0, 3.0) U in the older group, and 0.0 (0.0, 2.0) U in the younger group (p = 0.028; Tables 1 and 2).
When binary outcome variables for the two matched groups were compared, the respective incidences of postoperative atrial arrhythmias in the ICU only and at anytime during the entire postoperative hospital stay were greater in the older group, at 23.8% vs 10.7% (p = 0.007) and 60.7% vs 28.1% (p < 0.001; Table 3). The two groups had similar risks of major cardiac, neurologic, pulmonary, renal, gastrointestinal, hepatic, and infectious complications, as well as hospital mortality. Among the continuous variables, median initial and total lengths of stay (LOS) in the ICU (46.7 vs 28.3 hours, and 48.9 vs 29 hours, respectively, p = 0.002), and hospital (12 vs 9 days, p < 0.001) were greater in the older group (Table 4).
Before propensity matching, the two patient groups differed on incidence of bacteremia, catheter-related sepsis, pneumonia, initial intubation time of 72 hours or longer, serious infection morbidity, and mortality. Because each matched group in this investigation had only 122 patients, the incidences of these relatively rare complications are low in both groups. Therefore, multiple logistic regression analysis using all data was performed to confirm or contradict the propensity matching results. This analysis showed that the two groups differed on the incidence of postoperative atrial arrhythmias in the ICU that required treatment and catheter-related sepsis; both were greater in the older group after adjusting for other significant factors (p = 0.002, and p = 0.019 respectively; Table 5). Patients in both age groups in whom postoperative atrial arrhythmias developed while they were in the ICU had longer ICU and total hospital LOS compared with patients who did not have atrial arrhythmias (70.5 hours vs 29.1 hours, and 16 days vs 9 days, respectively; p 
0.001; Table 6).
The distribution of patients per operation year over the investigation period did not differ between the two matched groups (p = 0.07, Appendix A).
|
Comment
|
|---|
The most striking finding in this investigation is that the propensity-matched age groups did not differ significantly in major cardiac, neurologic, pulmonary, renal, gastrointestinal, hepatic, and infectious complications, or in hospital mortality. Patients in the two age groups were successfully matched on all baseline and intraoperative variables except intraoperative RBC transfusion. Transfusions were more common in the older age group, presumably because lower transfusion thresholds were applied to the older patients. This presumption could not be tested, however, because hemoglobin concentration values at the time of RBC transfusion were not recorded in the registry. Several studies suggest that RBC transfusions are associated with worse morbidity and mortality outcomes in patients undergoing cardiac operations [13–15]. Information from these investigations would lead to an expectation of worse outcome in the older group; however, the incidence of major morbidities and mortality in the older group was similar to the younger group even though the older group received more RBC transfusions.
Postoperative atrial arrhythmia was more common in the older patients in this investigation. This is consistent with the fact that atrial arrhythmias increase with age [16–20], possibly because of age-related atrial enlargement that increases the number and length of potential reentry pathways [21]. The peak onset of atrial fibrillation after cardiac surgery is on postoperative day 2 [22–23]. Because the median initial ICU LOS in both groups was less than 2 days, the incidence of postoperative atrial arrhythmia was evaluated using information from both the Department of Cardiothoracic Anesthesia Patient Registry, which tracks postoperative atrial arrhythmia in the ICU only, and from the Department of Cardiothoracic Surgery Cardiovascular Information Registry, which tracks postoperative atrial fibrillation and atrial flutter during the entire postoperative hospital stay.
Patients in both age groups in whom postoperative atrial arrhythmia developed had longer ICU and hospital stays, which suggests that postoperative atrial arrhythmia, and not advanced age per se, account for the increase in LOS. The increased incidence of catheter-related sepsis (based on multiple logistic regression analysis) in the older patient group may be associated with the prolonged ICU stay.
