- Email: [email protected]
Prognosis and Quality of Life After Valve Surgery in Patients Older Than 75 Years
Prognosis and Quality of Life After Valve Surgery in Patients Older Than 75 Years* Oz M. Shapira, MD, FCCP; Ruth M. Kelleher, RN; j ulian Zelingher, MD; Deborah W halen, CCRN;Carmel Fitzgerald, CCRN; Gabriel S. Aldea, M D, FCCP; and Richard ]. Shemin, MD, FCCP
Background: Assessment of quality of life has become an increasingly important aspe ct of the risk-benefit analysis of any therapeutic intervention, particularly in high-risk populations, such as the elderly. Methods: Clinical outcomes of 147 consecutive patients older than 75 years undergoing valve surgery between 1992 and 1995 were reviewe d . Long-term quality of life was assessed using a modified version of the Duke University Me dical Outcomes Study system. Results: Mean age was 79.3±4 years, range was 75 to 89 years, and 67% (99/147) were male. Preoperatively, 128 patients (87%) were in New York Heart Association (NYHA) functional class III/IV. Fifty-two percent (77/147) of the operations were nonelective. Concomitant coronary artery bypass grafting was performed in 69 patients (47%). Thirty-day hospital mortality was 7.5% (11/147). Mean follow-up for 98% (133/136) of hospital survivors was 30±13 months (range, 2 to 55 months). Actuarial survival at 55 months was 71±6%-equivalent to a g eneral age-, race-, and gender-matched population reported in the Life Tables of the US National Center for Health Sta tistics. At the time offollow-up, 112 patie nts (96%) lived at home, 78% (91/116) defined their health between good to excellent, and 81 % (93/114) stated that the operation improved their health status. Ninety-seven percent (112/116) were able to bathe and dress independently, 92% (104/113) could walk at least one block, and 88.5% (100/113) could climb at least one flight of stairs. Moderate to vigorous activities could be performe d b y59.2% (67/113). Overall, at the time of follow-up, 81 % (95/117) were in NYHA class 1/11. Conclu~ion: In a selected patient population, valve surgery in the elderly is associated with acceptable early morbidity and mortality. Long-term survival and quality of life are excellent. These facts strongly support the performance of these procedures in patients older than 75 years. (CHEST 1997; 112:885-94) Key words: elderly; qu ality of life; valve surgery Abbreviations: AVR =aortic valve replacement; CABG =coronary arte1y bypass grafting; CPB =cardiopulmonary bypass; CVA=cerebrovascular accident; DVR=double valve replacement; LVEF= left ventricular eje ction fraction; MI = myocardial infarction; MOS system= Medical Outcomes Study system; MVR=mitral valve replacement; NYHA=New York Heart Association
Jn the early 1990s,
3% ( 7.4) million of the US population were older than 80 years.1 The current life expectancy of octogenarians is inthe range of 6.0
Global theme issue on aging *From the Department of Cardiothoracic Surgery, Boston Medical Center. in abstract fo rm at the 46th Annual r This article was p esented Meeting of the American College of Cardiology, March 16-19, 1997, Anal1eim, Calif. Manuscript received April 8, 1997; revision accepted July 28. Reprint r equests: Oz M. Shap·ira, MD, Department ofCardiothoracic Su·rgery, Boston Medical Center, 88 E Newton St, Boston, MA 0211 8; email: [email protected]. edu
to 8. 1 years, leading to an expected rise in their number t o > 10 million (4.3%) within the next 10 years.l·2 Of these octogenarians, 40% have serious symptomatic heart disease. 2 This tr end is reflected in our institution b y the increased fraction of elderly patients undergoing o pen heart surgery from 1 2 to 22% of total cases over the past decade. Quality of life encompasses the total well-being of the person. Assessment of quality of life includes evaluation of functional capacity, symptoms, and the patient's perception of well-being. 3 In the e ra of managed care, quality of life has become an increasingly important end-point measure, being an integral part of the e valuation of the risk-benefit ratio and CHEST I 11 2 I 4 I OCTOBER, 1997
885
cost-effectiveness of a variety of therapeutic modalities. Advances in anesthetic and surgical techniques, as well as improved perioperative care, have resulted in a significant improvement in short- and long-term clinical outcome after open heart surgery in the elderly.4 - 17 However, very little is reported on the effect of these operations on quality of life. The purpose of this study was to evaluate the clinical outcome of valve surge1y in 14 7 patients older than 75 years undergoing valve surgery, focusing on longterm quality of life.
MATERIALS AND METHODS
(3) Comorbid risk factors included hypertension, diabetes mellitus, prior cerebrovascular disease, peripheral vascular disease, smoking, COPD, chronic renal failure (c reatinine > 2.5), dialysis, prior myocardial infarction, and preoperative use of anticoagulants. (4)Operative data included type of operation, concomitant procedures, type of CPB circuit, CPB and aortic cross-clamp times, use of inotropes and intra-aortic balloon pump, and intraoperative complications. (5) Postoperative complications included 30-day mortality, and significant morbidity was defined as reoperation for bleeding, mediastinal infection, pneumonia, respirator > 3 days, transient ischemic attack and/or cerebrovascular event, myocardial infarction (new Q wave, and/or elevation of creatine phosphokinase MB 2:50 U), low cardiac output (a newly placed intra-aortic balloon pump or the use of inotropes for > 24 h t o maintain a cardiac index > 2.0), valve thrombosis, and other major complications (vascular, GI, e tc) .6 Bleeding and transfu sion require ments were also noted.
Patients
Long-term Follou;-up and Assessment of Quality of Life
The study population consisted of 147 consecutive patients older than 75 years undergoing valve surgery in a tertiary teaching hospital between January 1992 and July 1995.
Two questionnaires, one for patients and one for physicians caring for them, were developed to collect long-term results. Each questionnaire included a brief explanation of the project. The patient's questionnaire (Appendix l ) was based on the Duke University Medical Outcomes Study (MOS ) system. To maximize the response rate, the questionnaire was k pe t as brief as possible, and consisted of three of the six items originally included in the MOS system. The focus was on functional capacity, the patient's perception of well-being, and the impact of surgery on overall health status. The physician questionnaire (Appendix 2) focused on long-term morbidity, mortality, and fun ctional capacity as determined b y NYHA classification. Patients, families, and/or physicians who failed to respond were directly contacted by phone to achieve as complete follow-up as possible.
Surgical Techniques
Operations were performed using cardiopulmonary bypass (CPB) and mild systemic hypothermia (34°C to 35°C). Myocardial protection was achieved using antegrade and retrograde cold W) blood cardioplegia, supplemented by topical cooling with cold saline solution. Aminocaproic acid (Amicar; American Reagent; Shirley, NY) was used routinely in this study. Departmental thresholds for homologous blood transfusion were defined a s hematocrit of 20% durin g CPB and 25% postope ratively. Use of clotting factors was based on clinical assessment of bleeding and he matologic evaluation, s pecifically correcting factor deficits. Persiste nt bleeding in excess of 300 mL in the first hour or 500 mL in the first 2 hwas considered an indication for transfusion of 5 to 10 U of platelets and 2 to 4 U of fresh frozen plasma. The departmental transfusion guidelines w ere rigid and did not change ove r the study period. Data Collection
Hospital and clinical records were re viewed r etrospectively. Data retrieved included the following: (l ) demographic information-age, gender, body surface area, indication for surgery, date of ope ration, surgeon , status of the procedure (elective vs urgent/emergent). We used the Society of Thoracic Surgeons criteria to define the status of the procedure. 18 Elective: an elective operation was one that was performed on a patient with cardiac function that had been stable in days or weeks prior to the operation. Urgent: an urgent operation was one in which surgery was required within 24 h in order to minimi ze the chance of further d eterioration. Delay in operation was necessitated only by attempts to improve the patient's condition, availability of a spouse or a parent for inform ed consent, availability of blood products, or the availability of results of essential laboratory procedures or tests. Emergent: patients requiring emergency operations had ongoing, refractory unrelenting cardiac compromise, with or without hemodynamic instability, and were not responsive to any form of therapy except cardiac surgery. An emergency operation was one in which there was no delay in providing operative intervention 1 8 (2) Preoperative left ventricular ejection fraction (LVEF) and functional status were determined b ythe NewYork Heart Association (NYHA) classification. 886
Statistical Analysis
Statistical analysis was performed using the statistical software (SAS version 6.11; SAS Institute; Cary, NC ). Data are expressed as mean±SD, and for selected variables also by the median and range. Two-tailed Student t test was used to analyze continuous variables. Categorical variables were analyzed u sing x2 with Yates' correction or Fisher's Exact Test when appropriate . A p value of < 0.05 was considered significant. Multivariate analysis using logistic regression was used to identify factors associated with early and late mortality. For the logistic regression analysis, the LOGISTIC procedure was utilized. A backward selection method was used to select the variables in the model, with a significance level for entry into the model of 0.1 and a significance level of 0.05 for staying in the model. Long-term survival was analyzed using the Kaplan-M eier method. The survival of the study group was compared to age-, race-, and gender-matched general US population, obtained from the Life Tables of the US National Center for Health Statistics,19 using the log rank test. All the calculations were performed using a procedure (SAS LIFETEST)
RESULTS
Baseline Characteristics of the Study Group The baseline clinical and operative profiles of the study group are summarized in Table l. There were 99 men and 48 women with a mean age of 79.3::!:::3.8 years (range, 75 to 89 years). Mean LVEF was Clinical Investigations
Table !-Baseline Characteristics of the Study Group Operative Profile
Clinical Profile Variable
n=147
Age, (yr) Male/female Preoperative NYHA class
79.3::+::3.8 (range 75-89 years ) 99/48 3.3 : +: 0.7 0 (0%) 19 (13%) 57 (39%) 71 (48%) 53.9::+::16.1% 40 (27%) 77 (52%) 13 (9%) 80 (54%) 69 (47%) 45 (31%) 32 (22%) 20 (14%) 15 (10%) 13 ( 9%) 6 (4%)
I II III
IV LVEF LVEF < 45% Nonelective Reoperation Hypertension CAD* Cigarette smoking Diabetes mellitus COPD PVD* Prior CVA Renal failure
n = l47
Variable Procedure AVR AVR+ CABG MRV
54 (37%) 51 (35%) 11 (7%) 10 (7%)
MVR + CABG
6 (4%) 3 (2%) 7 (5%) 5 (3%) 69 (47%)
Mitral Valve Repair Mitral valve repair+CABG DVR DVR+CABG Overall concomitant CABG Prosthesis Bioprosthesis Mechanical CPB time, min Aortic cross-clamp time, min
121 (82%) 17 (12%) 119.8::+::45 72.6::+:: 25
*CAD=coronary artety disease; PVD=peripheral vascular disease.
53.9±16.1% with 27% having LVEF of <45%. Mean preoperative NYHA functional class was 3.3±0.7 with 87% being in class III/IV. A significant number of patients had multiple comorbidities such as hypertension, coronary artery disease, smoking, diabetes, prior stroke, and others. Aortic valve replacement (AVR) was performed in 105 patients mostly for senile calcific aortic stenosis; mitral valve repair (n =9) or replacement (n = 21) (MVR ) was performed in 30 patients, one third of whom had ischemic mitral regurgitation; and double valve replacement (DVR) procedures were performed in 12 patients. Concomitant coronary artery bypass grafting (CABG) was performed in 69 patients (47%) . More than 50% of the operations (73/147) were nonelective, and 9% were reoperations. Bioprostheses were implanted in most patients (1211147 [82%] ). There was no significant difference between mechanical valves and bioprostheses with regards to the size of the prosthesis implanted: AVR prosthesis size-mechanical: 20.6±2 mm vs bioprosthesis: 21.6±3 mm (p=0.36); MVR prosthesis sizemechanical: 30.0±4 mm vs bioprosthesis: 28.8±3 (p=0.46). Conventional CPB circuits were used in 91 patients and heparin-bonded circuits were used in 56. Early Mortality and Morbidity Early clinical outcomes are summarized in Table 2. Thirty-day operative mortality was 7.5% (ll/147). The causes of hospital deaths included the following: CVA, two patients; ischemic bowel, two patients;
coagulopathy, two patients; aortic dissection, postinfarction ventricular septal defect, perioperative myocardial infarction (MI ), multiorgan failure , and atrioventricular rupture each in one patient. Patients with bioprostheses had a mortality of 5.8% (7/121), and patients with mechanical valves had a mortality of 23.5% (4/17) , p=0.04. The causes of deaths in patients with mechanical valves included CVA in two patients, atrioventricular rupture , and multiorgan
Table 2-Early Clinical Outcomes Variable
n=147
30-d operative mortality Overall morbidity Morbidity excluding uncomplicated atrial fibrillation Atrial fibrillation/flutter Pneumonia Ventilator support > 3 d Tracheostomy Reoperation fo r bleeding CVA Renal Failure MI Mediastinal infection Total blood transfusion s, donor units Patients transfus ed Hours of ventilatory suppott Surgical ICU length of stay, d Total hospital l ength of stay, d
11 (7.5%) 89 (61 %) 52 (35%) 53 (36%) 9 (6%) 24 (16%) 11 (7.5%) 15 (10%) 7 (4.8%) 3(2%) 1 (0.7%) 0 (0%) 22.3::+:: 26 (median, 15; range, 0-34) 136 (94%) 37.8 : +: 63 (median, 20; range, 6-342) 3.9::+::6 (median, 2; range, 1-49) 10.3::+:: 7 (median , 8; range, 5-52)
CHEST I 112 I 4 I OCTOBER, 1997
887
Table 3-Univariate (Top) and Multivariate (Bottom) Analysis of Early Mortality* Variable
Alive (n =136)
D ead (n = ll )
Nonelective CPB time, min Aortic cross-clamp time , min Intra-aortic balloon pump 24-h chest tube drainage, mL Reoperation for bleeding Total blood donor exposure, U Postoperative CVA Bioprosthesis
68 (50%) 115.5::!:31 71.2::!:23 2 (1.5%) 979::!:720 12 (8.8%) 19.3::!:22 5(3.7%) 114 (83 .8%)
9 (82%) 158.0::!:119 89.8::!:34.2 3 (27.2%) 2078::!:2534 3 (27.2%) 63.3::!:45 2 (18.2%) 7 (63.6%)
Variable Total blood donor exposure, U Postoperative CV A Bioprosthesis
~
Coefficient 0.0499 2.8001 -3.5104
p Value 0.04 0.002 0.015 0.015 0.0004 0.05 <0.00001 0.05 0.077
Odds Ratio (95% CI)
p Value
1.05 (1.02-1.08) 16.45 (1.49-18.58) 0.03 (0.002-0.38)
0.044 0.022 0.007
*Early mortality: 30-day in-hospital deaths. 1 CI=confidence interval.
failure. The causes of death in these patients could not be directly attributed to the use of mechanical valves. Overall morbidity was quite high (61 %), mainly due to high rates of respiratory complications, postoperative atrial fibrillation, and reoperation for bleeding. Excluding uncomplicated episodes of atrial fibrillation overall, morbidity was still significant35% (521147). Postoperative CVA occurred in 4.8% (7/147). Three patients clearly had an intraoperative neurologic event. However, four patients were neurologically intact immediately after the operation, and the neurologic deficit developed 3 to 5 days after surgery. In all four patients, the development of CVA was strongly related to new-onset atrial fibrillation of 2:24-h duration. Of note are the very low rates of perioperative MI and mediastinal infection. Most patients (138/147, 94%) required significant amount of homologous blood transfusions, an average of 22.3:±:26 donors (median, 15 U; range, 0 to 134 U). Blood product utilization was high for each of the components: RBCs: 5.5:±:6 U (median, 4 U; range, 0 to 34 U); fresh frozen plasma: 3.4:±:4 U (median, 2 U; range, 0 to 22 U); platelets: 9.9:±:11 U (median, 6 U; range, 0 to 59 U). Even after exclusion of the 15 patients (10%) who underwent reoperation for bleeding, 93% (123/132) required transfusion of an average of 16.3:±: 17 donors (median, 11 U; range, 0 to 78 U). The relatively high transfusion and complication rates resulted in an average of 37.8:±:63 h (median, 20 h; range, 6 to 432 h) of ventilatory support, 3.9:±:6.5 days (median, 2days; range, 1 to 49 days ) of ICU stay, and total hospital length of stay of 10.3±7.3 days (median, 8 days; range, 5 to 52 days ). These figures are significantly higher than our results following similar operations performed in younger patients (data not shown). Factors associated with 888
increased early mortality in a univariate analysis (Table 3, top) included nonelective operations, increased cardiopulmonary and aortic cross-clamp times, use of intra-aortic balloon pump, increased postoperative mediastinal chest tube output, reoperation for bleeding, increased homologous transfusions, and postoperative CVA. However, in a multiple logistic regression analysis, only postoperative CVA and increased homologous transfusions were identified as independent predictors of increased mortality, and the use of bioprosthetic valve was associated with decreased mortality (Table 3, bottom). There was a trend toward increased mortality for mitral valve procedures compared to aortic valve, but the difference did not reach significance (MVR, 5/30 [16.6%] vs AVR 5/105 [4.8%], vs DVR 1/12 [9.1 %], p==0.09). Performing CABG in addition to valve replacement did not significantly affect early mortality (valve procedure alone 4/74 [5.1 %] vs valve procedure with CABG 7/69 [10.1 %], p==0.15). Long-term Mortality Follow-up was complete for 98% (133/136) of hospital survivors and ranged from 2 to 55 months (mean, 30±13 months ). Fifteen patients (10.2%) died during the follow-up period. The most common cause of long-term mortality was congestive heart failure , 7 of 15 (46.6%). Actuarial survival at 55 months was 71±6%, equivalent to a general age-, race-, and gender-matched population reported in the Life Tables of the US National Center for Health Statistics (Fig 1).1 9 Factors associated with late mortality in a univariate analysis (Table 4, top ) included advanced age, male gender, preoperative history of CVA, postoperative respiratory failure requiring ventilatory support >3 days and tracheostomy, postopClinical Investigations
1.0 0.9 0.8 ii
>
0.7
·s:...
0.6
th
0.5
:::s
...c
Cl)
:;::;
ca
a.
......... Study Group US General Population
I3-B-£J
0.4
4-year Survival 76.9 ± 5.8% Study Group 77.0 ± 6.8% US Population
0.3
p
0.2 0.1 0.0
147
136
120
Number of Patients at Risk 102 82 65 49
0
6
12
18
24
30
36
=0.7594
36
15
42
48
54
Follow Up Months after Surgery FIGURE 1. Kaplan-Meier analysis of survival, comparing the study group to an age-, race-, and gender-matched US general population.
erative CVA, and postoperative atrial fibrillation. In a multiple logistic regression analysis, advanced age, male gender, preoperative and postoperative CVA, and the need for tracheostomy remained significant independent predictors of long-term mortality (Table 4, bottom).