Although most patients who underwent CABG at our institution were aged 65 to 75, an age range of 55 to 65 for the younger group was selected to better separate median age between the groups. The results of this investigation differed from those of Scott and colleagues [24], who conducted a retrospective investigation using multivariable risk factor analysis to compare outcomes after CABG in patients younger and older than 80 years and reported that age 80 years and older was an independent predictor of increased postoperative morbidity and mortality. The results also differed from results reported by Johnson and colleagues [25], who conducted a prospective cohort study that used multivariable risk factor analysis to compare octogenarians with patients aged 19 to 79 years undergoing isolated CABG, isolated valve procedures, or both. They reported that octogenarians had more comorbidities and higher mortality and concluded that age alone is associated with worse outcome after cardiac surgery. An important difference between these investigations and ours is that they attempted to control for possible confounding variables with multivariable risk factor analysis only. Perhaps as a consequence, their older patients had a significantly greater incidence of several preoperative comorbid conditions that may have contributed to worse outcomes in the older group. This is an important limitation, because it is likely that a 19-year-old would have fewer morbidities and lower mortality than an 80-year-old patient.
This investigation is limited by the lack of long-term follow-up. Another limitation is the small number of patients in each propensity-matched group; in this investigation, 92% of the older patient group was successfully matched to obtain 122 patients in each group for the analysis. Although this is a larger patient cohort than reported in many studies, 122 patients provide relatively low statistical power and small outcome differences in low incidence variables might have been missed.
This investigation was conducted at a single institution, which increases bias relative to patient selection, surgical techniques, and perioperative management. Relative morbidity and mortality risk in older cardiac surgery patients may differ at other institutions.
Another possible limitation in this investigation is the effect of year of operation on outcome. Because there was no difference in patient distribution by operation year, the effect of year of operation on outcome should be similar between groups. Finally, this study was observational, as are all others evaluating the effect of age on outcome, because patient age cannot be randomly assigned.
In conclusion, we compared outcomes after isolated CABG in patients aged 85 and older with propensity score matched patients aged 55 to 65. The main outcome difference was a higher incidence of atrial arrhythmia in the older group and, possibly as a consequence, longer ICU and hospital LOS. Major cardiac, neurologic, pulmonary, renal, gastrointestinal, hepatic, and infectious morbidities, as well as hospital mortality, did not differ between the two age groups. Therefore, age 85 years or older should not per se exclude otherwise appropriate patients from consideration for CABG.
|
Appendix
|
|---|
|
Acknowledgments
|
|---|
We are grateful to Marvin Leventhal, MS, Cardiothoracic Anesthesia Patient Registry Manager, and to the Cardiovascular Information Registry Staff for providing us with information for this investigation. This study is supported by internal Cleveland Clinic funds, National Institutes of Health (NIH) Grant GM 061655 (Bethesda, MD), and the Joseph Drown Foundation (Los Angeles, CA).
|
References
|
|---|
- U.S. Census Bureau U.S. interim projections by Age, Sex, Race, and Hispanic OriginReleased March 2004http://www.census.gov/ipc/www/usinterimproj/natprojtab02a.pdfAccessed Jan 8, 2008.
- Arias E. United States life tables, 2003: Life expectancy by age, race, and sex. National vital statistics reports. Vol 54, No. 14. Released April 2006http://www.cdc.gov/nchs/data/nvsr/nvsr54/nvsr54_14.pdfAccessed Jan 8, 2008.