Quality of Life Quality of life results are summarized in Figures 2-4. At the time of follow-up, 112 (96%) lived at
Table 4-Univariate (Top) and Multiple Logistic Regression (Bottom) Analysis of Long-term Mortality* Variable
Alive (n=121)
Dead (n=15)
p Value
Ventilator >3 d Postoperative atrial fibrillation Age, yr Gender, male/female Preoperative CVA Postoperative CVA Tracheostomy
14 (11.6%) 42 (34.7%) 79.0:':4 35/86 8(66%) 2 (1.6%) 5 (4.1% )
7 (46.6%) 9(60%) 81.1:':4 9/6 4 (26.6%) 3(20%) 5 (33.3%)
0.023 0.053 0.039 0.019 0.028 0.009 0.001
J3 Coefficient Odds Ratio (95% CI 1 ) p Value
Age, yr Gender, male/female Preoperative CVA Postoperative CVA Tracheostomy
0.1803 1.6331
1.19 (1.005-1.43) 5.12 (1.32-19.77)
0.044 0.019
2.8956 2.2770 2.3340
18.09 (3.69-88.75) 9.753 (1.52-62.71) 10.32 (1.71-62.33)
0.0004 0.016 0.011
*Long-term follow-up: range, 2-55 months (mean, 30:':13 months). 1 See Table 3 footnote.
home, 78% (91/116) defined their health status between good and excellent (Fig 2), and 81% (93/ 114) stated that the operation improved their health status (Fig 3). Ninety-seven percent (112/116) were able to bathe and dress independently, 92% (104/ 113) could walk at least one block, and 88.5% (100/113) could climb at least one flight of stairs. Moderate to vigorous activities could be performed by 59.2% (67/113) (Fig 4, A [top]) . Most patients who could not perform vigorous activities were limited by concomitant disorders such as CVA, degenerative joint disease, and impaired visual acuity, rather than by their cardiovascular performance. Overall, at the time of follow-up, 81% (95/117) were in NYHA class IIII-a significant change compared to preoperative status (Fig 4, B [bottom]).
Exellent
V.good Good Fair
0
10
20
30
% of patients
FIGURE 2. Health self-perception. Summary of the patients' responses to the first question being asked to define their overall health status at the time of follow-up.
CHEST I 112 I 41 OCTOBER, 1997
889
Same Somewhat worse
Much worse 0
20
10
30
40
50
60 %of Patients
FIGURE 3. Impact of surgery on health status. Summary of the patients' responses to the second question being asked to compare their health status at the time of follow-up to the preoperative status.
DISCUSSION
Evaluation of the short- and long-term results of surgery in the elderly is difficult, since there is no true control group. A randomized trial comparing surgery to medical treatment would be ideal, but it is unlikely to be performed based on the existing data of the natural history of hemodynamically significant valvular heart disease treated medically. 20 -22 Comparison to age-, race-, and gender-matched population of patients with the same cardiac abnormalities
A Bathe and dress Walk one block Climb one flight Moderate to vigorous
•
l
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
0
20
40
60
I
80
100
% of patients
B 70
II
Ill
IV
NYHAclass
FIGURE 4.Functional capacity. A, top: summary of the patients' responses to the third question evaluating their own functional capacity. B, bottom: preoperative vs follow-up NYHA fun ctional class. 890
who were eligible for surgery and refused is another theoretical option, but it would be difficult to perform . Compa1ison to younger patients carries significant limitations as older patients have different cardiac diseases, significantly more comorbidities, and usually are referred later for surgery. 2 0- 22 The most common aortic valve abnormality in the elderly is senile calcific aortic stenosis, whereas young patients are frequently operated on for aortic regurgitation or a mixed lesion. Rheumatic and degenerative mitral valve disease are the most frequent indications for surgery in the young, whereas ischemic mitral regurgitation is much more common in the older age group. 20 -22 Also, concomitant coronary artery disease with the need for myocardial revascularization is more common in the elderly.2 1·22 These trends are reflected in our study group, where 90% of aortic valve replacements were performed in patients with calcific aortic stenosis, one third of mitral valve procedures were performed in patients with ischemic mitral regurgitation, and concomitant CABG was indicated in 47% of patients. The late referral is reflected by the advanced preoperative NYHA functional class (87% in class III/IV), and the fact that > 50% of the operations were nonelective. Considering this very high-risk profile, the observed 7.5% operative mortality seems to be acceptable, and compares favorably with the reported mortality rates in the most recent Society of Thoracic Surgeons national database 23 and previous studies evaluating valve surgery in the elderly, with reported mortality rates of9 to 37%.4 - 17 Most likely, the lower mortality is a r esult of continued improvement in anesthesia, surgical techniques, and perioperative care, but also of patient selection. All the patients operated on during the study period were included in our analysis with no exclusion. However, data were not available on patients who had significant valvular disease but have not been referred for surgery by their primary care physicians and/or cardiologist, or have been rejected by the surgeon. The issue of selection bias is one of the major limitations of a retrospective study. Elderly patients had prolonged ventilatory support and longer ICU and total hospital lengths of stay. These figures reflect not only the high risk of this specific group, but also our practice during the study period. Since than, early extubation and a "fast track" protocol have been aggressively implemented in all of our patients, resulting in a significant reduction of ventilatory support time and lengths of ICU and hospital stays. Similar to previous reports,6,lO,ll mortality following mitral valve surgery was higher compared to aortic valve surgery. This did not reach significance in our study, most likely due to a small sample size. Also, concomitant CABG was not associated with increased mortality, an issue that is still Clinical Investigations
debatable in the literature. 4 •11 · 17 •23· 25 The only independent predictors of early mortality in this series were increased homologous blood transfusions and postoperative CVA. The rate of reoperation for bleeding was relatively high. 23 However, even after exclusion of this subset of patients, blood transfusions were required in >90% of patients. Thus, the high transfusion rates in this age group may be related to other factors such as higher incidence of preoperative anemia, high percentage of women with small body surface area, and higher incidence of postoperative coagulopathy. In the latter part of the study, we switched to heparin-bonded CPB circuits with lower systemic anticoagulation protocol with a dramatic reduction of the rate of bleeding and blood product utilization. 26 Postoperative CVA occurred in 4.8% of our patients, which is somewhat lower than that previously reported in this age population. 4 · 17 In our study, postoperative CVA was related to newonset atrial fibrillation in nvo thirds of the patients. Undoubtedly, efforts aimed at decreasing the incidence of postoperative atrial fibrillation are of prime importance. We have shown previously that strategies such as the use of heparin-bonded circuits effectively reduced the rate of atrial fibrillation in patients undergoing CABG.27 Similar approaches should be studied in this patient population. In this study, the use of bioprostheses was associated with lower early mortality, although the reason is not entirely clear. The superior short- and long-term results with bioprosthetic valves in this age population have been documented previously. 28•29 Analyzing the effect of the type of prosthesis on long-term result was impossible in our study due to small number of patients with mechanical valves. A major finding in this study was that patients who left the hospital had excellent long-term survival, similar to that of an age-, gender-, and race-matched US general population. Of note is that postoperative CVA was the only variable identified as a significant predictor of both early and late mortality. Although occurring in only 4.8%, when it does occur, it is a bad prognostic indicator. Since the long-term survival of these patients is excellent, the issue of quality of life becomes an increasingly important aspect of the evaluation of the risk-benefit ratio and cost-effectiveness. Previous studies evaluating long-term results of valve surgery in the elderly have used survival and NYHA classification as the major end points and have reported improving results. 4 · 17•24 •25 However, other measures such as placement of the patient, health self-percep-
tion, and other important components of overall quality of life assessment have not been investigated previously in this specific patient population. Assessment of quality of life is a complex process because it relies heavily on subjective data, it requires evaluation of many variables, and it is very difficult to quantify. 3 Comprehensive health surveys such as the McMaster Health Index Questionnaire,3 0 the Sickness Impact Profile,31 and others are lengthy instruments that make them clinically impractical. Single-item measures such as the National Health and Nutrition Examination Survef 2 are short and result in an improved response rate but are usually less precise, less reliable, and less valid. The Duke University MOS system,33 which was developed as a compromise benveen these extremes, subsequently has become a very effective tool in evaluating quality of life in a variety of clinical trials. 34· 36 In this study, we used a modified version of the MOS system because of the specific characteristics of the study group, being aware of the lower compliance rate for these type of studies in the elderly. 33 The focus was on functional capacity, the patient's perception of well-being, and the impact of surgery on overall health status. Bodily pain was not assessed as most patients were far out from surgery. Work status was believed to be not applicable as the patients were all retired. The psychosocial status was not included as it is a more complex issue, and we believed that it should be assessed separately. Using this short version, we were able to achieve a very high response rate and obtain a reliable evaluation of long-term quality of life of this patient population. The overall results were very encouraging. Most patients stated that the operation had a significant positive impact on their overall health status, and estimated their health status at the time of follow-up to be benveen good to excellent. Also, the vast majority of patients lived at home, either alone or with family, and were functionally independent. Most patients were able to function at a relatively high level. Those who did not were limited by other comorbidities and not by their cardiovascular performance. In conclusion, valve surgery in the elderly is associated with acceptable early morbidity and mortality. Hospital survivors are expected to have a life expectancy that is equal to that of age-, race-, and gender-matched population. Long-term quality of life in the elderly after valve surgery is excellent. These facts strongly support the performance of valve surgery in patients older than 75 years.
CHEST I 112 I 4 I OCTOBER, 1997
891
BOSTON
UN IVERSITY
SCHOOL
OF
MEDICINE
I
SCHOOL
OF
PUBUC
HEALTH
•
bOSTON
UNIVERSITY
MEDICAL
CENTER
HOSPITAL
•
UNIVERSITY
GOLDMAN
SCHOOL
OF
GRADUATE
OENnSntY
Department of Cardiothoracic Surgery
Dear_____________________________________ The Department of Cardiothoracic Surgery at BUMC is conducting a survey of patients who have undergone valvular replacement surgery at our facility. The purpose of this survey is to assess activity levels of patients after surgery. It would be most appreciated if you would fill out the questionnaire below and return it to us in the envelope provided by Thank you for assisting us with this important project which we hoDe will enable us to better evaluate how valve replacement sur·gery affects quality of life.
Our records show that you received a on 1. In general, would you say your health is: that best describes your health today) Excellent Very good Good Fair
(circle the answer Poor
2. Compared to before your valve surgery, how would you rate your health in general now? (Select one) * Much better than before my surgery * Somewhat better than before my surgery * About the same as before my surgery ----* Somewhat worse than before my surgery ____ * Much worse than before my surgery ____ 3. The following items are about activities that you might do during a typical day. (Please indicate by circling yes (Y) or no (N} whether you are able to participate in these activities:) * Vigorous activities, such as lifting heavy objects, playing tennis: Y N * Moderate activities, such as moving a table, pushing a vacuum cleaner, bowling, or playing golf: Y N * Climbing two (2) flights of stairs: Y N Y N * Climbing one (1) flight of stairs : * Walking two (2} blocks or more: Y N * Walking one (1} block: Y N * Bathing and dressing yourself: Y N .\
Care financins .\dminsrrarion designared :'vte dicare Participaring Heart Bypaos c~mer
H~alrh
APPENDIX
892
1: Patient questionnaire.
Clinical Investigations
BOSTON
UNIVERSITY
SCHOOL
OF
MEDICINE
1 SCHOOL
Of
PUBUC
HEALTH
•
BOSTON
I
I
•
MEDICAL
CENTER
HOSPIT... L
·
BOSTON
UNI VERSITY
GOLD M AN
SCHOOL
OF
GR ... OUATE
DENTISTRY
Department of Cardiothoracic Surgery
~.-··. i
UN I VERSITY
0
! HOSPITAL
i
Dear Doctor ____________________ The Depar t ment of Cardi othoracic Surgery at BUMC is conducting a su r vey of elderly patients ( > 75 year s old) who have und ergone valve replaceme nt sur g ery at our fac i lity . The purpose of the survey is to assess the funct i ona l health status of the se patients after surgery. As part of the project we are also survey i ng the physicians who are caring for these patients. I t would be most aoprec:cated if you would fill out the questionnaire below and return it to us in the envelope provided by____________ Thank you fo r a ssisting us with this important p r oject which we hope wi ll enable us to better evaluate how valvular replaceme:1t surgery a f fects quality of life of elderly patients . Our records show that your patient ____________ ______ under went
______ ______________ on _ ___ ______
1. PATIENT STATUS Liv i ng_ _ __
Deceased_ _ _ ___ If Livi ng
2. FUNCTI ONAL STATUS In your opinion which of the following New York Heart Association class i fication best describes the above patient? Class I Class II Class III Class IV 3 . COMPLICATIONS - - - -Has the above patient experienced any of the following? CHF: Y I N AV Block: Y I N Pacer : Y I N Endocarditis: Y I N Anticoagulation / Bleeding problems: Y I N (specify) Thromboembolic complications: Y I N (if yes cir-c~l-e--a~l~l~~t h-a~t- apply) CVA TIA Mesenteric Cardiac Peripheral Vascular Other Reoperation Y I N 4. ECHOCARDIOGRAM Has the patient had an echocardiogram to assess valve function~ Yes No ( If yes ) Date Results ( i f available ) : EF =-=v:---a-:::1v _e _F=un--c--,t-~ ,--.o_n ________ Valve / perivalvular leak Y I N Structural abnormalit i es Y I N 7
5 . OTHER COMMENTS :\. Health C:1re Financmg .".ciminsrr:ltion designated .Viediore P:1nicipating Heart Bvpuss C.;mer
APPENDIX
2: Physician question naire.
CHEST I 112 I 4 I OCTOBER, 1997
893
REFERENCES 1 US Bureau of the Census. Statistical abstract of the United States: 1994, 109'" ed. Washington, DC: US Bureau of the Census, 1994 2 US Bureau of the Census. Major improvement in life expectancy: 1989. Statistical bulletin, July-September 1990. Washington, DC: US Bureau of the Census, 1990 3 Stewart AL, Ware JE. Measuring functioning and well being-the Medical Outcomes Study approach. 1st ed. Durham, NC: Duke University, 1992; 1-449 4 Tsai TP, Matloff JM , Chaux A, et a!. Combined valve and coronary artery bypass procedures in septuagenarians and octogenarians: results in 120 patients. Ann Thorac Surg 1986; 42:681-84 5 Bessone LN, Pupello, DF, Hiro SP, et a!. Surgical management of aortic valve disease in the elderly: a longitudinal analysis. Ann Thorac Surg 1988; 46:264-69 6 Fremes SE, Goldman BS, Ivanov J, et a!. Valvular surgery in the elderly. Circulation 1989; 80(suppl I):I-77-I-90 7 Fiore AC, Naunheim KS, Barner HB, et al. Valve replacement in octogenarians. Ann Thorac Surg 1989; 48:104-08 8 Merrill WH, Stewart JR, Frist WH, et al. Cardiac surgery in patients age 80 years or older. Ann Surg 1990; 211:772-75 9 Pifarre R. Open heart operations in the elderly: changing risk parameters. Ann Thorac Surg 1993; 56(suppl):S-71-S-73 10 Davis EA, Gardner TJ, Gillinov AM, eta!. Valvular disease in the elderly: influence on surgical results. Ann Thorac Surg 1993; 55:333-38 11 Tsai TP, Chaux A, Matloff JM, et a!. Ten-year experience of cardiac surgery in patients aged 80 years and over. Ann Thorac Surg 1994; 58:445-51 12 Adkins MS, Amalfitano D , Harnum NA, et a!. Efficacy of combined coronmy revascularization and valve procedures in octogenarians. Chest 1995; 108:927-31 13 Kumar P, Zehr KJ, Chang A, et a!. Quality of life in octogenarians after open heart surgery. Chest 1995; 108: 919-26 14 He GW, Grunkemeier GL, Starr A. Aortic valve replacement in elderly patients: influence of concomitant coronaty grafting on late survival. Ann Thorac Surg 1996; 61:1746-51 15 Chocron S, Etievent JP, Clement F, eta!. Is surge1y for aortic stenosis justified in patients over 75 years of age? J Cardiovasc Surg 1996; 37:255-59 16 'Nixon CL, Lust RM, Meadows LW, et al. Mitral valve surgery in the elderly [abstract]. Chest 1996; llO(suppl):62S 17 Geholt A, Mullany CJ, Ilstrup D, eta!. Aortic valve replacement in patients aged 80 years and older: early and long-term results. J Thorac Cardiovasc Surg 1996; 111:1026-36 18 The Society of Thoracic Surgeons National Cardiac Surgery Database. Manual for data managers-definitions of terms. Minneapolis: Summit Medical Systems, 1995 19 Vital Statistics of the United States, 1992. Life-tables (volume II, section 6). Hyattsville, Md: US Dept of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Center for Health Statistics, 1996 20 Braunwald E. Valvular heart disease. In: Braunwald E, ed.