- Heart disease and stroke statistics-2006 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee Circulation 2006;113:e85-e151http://circ.ahajournals.org/cgi/content/short/113/6/e85 2006Accessed Jan 8, 2008.[Free Full Text]
- Sakamoto S, Matsubara J, Matsubara T, et al. Coronary artery bypass grafting in octogenarians Cardiovascular Surgery 2001;9:487-491.[Medline]
- Baskett R, Buth K, Ghali W, et al. Outcomes in octogenarians undergoing coronary artery bypass grafting CMAJ 2005;172:1183-1186.[Abstract/Free Full Text]
- Hirose H, Amano A, Takahashi A. Coronary artery bypass grafting for octogenarians: experience in a private hospital and review of the literature Ann Thorac Cardiovasc Surg 2001;7:282-291.[Medline]
- Miller DJ, Samuels LE, Kaufman MS, Morris RJ, Thomas MP, Brockman SK. Coronary artery bypass surgery in nonagenarians Angiology 1999;50:613-617.[Medline]
- Williams DB, Carrillo RG, Traad EA, et al. Determinants of operative mortality in octogenarians undergoing coronary bypass Ann Thorac Surg 1995;60:1038-1043.[Abstract/Free Full Text]
- Glower DD, Christopher TD, Milano CA, et al. Performance status and outcome after coronary artery bypass grafting in persons aged 80 to 93 years Am J Cardiol 1992;70:567-571.[Medline]
- Blackstone EH. Comparing apples and oranges J Thorac Cardiovasc Surg 2002;123:8-15.[Free Full Text]
- Rubin DB. Estimating causal effects from large data sets using propensity scores Ann Intern Med 1997;127:757-763.[Abstract/Free Full Text]
- Parsons LS. Reducing bias in a propensity score matched-pair sample using greedy matching techniques Proceedings of the Twenty-Sixth Annual SAS Users Group International Conference. Cary, NC: SAS Institute Inc; 2001http://www2.sas.com/proceedings/sugi26/p214–26.pdf 2001Accessed Jan 8, 2008.
- Vamvakas EC, Carven JH. Allogeneic blood transfusion and postoperative duration of mechanical ventilation: effects of red cell supernatant, platelet supernatant, plasma components and total transfused fluid Vox Sang 2002;82:141-149.[Medline]
- Leal-Noval SR, Rincon-Ferrari, MD, Garcia-Curiel A, et al. Transfusion of blood components and postoperative infection in patients undergoing cardiac surgery Chest 2001;119:1461-1468.[Medline]
- Michalopoulos A, Tzelepis G, Dafni U, Geroulanos S. Determinants of hospital mortality after coronary artery bypass grafting Chest 1999;115:1598-1603.[Medline]
- Fuller JA, Adams GG, Buxton B. Atrial fibrillation after coronary artery bypass grafting. Is it a disorder of the elderly?. J Thorac Cardiovasc Surg 1989;97:821-825.[Abstract]
- Almassi GH, Schowalter T, Nicolosi AC, et al. Atrial fibrillation after cardiac surgery: a major morbid event? Ann Surg 1997;226:501-511.[Medline]
- Furberg CD, Psaty BM, Manolio TA, Gardin JM, Smith VE, Rautaharju PM. Prevalence of atrial fibrillation in elderly subjects (the Cardiovascular Health Study) Am J Cardiol 1994;74:236-241.[Medline]
- Hogue Jr CW, Creswell LL, Gutterman DD, Fleisher LA, American College of Chest Physicians Epidemiology, mechanisms, and risks: American College of Chest Physicians guidelines for the prevention and management of postoperative atrial fibrillation after cardiac surgery Chest 2005;128:9S-16S.[Medline]
- Leitch JW, Thomson D, Baird DK, Harris PJ. The importance of age as a predictor of atrial fibrillation and flutter after coronary artery bypass grafting J Thorac Cardiovasc Surg 1990;100:338-342.[Abstract]
- Liu XK, Jahangir A, Terzic A, Gersh BJ, Hammill SC, Shen WK. Age- and sex-related atrial electrophysiologic and structural changes Am J Cardiol 2004;94:373-375.[Medline]
- Mathew JP, Fontes ML, Tudor IC, et al. A multicenter risk index for atrial fibrillation after cardiac surgery JAMA 2004;291:1720-1729.[Abstract/Free Full Text]
- Funk M, Richards SB, Desjardins J, Bebon C, Wilcox H. Incidence, timing, symptoms, and risk factors for atrial fibrillation after cardiac surgery Am J Crit Care 2003;12:424-433.[Abstract/Free Full Text]
- Scott BH, Seifert FC, Grimson R, Glass PS. Octogenarians undergoing coronary artery bypass graft surgery: resource utilization, postoperative mortality, and morbidity J Cardiothorac Vasc Anesth 2005;19:583-588.[Medline]
- Johnson WM, Smith JM, Woods SE, Hendy MP, Hiratzka LF. Cardiac surgery in octogenarians: does age alone influence outcomes? Arch Surg 2005;140:1089-1093.[Abstract/Free Full Text]
Top
Abstract
Introduction