894
21
22
23
24
25
26
27
28
29
30
31 32 33 34 35
36
Heart disease: a textbook of cardiovascular medicine. Philadelphia: WB Saunders, 1997; 1007-76 Kirklin JW, Barratt-Bayes BG. Mitral valve disease vvith or without tricuspid valve disease. In: Kirklin JW, Barratt-Bayes BG, eds. Cardiac surge1y. New York: Churchill Livingstone, 1993; 425-89 Kirklin JW, Barratt-Bayes BG. Aortic valve disease. In: Kirklin JW, Barratt-Bayes BG, eds. Cardiac surgery. New York: Churchill Livingstone, 1993; 491-571 Society of Thoracic Surgeons. Data analysis of the STS national cardiac surgery database: the sixth year. Chicago: Society of Thoracic Surgeons, 1997; 94-167 Flameng WJ, Herijgers P, Scecsi J, et a!. Determinants of early and late results of combined valve operations and coronary artery bypass grafting. Ann Thorac Surg 1996; 61:621-28 Verheul HA, van den Brink RB, Bouma BJ, et al. Analysis of 1isk factors for excess mortality after aortic valve replacement. J Am Coli Cardiol 1995; 26:1280-86 Shapira OM , Aida GS, Zelingher J, et a!. Enhanced blood conservation and improved clinical outcome after valve surgery using heparin-bonded cardiopulmonary bypass circuits. J Cardiac Surg 1996; 11:307-17 Aldea GS, Doursounian M, O'Gara P, et al. Heparin-bonded circuits with a reduced anticoagulation protocol in primary CABG: a prospective randomized study. Ann Thorac Surg 1996; 62:410-18 Borkon AM, Soule LM, Baughman KL, et a!. Aortic valve replacement in the elderly patient. Ann Thorac Surg 1988; 46:270-77 Jamieson WRE, Burr LH, Munro AI, et a!. Cardiac valve replacement in the elderly: clinical performance of biological prostheses. Ann Thorac Surg 1989; 48:173-85 Chambers LW, MacDonald LA, Tugwell P, et a!. The McMaster H ealth Index Questionnaire as a measure of quality of life for patients with rheumatoid disease. J Rheumatol 1982; 9:780-84 Bergner M, Bobbitt RA, Carter \VB, et a!. The Sickness Impact Profile: development and final revision of a health status measure. Med Care 1981; 19:787-805 Wan TT, Livieratos B. Interpreting a general index of subjective well-being. Milbank Mem Fund Q Health Soc 1978; 56:531-56 Stewart AL, Hays RD, Ware JE. The MOS short-form health survey: reliability and validity in patient population. Med Care 1988; 26:724-35 Lavie CJ, Milani RV. Effects of cardiac rehabilitation and exercise training programs in patients ?.75 years of age. Am J Cardia! 1996; 78:675-77 Bubien RS, Knotts-Dolson SM, Plumb VJ. Effect of radiofrequency catheter ablation on health-related quality of life and activities of daily living in patients \vith recurrent arrhythmias. Circulation 1996; 94:1585-91 Hlatky MA, Rogers WJ, Johnstone I, eta!. Medical care costs and quality of life after randomization to coronary angioplasty or coronary bypass surgery. N Eng! J Med 1997; 336:92-99
Clinical Investigations
Background: Assessment of quality of life has become an increasingly important aspe ct of the risk-benefit analysis of any therapeutic intervention, particularly in high-risk populations, such as the elderly. Methods: Clinical outcomes of 147 consecutive patients older than 75 years undergoing valve surgery between 1992 and 1995 were reviewe d . Long-term quality of life was assessed using a modified version of the Duke University Me dical Outcomes Study system. Results: Mean age was 79.3±4 years, range was 75 to 89 years, and 67% (99/147) were male. Preoperatively, 128 patients (87%) were in New York Heart Association (NYHA) functional class III/IV. Fifty-two percent (77/147) of the operations were nonelective. Concomitant coronary artery bypass grafting was performed in 69 patients (47%). Thirty-day hospital mortality was 7.5% (11/147). Mean follow-up for 98% (133/136) of hospital survivors was 30±13 months (range, 2 to 55 months). Actuarial survival at 55 months was 71±6%-equivalent to a g eneral age-, race-, and gender-matched population reported in the Life Tables of the US National Center for Health Sta tistics. At the time offollow-up, 112 patie nts (96%) lived at home, 78% (91/116) defined their health between good to excellent, and 81 % (93/114) stated that the operation improved their health status. Ninety-seven percent (112/116) were able to bathe and dress independently, 92% (104/113) could walk at least one block, and 88.5% (100/113) could climb at least one flight of stairs. Moderate to vigorous activities could be performe d b y59.2% (67/113). Overall, at the time of follow-up, 81 % (95/117) were in NYHA class 1/11. Conclu~ion: In a selected patient population, valve surgery in the elderly is associated with acceptable early morbidity and mortality. Long-term survival and quality of life are excellent. These facts strongly support the performance of these procedures in patients older than 75 years. (CHEST 1997; 112:885-94) Key words: elderly; qu ality of life; valve surgery Abbreviations: AVR =aortic valve replacement; CABG =coronary arte1y bypass grafting; CPB =cardiopulmonary bypass; CVA=cerebrovascular accident; DVR=double valve replacement; LVEF= left ventricular eje ction fraction; MI = myocardial infarction; MOS system= Medical Outcomes Study system; MVR=mitral valve replacement; NYHA=New York Heart Association
Jn the early 1990s,
3% ( 7.4) million of the US population were older than 80 years.1 The current life expectancy of octogenarians is inthe range of 6.0
Global theme issue on aging *From the Department of Cardiothoracic Surgery, Boston Medical Center. in abstract fo rm at the 46th Annual r This article was p esented Meeting of the American College of Cardiology, March 16-19, 1997, Anal1eim, Calif. Manuscript received April 8, 1997; revision accepted July 28. Reprint r equests: Oz M. Shap·ira, MD, Department ofCardiothoracic Su·rgery, Boston Medical Center, 88 E Newton St, Boston, MA 0211 8; email: [email protected]. edu
to 8. 1 years, leading to an expected rise in their number t o > 10 million (4.3%) within the next 10 years.l·2 Of these octogenarians, 40% have serious symptomatic heart disease. 2 This tr end is reflected in our institution b y the increased fraction of elderly patients undergoing o pen heart surgery from 1 2 to 22% of total cases over the past decade. Quality of life encompasses the total well-being of the person. Assessment of quality of life includes evaluation of functional capacity, symptoms, and the patient's perception of well-being. 3 In the e ra of managed care, quality of life has become an increasingly important end-point measure, being an integral part of the e valuation of the risk-benefit ratio and CHEST I 11 2 I 4 I OCTOBER, 1997
885
cost-effectiveness of a variety of therapeutic modalities. Advances in anesthetic and surgical techniques, as well as improved perioperative care, have resulted in a significant improvement in short- and long-term clinical outcome after open heart surgery in the elderly.4 - 17 However, very little is reported on the effect of these operations on quality of life. The purpose of this study was to evaluate the clinical outcome of valve surge1y in 14 7 patients older than 75 years undergoing valve surgery, focusing on longterm quality of life.
MATERIALS AND METHODS
(3) Comorbid risk factors included hypertension, diabetes mellitus, prior cerebrovascular disease, peripheral vascular disease, smoking, COPD, chronic renal failure (c reatinine > 2.5), dialysis, prior myocardial infarction, and preoperative use of anticoagulants. (4)Operative data included type of operation, concomitant procedures, type of CPB circuit, CPB and aortic cross-clamp times, use of inotropes and intra-aortic balloon pump, and intraoperative complications. (5) Postoperative complications included 30-day mortality, and significant morbidity was defined as reoperation for bleeding, mediastinal infection, pneumonia, respirator > 3 days, transient ischemic attack and/or cerebrovascular event, myocardial infarction (new Q wave, and/or elevation of creatine phosphokinase MB 2:50 U), low cardiac output (a newly placed intra-aortic balloon pump or the use of inotropes for > 24 h t o maintain a cardiac index > 2.0), valve thrombosis, and other major complications (vascular, GI, e tc) .6 Bleeding and transfu sion require ments were also noted.
Patients
Long-term Follou;-up and Assessment of Quality of Life
The study population consisted of 147 consecutive patients older than 75 years undergoing valve surgery in a tertiary teaching hospital between January 1992 and July 1995.
Two questionnaires, one for patients and one for physicians caring for them, were developed to collect long-term results. Each questionnaire included a brief explanation of the project. The patient's questionnaire (Appendix l ) was based on the Duke University Medical Outcomes Study (MOS ) system. To maximize the response rate, the questionnaire was k pe t as brief as possible, and consisted of three of the six items originally included in the MOS system. The focus was on functional capacity, the patient's perception of well-being, and the impact of surgery on overall health status. The physician questionnaire (Appendix 2) focused on long-term morbidity, mortality, and fun ctional capacity as determined b y NYHA classification. Patients, families, and/or physicians who failed to respond were directly contacted by phone to achieve as complete follow-up as possible.
Surgical Techniques
Operations were performed using cardiopulmonary bypass (CPB) and mild systemic hypothermia (34°C to 35°C). Myocardial protection was achieved using antegrade and retrograde cold W) blood cardioplegia, supplemented by topical cooling with cold saline solution. Aminocaproic acid (Amicar; American Reagent; Shirley, NY) was used routinely in this study. Departmental thresholds for homologous blood transfusion were defined a s hematocrit of 20% durin g CPB and 25% postope ratively. Use of clotting factors was based on clinical assessment of bleeding and he matologic evaluation, s pecifically correcting factor deficits. Persiste nt bleeding in excess of 300 mL in the first hour or 500 mL in the first 2 hwas considered an indication for transfusion of 5 to 10 U of platelets and 2 to 4 U of fresh frozen plasma. The departmental transfusion guidelines w ere rigid and did not change ove r the study period. Data Collection
Hospital and clinical records were re viewed r etrospectively. Data retrieved included the following: (l ) demographic information-age, gender, body surface area, indication for surgery, date of ope ration, surgeon , status of the procedure (elective vs urgent/emergent). We used the Society of Thoracic Surgeons criteria to define the status of the procedure. 18 Elective: an elective operation was one that was performed on a patient with cardiac function that had been stable in days or weeks prior to the operation. Urgent: an urgent operation was one in which surgery was required within 24 h in order to minimi ze the chance of further d eterioration. Delay in operation was necessitated only by attempts to improve the patient's condition, availability of a spouse or a parent for inform ed consent, availability of blood products, or the availability of results of essential laboratory procedures or tests. Emergent: patients requiring emergency operations had ongoing, refractory unrelenting cardiac compromise, with or without hemodynamic instability, and were not responsive to any form of therapy except cardiac surgery. An emergency operation was one in which there was no delay in providing operative intervention 1 8 (2) Preoperative left ventricular ejection fraction (LVEF) and functional status were determined b ythe NewYork Heart Association (NYHA) classification. 886
Statistical Analysis
Statistical analysis was performed using the statistical software (SAS version 6.11; SAS Institute; Cary, NC ). Data are expressed as mean±SD, and for selected variables also by the median and range. Two-tailed Student t test was used to analyze continuous variables. Categorical variables were analyzed u sing x2 with Yates' correction or Fisher's Exact Test when appropriate . A p value of < 0.05 was considered significant. Multivariate analysis using logistic regression was used to identify factors associated with early and late mortality. For the logistic regression analysis, the LOGISTIC procedure was utilized. A backward selection method was used to select the variables in the model, with a significance level for entry into the model of 0.1 and a significance level of 0.05 for staying in the model. Long-term survival was analyzed using the Kaplan-M eier method. The survival of the study group was compared to age-, race-, and gender-matched general US population, obtained from the Life Tables of the US National Center for Health Statistics,19 using the log rank test. All the calculations were performed using a procedure (SAS LIFETEST)
RESULTS
Baseline Characteristics of the Study Group The baseline clinical and operative profiles of the study group are summarized in Table l. There were 99 men and 48 women with a mean age of 79.3::!:::3.8 years (range, 75 to 89 years). Mean LVEF was Clinical Investigations
Table !-Baseline Characteristics of the Study Group Operative Profile
Clinical Profile Variable
n=147
Age, (yr) Male/female Preoperative NYHA class
79.3::+::3.8 (range 75-89 years ) 99/48 3.3 : +: 0.7 0 (0%) 19 (13%) 57 (39%) 71 (48%) 53.9::+::16.1% 40 (27%) 77 (52%) 13 (9%) 80 (54%) 69 (47%) 45 (31%) 32 (22%) 20 (14%) 15 (10%) 13 ( 9%) 6 (4%)
I II III
IV LVEF LVEF < 45% Nonelective Reoperation Hypertension CAD* Cigarette smoking Diabetes mellitus COPD PVD* Prior CVA Renal failure
n = l47
Variable Procedure AVR AVR+ CABG MRV
54 (37%) 51 (35%) 11 (7%) 10 (7%)
MVR + CABG
6 (4%) 3 (2%) 7 (5%) 5 (3%) 69 (47%)
Mitral Valve Repair Mitral valve repair+CABG DVR DVR+CABG Overall concomitant CABG Prosthesis Bioprosthesis Mechanical CPB time, min Aortic cross-clamp time, min
121 (82%) 17 (12%) 119.8::+::45 72.6::+:: 25
*CAD=coronary artety disease; PVD=peripheral vascular disease.
53.9±16.1% with 27% having LVEF of <45%. Mean preoperative NYHA functional class was 3.3±0.7 with 87% being in class III/IV. A significant number of patients had multiple comorbidities such as hypertension, coronary artery disease, smoking, diabetes, prior stroke, and others. Aortic valve replacement (AVR) was performed in 105 patients mostly for senile calcific aortic stenosis; mitral valve repair (n =9) or replacement (n = 21) (MVR ) was performed in 30 patients, one third of whom had ischemic mitral regurgitation; and double valve replacement (DVR) procedures were performed in 12 patients. Concomitant coronary artery bypass grafting (CABG) was performed in 69 patients (47%) . More than 50% of the operations (73/147) were nonelective, and 9% were reoperations. Bioprostheses were implanted in most patients (1211147 [82%] ). There was no significant difference between mechanical valves and bioprostheses with regards to the size of the prosthesis implanted: AVR prosthesis size-mechanical: 20.6±2 mm vs bioprosthesis: 21.6±3 mm (p=0.36); MVR prosthesis sizemechanical: 30.0±4 mm vs bioprosthesis: 28.8±3 (p=0.46). Conventional CPB circuits were used in 91 patients and heparin-bonded circuits were used in 56. Early Mortality and Morbidity Early clinical outcomes are summarized in Table 2. Thirty-day operative mortality was 7.5% (ll/147). The causes of hospital deaths included the following: CVA, two patients; ischemic bowel, two patients;
coagulopathy, two patients; aortic dissection, postinfarction ventricular septal defect, perioperative myocardial infarction (MI ), multiorgan failure , and atrioventricular rupture each in one patient. Patients with bioprostheses had a mortality of 5.8% (7/121), and patients with mechanical valves had a mortality of 23.5% (4/17) , p=0.04. The causes of deaths in patients with mechanical valves included CVA in two patients, atrioventricular rupture , and multiorgan
Table 2-Early Clinical Outcomes Variable
n=147
30-d operative mortality Overall morbidity Morbidity excluding uncomplicated atrial fibrillation Atrial fibrillation/flutter Pneumonia Ventilator support > 3 d Tracheostomy Reoperation fo r bleeding CVA Renal Failure MI Mediastinal infection Total blood transfusion s, donor units Patients transfus ed Hours of ventilatory suppott Surgical ICU length of stay, d Total hospital l ength of stay, d
11 (7.5%) 89 (61 %) 52 (35%) 53 (36%) 9 (6%) 24 (16%) 11 (7.5%) 15 (10%) 7 (4.8%) 3(2%) 1 (0.7%) 0 (0%) 22.3::+:: 26 (median, 15; range, 0-34) 136 (94%) 37.8 : +: 63 (median, 20; range, 6-342) 3.9::+::6 (median, 2; range, 1-49) 10.3::+:: 7 (median , 8; range, 5-52)
CHEST I 112 I 4 I OCTOBER, 1997
887
Table 3-Univariate (Top) and Multivariate (Bottom) Analysis of Early Mortality* Variable
Alive (n =136)
D ead (n = ll )
Nonelective CPB time, min Aortic cross-clamp time , min Intra-aortic balloon pump 24-h chest tube drainage, mL Reoperation for bleeding Total blood donor exposure, U Postoperative CVA Bioprosthesis
68 (50%) 115.5::!:31 71.2::!:23 2 (1.5%) 979::!:720 12 (8.8%) 19.3::!:22 5(3.7%) 114 (83 .8%)
9 (82%) 158.0::!:119 89.8::!:34.2 3 (27.2%) 2078::!:2534 3 (27.2%) 63.3::!:45 2 (18.2%) 7 (63.6%)
Variable Total blood donor exposure, U Postoperative CV A Bioprosthesis
~
Coefficient 0.0499 2.8001 -3.5104
p Value 0.04 0.002 0.015 0.015 0.0004 0.05 <0.00001 0.05 0.077
Odds Ratio (95% CI)
p Value
1.05 (1.02-1.08) 16.45 (1.49-18.58) 0.03 (0.002-0.38)
0.044 0.022 0.007
*Early mortality: 30-day in-hospital deaths. 1 CI=confidence interval.
failure. The causes of death in these patients could not be directly attributed to the use of mechanical valves. Overall morbidity was quite high (61 %), mainly due to high rates of respiratory complications, postoperative atrial fibrillation, and reoperation for bleeding. Excluding uncomplicated episodes of atrial fibrillation overall, morbidity was still significant35% (521147). Postoperative CVA occurred in 4.8% (7/147). Three patients clearly had an intraoperative neurologic event. However, four patients were neurologically intact immediately after the operation, and the neurologic deficit developed 3 to 5 days after surgery. In all four patients, the development of CVA was strongly related to new-onset atrial fibrillation of 2:24-h duration. Of note are the very low rates of perioperative MI and mediastinal infection. Most patients (138/147, 94%) required significant amount of homologous blood transfusions, an average of 22.3:±:26 donors (median, 15 U; range, 0 to 134 U). Blood product utilization was high for each of the components: RBCs: 5.5:±:6 U (median, 4 U; range, 0 to 34 U); fresh frozen plasma: 3.4:±:4 U (median, 2 U; range, 0 to 22 U); platelets: 9.9:±:11 U (median, 6 U; range, 0 to 59 U). Even after exclusion of the 15 patients (10%) who underwent reoperation for bleeding, 93% (123/132) required transfusion of an average of 16.3:±: 17 donors (median, 11 U; range, 0 to 78 U). The relatively high transfusion and complication rates resulted in an average of 37.8:±:63 h (median, 20 h; range, 6 to 432 h) of ventilatory support, 3.9:±:6.5 days (median, 2days; range, 1 to 49 days ) of ICU stay, and total hospital length of stay of 10.3±7.3 days (median, 8 days; range, 5 to 52 days ). These figures are significantly higher than our results following similar operations performed in younger patients (data not shown). Factors associated with 888
increased early mortality in a univariate analysis (Table 3, top) included nonelective operations, increased cardiopulmonary and aortic cross-clamp times, use of intra-aortic balloon pump, increased postoperative mediastinal chest tube output, reoperation for bleeding, increased homologous transfusions, and postoperative CVA. However, in a multiple logistic regression analysis, only postoperative CVA and increased homologous transfusions were identified as independent predictors of increased mortality, and the use of bioprosthetic valve was associated with decreased mortality (Table 3, bottom). There was a trend toward increased mortality for mitral valve procedures compared to aortic valve, but the difference did not reach significance (MVR, 5/30 [16.6%] vs AVR 5/105 [4.8%], vs DVR 1/12 [9.1 %], p==0.09). Performing CABG in addition to valve replacement did not significantly affect early mortality (valve procedure alone 4/74 [5.1 %] vs valve procedure with CABG 7/69 [10.1 %], p==0.15). Long-term Mortality Follow-up was complete for 98% (133/136) of hospital survivors and ranged from 2 to 55 months (mean, 30±13 months ). Fifteen patients (10.2%) died during the follow-up period. The most common cause of long-term mortality was congestive heart failure , 7 of 15 (46.6%). Actuarial survival at 55 months was 71±6%, equivalent to a general age-, race-, and gender-matched population reported in the Life Tables of the US National Center for Health Statistics (Fig 1).1 9 Factors associated with late mortality in a univariate analysis (Table 4, top ) included advanced age, male gender, preoperative history of CVA, postoperative respiratory failure requiring ventilatory support >3 days and tracheostomy, postopClinical Investigations
1.0 0.9 0.8 ii
>
0.7
·s:...
0.6
th
0.5
:::s
...c
Cl)
:;::;
ca
a.
......... Study Group US General Population
I3-B-£J
0.4
4-year Survival 76.9 ± 5.8% Study Group 77.0 ± 6.8% US Population
0.3
p
0.2 0.1 0.0
147
136
120
Number of Patients at Risk 102 82 65 49
0
6
12
18
24
30
36
=0.7594
36
15
42
48
54
Follow Up Months after Surgery FIGURE 1. Kaplan-Meier analysis of survival, comparing the study group to an age-, race-, and gender-matched US general population.
erative CVA, and postoperative atrial fibrillation. In a multiple logistic regression analysis, advanced age, male gender, preoperative and postoperative CVA, and the need for tracheostomy remained significant independent predictors of long-term mortality (Table 4, bottom).
Quality of Life Quality of life results are summarized in Figures 2-4. At the time of follow-up, 112 (96%) lived at
Table 4-Univariate (Top) and Multiple Logistic Regression (Bottom) Analysis of Long-term Mortality* Variable
Alive (n=121)
Dead (n=15)
p Value
Ventilator >3 d Postoperative atrial fibrillation Age, yr Gender, male/female Preoperative CVA Postoperative CVA Tracheostomy
14 (11.6%) 42 (34.7%) 79.0:':4 35/86 8(66%) 2 (1.6%) 5 (4.1% )
7 (46.6%) 9(60%) 81.1:':4 9/6 4 (26.6%) 3(20%) 5 (33.3%)
0.023 0.053 0.039 0.019 0.028 0.009 0.001
J3 Coefficient Odds Ratio (95% CI 1 ) p Value
Age, yr Gender, male/female Preoperative CVA Postoperative CVA Tracheostomy
0.1803 1.6331
1.19 (1.005-1.43) 5.12 (1.32-19.77)
0.044 0.019
2.8956 2.2770 2.3340
18.09 (3.69-88.75) 9.753 (1.52-62.71) 10.32 (1.71-62.33)
0.0004 0.016 0.011
*Long-term follow-up: range, 2-55 months (mean, 30:':13 months). 1 See Table 3 footnote.
home, 78% (91/116) defined their health status between good and excellent (Fig 2), and 81% (93/ 114) stated that the operation improved their health status (Fig 3). Ninety-seven percent (112/116) were able to bathe and dress independently, 92% (104/ 113) could walk at least one block, and 88.5% (100/113) could climb at least one flight of stairs. Moderate to vigorous activities could be performed by 59.2% (67/113) (Fig 4, A [top]) . Most patients who could not perform vigorous activities were limited by concomitant disorders such as CVA, degenerative joint disease, and impaired visual acuity, rather than by their cardiovascular performance. Overall, at the time of follow-up, 81% (95/117) were in NYHA class IIII-a significant change compared to preoperative status (Fig 4, B [bottom]).
Exellent
V.good Good Fair
0
10
20
30
% of patients
FIGURE 2. Health self-perception. Summary of the patients' responses to the first question being asked to define their overall health status at the time of follow-up.
CHEST I 112 I 41 OCTOBER, 1997
889
Same Somewhat worse
Much worse 0
20
10
30
40
50
60 %of Patients
FIGURE 3. Impact of surgery on health status. Summary of the patients' responses to the second question being asked to compare their health status at the time of follow-up to the preoperative status.
DISCUSSION
Evaluation of the short- and long-term results of surgery in the elderly is difficult, since there is no true control group. A randomized trial comparing surgery to medical treatment would be ideal, but it is unlikely to be performed based on the existing data of the natural history of hemodynamically significant valvular heart disease treated medically. 20 -22 Comparison to age-, race-, and gender-matched population of patients with the same cardiac abnormalities
A Bathe and dress Walk one block Climb one flight Moderate to vigorous
•
l
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
0
20
40
60
I
80
100
% of patients
B 70
II
Ill
IV
NYHAclass
FIGURE 4.Functional capacity. A, top: summary of the patients' responses to the third question evaluating their own functional capacity. B, bottom: preoperative vs follow-up NYHA fun ctional class. 890
who were eligible for surgery and refused is another theoretical option, but it would be difficult to perform . Compa1ison to younger patients carries significant limitations as older patients have different cardiac diseases, significantly more comorbidities, and usually are referred later for surgery. 2 0- 22 The most common aortic valve abnormality in the elderly is senile calcific aortic stenosis, whereas young patients are frequently operated on for aortic regurgitation or a mixed lesion. Rheumatic and degenerative mitral valve disease are the most frequent indications for surgery in the young, whereas ischemic mitral regurgitation is much more common in the older age group. 20 -22 Also, concomitant coronary artery disease with the need for myocardial revascularization is more common in the elderly.2 1·22 These trends are reflected in our study group, where 90% of aortic valve replacements were performed in patients with calcific aortic stenosis, one third of mitral valve procedures were performed in patients with ischemic mitral regurgitation, and concomitant CABG was indicated in 47% of patients. The late referral is reflected by the advanced preoperative NYHA functional class (87% in class III/IV), and the fact that > 50% of the operations were nonelective. Considering this very high-risk profile, the observed 7.5% operative mortality seems to be acceptable, and compares favorably with the reported mortality rates in the most recent Society of Thoracic Surgeons national database 23 and previous studies evaluating valve surgery in the elderly, with reported mortality rates of9 to 37%.4 - 17 Most likely, the lower mortality is a r esult of continued improvement in anesthesia, surgical techniques, and perioperative care, but also of patient selection. All the patients operated on during the study period were included in our analysis with no exclusion. However, data were not available on patients who had significant valvular disease but have not been referred for surgery by their primary care physicians and/or cardiologist, or have been rejected by the surgeon. The issue of selection bias is one of the major limitations of a retrospective study. Elderly patients had prolonged ventilatory support and longer ICU and total hospital lengths of stay. These figures reflect not only the high risk of this specific group, but also our practice during the study period. Since than, early extubation and a "fast track" protocol have been aggressively implemented in all of our patients, resulting in a significant reduction of ventilatory support time and lengths of ICU and hospital stays. Similar to previous reports,6,lO,ll mortality following mitral valve surgery was higher compared to aortic valve surgery. This did not reach significance in our study, most likely due to a small sample size. Also, concomitant CABG was not associated with increased mortality, an issue that is still Clinical Investigations
debatable in the literature. 4 •11 · 17 •23· 25 The only independent predictors of early mortality in this series were increased homologous blood transfusions and postoperative CVA. The rate of reoperation for bleeding was relatively high. 23 However, even after exclusion of this subset of patients, blood transfusions were required in >90% of patients. Thus, the high transfusion rates in this age group may be related to other factors such as higher incidence of preoperative anemia, high percentage of women with small body surface area, and higher incidence of postoperative coagulopathy. In the latter part of the study, we switched to heparin-bonded CPB circuits with lower systemic anticoagulation protocol with a dramatic reduction of the rate of bleeding and blood product utilization. 26 Postoperative CVA occurred in 4.8% of our patients, which is somewhat lower than that previously reported in this age population. 4 · 17 In our study, postoperative CVA was related to newonset atrial fibrillation in nvo thirds of the patients. Undoubtedly, efforts aimed at decreasing the incidence of postoperative atrial fibrillation are of prime importance. We have shown previously that strategies such as the use of heparin-bonded circuits effectively reduced the rate of atrial fibrillation in patients undergoing CABG.27 Similar approaches should be studied in this patient population. In this study, the use of bioprostheses was associated with lower early mortality, although the reason is not entirely clear. The superior short- and long-term results with bioprosthetic valves in this age population have been documented previously. 28•29 Analyzing the effect of the type of prosthesis on long-term result was impossible in our study due to small number of patients with mechanical valves. A major finding in this study was that patients who left the hospital had excellent long-term survival, similar to that of an age-, gender-, and race-matched US general population. Of note is that postoperative CVA was the only variable identified as a significant predictor of both early and late mortality. Although occurring in only 4.8%, when it does occur, it is a bad prognostic indicator. Since the long-term survival of these patients is excellent, the issue of quality of life becomes an increasingly important aspect of the evaluation of the risk-benefit ratio and cost-effectiveness. Previous studies evaluating long-term results of valve surgery in the elderly have used survival and NYHA classification as the major end points and have reported improving results. 4 · 17•24 •25 However, other measures such as placement of the patient, health self-percep-
tion, and other important components of overall quality of life assessment have not been investigated previously in this specific patient population. Assessment of quality of life is a complex process because it relies heavily on subjective data, it requires evaluation of many variables, and it is very difficult to quantify. 3 Comprehensive health surveys such as the McMaster Health Index Questionnaire,3 0 the Sickness Impact Profile,31 and others are lengthy instruments that make them clinically impractical. Single-item measures such as the National Health and Nutrition Examination Survef 2 are short and result in an improved response rate but are usually less precise, less reliable, and less valid. The Duke University MOS system,33 which was developed as a compromise benveen these extremes, subsequently has become a very effective tool in evaluating quality of life in a variety of clinical trials. 34· 36 In this study, we used a modified version of the MOS system because of the specific characteristics of the study group, being aware of the lower compliance rate for these type of studies in the elderly. 33 The focus was on functional capacity, the patient's perception of well-being, and the impact of surgery on overall health status. Bodily pain was not assessed as most patients were far out from surgery. Work status was believed to be not applicable as the patients were all retired. The psychosocial status was not included as it is a more complex issue, and we believed that it should be assessed separately. Using this short version, we were able to achieve a very high response rate and obtain a reliable evaluation of long-term quality of life of this patient population. The overall results were very encouraging. Most patients stated that the operation had a significant positive impact on their overall health status, and estimated their health status at the time of follow-up to be benveen good to excellent. Also, the vast majority of patients lived at home, either alone or with family, and were functionally independent. Most patients were able to function at a relatively high level. Those who did not were limited by other comorbidities and not by their cardiovascular performance. In conclusion, valve surgery in the elderly is associated with acceptable early morbidity and mortality. Hospital survivors are expected to have a life expectancy that is equal to that of age-, race-, and gender-matched population. Long-term quality of life in the elderly after valve surgery is excellent. These facts strongly support the performance of valve surgery in patients older than 75 years.
CHEST I 112 I 4 I OCTOBER, 1997
891
BOSTON
UN IVERSITY
SCHOOL
OF
MEDICINE
I
SCHOOL
OF
PUBUC
HEALTH
•
bOSTON
UNIVERSITY
MEDICAL
CENTER
HOSPITAL
•
UNIVERSITY
GOLDMAN
SCHOOL
OF
GRADUATE
OENnSntY
Department of Cardiothoracic Surgery
Dear_____________________________________ The Department of Cardiothoracic Surgery at BUMC is conducting a survey of patients who have undergone valvular replacement surgery at our facility. The purpose of this survey is to assess activity levels of patients after surgery. It would be most appreciated if you would fill out the questionnaire below and return it to us in the envelope provided by Thank you for assisting us with this important project which we hoDe will enable us to better evaluate how valve replacement sur·gery affects quality of life.
Our records show that you received a on 1. In general, would you say your health is: that best describes your health today) Excellent Very good Good Fair
(circle the answer Poor
2. Compared to before your valve surgery, how would you rate your health in general now? (Select one) * Much better than before my surgery * Somewhat better than before my surgery * About the same as before my surgery ----* Somewhat worse than before my surgery ____ * Much worse than before my surgery ____ 3. The following items are about activities that you might do during a typical day. (Please indicate by circling yes (Y) or no (N} whether you are able to participate in these activities:) * Vigorous activities, such as lifting heavy objects, playing tennis: Y N * Moderate activities, such as moving a table, pushing a vacuum cleaner, bowling, or playing golf: Y N * Climbing two (2) flights of stairs: Y N Y N * Climbing one (1) flight of stairs : * Walking two (2} blocks or more: Y N * Walking one (1} block: Y N * Bathing and dressing yourself: Y N .\
Care financins .\dminsrrarion designared :'vte dicare Participaring Heart Bypaos c~mer
H~alrh
APPENDIX
892
1: Patient questionnaire.
Clinical Investigations
BOSTON
UNIVERSITY
SCHOOL
OF
MEDICINE
1 SCHOOL
Of
PUBUC
HEALTH
•
BOSTON
I
I
•
MEDICAL
CENTER
HOSPIT... L
·
BOSTON
UNI VERSITY
GOLD M AN
SCHOOL
OF
GR ... OUATE
DENTISTRY
Department of Cardiothoracic Surgery
~.-··. i
UN I VERSITY
0
! HOSPITAL
i
Dear Doctor ____________________ The Depar t ment of Cardi othoracic Surgery at BUMC is conducting a su r vey of elderly patients ( > 75 year s old) who have und ergone valve replaceme nt sur g ery at our fac i lity . The purpose of the survey is to assess the funct i ona l health status of the se patients after surgery. As part of the project we are also survey i ng the physicians who are caring for these patients. I t would be most aoprec:cated if you would fill out the questionnaire below and return it to us in the envelope provided by____________ Thank you fo r a ssisting us with this important p r oject which we hope wi ll enable us to better evaluate how valvular replaceme:1t surgery a f fects quality of life of elderly patients . Our records show that your patient ____________ ______ under went
______ ______________ on _ ___ ______
1. PATIENT STATUS Liv i ng_ _ __
Deceased_ _ _ ___ If Livi ng
2. FUNCTI ONAL STATUS In your opinion which of the following New York Heart Association class i fication best describes the above patient? Class I Class II Class III Class IV 3 . COMPLICATIONS - - - -Has the above patient experienced any of the following? CHF: Y I N AV Block: Y I N Pacer : Y I N Endocarditis: Y I N Anticoagulation / Bleeding problems: Y I N (specify) Thromboembolic complications: Y I N (if yes cir-c~l-e--a~l~l~~t h-a~t- apply) CVA TIA Mesenteric Cardiac Peripheral Vascular Other Reoperation Y I N 4. ECHOCARDIOGRAM Has the patient had an echocardiogram to assess valve function~ Yes No ( If yes ) Date Results ( i f available ) : EF =-=v:---a-:::1v _e _F=un--c--,t-~ ,--.o_n ________ Valve / perivalvular leak Y I N Structural abnormalit i es Y I N 7
5 . OTHER COMMENTS :\. Health C:1re Financmg .".ciminsrr:ltion designated .Viediore P:1nicipating Heart Bvpuss C.;mer
APPENDIX
2: Physician question naire.
CHEST I 112 I 4 I OCTOBER, 1997
893
REFERENCES 1 US Bureau of the Census. Statistical abstract of the United States: 1994, 109'" ed. Washington, DC: US Bureau of the Census, 1994 2 US Bureau of the Census. Major improvement in life expectancy: 1989. Statistical bulletin, July-September 1990. Washington, DC: US Bureau of the Census, 1990 3 Stewart AL, Ware JE. Measuring functioning and well being-the Medical Outcomes Study approach. 1st ed. Durham, NC: Duke University, 1992; 1-449 4 Tsai TP, Matloff JM , Chaux A, et a!. Combined valve and coronary artery bypass procedures in septuagenarians and octogenarians: results in 120 patients. Ann Thorac Surg 1986; 42:681-84 5 Bessone LN, Pupello, DF, Hiro SP, et a!. Surgical management of aortic valve disease in the elderly: a longitudinal analysis. Ann Thorac Surg 1988; 46:264-69 6 Fremes SE, Goldman BS, Ivanov J, et a!. Valvular surgery in the elderly. Circulation 1989; 80(suppl I):I-77-I-90 7 Fiore AC, Naunheim KS, Barner HB, et al. Valve replacement in octogenarians. Ann Thorac Surg 1989; 48:104-08 8 Merrill WH, Stewart JR, Frist WH, et al. Cardiac surgery in patients age 80 years or older. Ann Surg 1990; 211:772-75 9 Pifarre R. Open heart operations in the elderly: changing risk parameters. Ann Thorac Surg 1993; 56(suppl):S-71-S-73 10 Davis EA, Gardner TJ, Gillinov AM, eta!. Valvular disease in the elderly: influence on surgical results. Ann Thorac Surg 1993; 55:333-38 11 Tsai TP, Chaux A, Matloff JM, et a!. Ten-year experience of cardiac surgery in patients aged 80 years and over. Ann Thorac Surg 1994; 58:445-51 12 Adkins MS, Amalfitano D , Harnum NA, et a!. Efficacy of combined coronmy revascularization and valve procedures in octogenarians. Chest 1995; 108:927-31 13 Kumar P, Zehr KJ, Chang A, et a!. Quality of life in octogenarians after open heart surgery. Chest 1995; 108: 919-26 14 He GW, Grunkemeier GL, Starr A. Aortic valve replacement in elderly patients: influence of concomitant coronaty grafting on late survival. Ann Thorac Surg 1996; 61:1746-51 15 Chocron S, Etievent JP, Clement F, eta!. Is surge1y for aortic stenosis justified in patients over 75 years of age? J Cardiovasc Surg 1996; 37:255-59 16 'Nixon CL, Lust RM, Meadows LW, et al. Mitral valve surgery in the elderly [abstract]. Chest 1996; llO(suppl):62S 17 Geholt A, Mullany CJ, Ilstrup D, eta!. Aortic valve replacement in patients aged 80 years and older: early and long-term results. J Thorac Cardiovasc Surg 1996; 111:1026-36 18 The Society of Thoracic Surgeons National Cardiac Surgery Database. Manual for data managers-definitions of terms. Minneapolis: Summit Medical Systems, 1995 19 Vital Statistics of the United States, 1992. Life-tables (volume II, section 6). Hyattsville, Md: US Dept of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Center for Health Statistics, 1996 20 Braunwald E. Valvular heart disease. In: Braunwald E, ed.
894
21
22
23
24
25
26
27
28
29
30
31 32 33 34 35
36
Heart disease: a textbook of cardiovascular medicine. Philadelphia: WB Saunders, 1997; 1007-76 Kirklin JW, Barratt-Bayes BG. Mitral valve disease vvith or without tricuspid valve disease. In: Kirklin JW, Barratt-Bayes BG, eds. Cardiac surge1y. New York: Churchill Livingstone, 1993; 425-89 Kirklin JW, Barratt-Bayes BG. Aortic valve disease. In: Kirklin JW, Barratt-Bayes BG, eds. Cardiac surgery. New York: Churchill Livingstone, 1993; 491-571 Society of Thoracic Surgeons. Data analysis of the STS national cardiac surgery database: the sixth year. Chicago: Society of Thoracic Surgeons, 1997; 94-167 Flameng WJ, Herijgers P, Scecsi J, et a!. Determinants of early and late results of combined valve operations and coronary artery bypass grafting. Ann Thorac Surg 1996; 61:621-28 Verheul HA, van den Brink RB, Bouma BJ, et al. Analysis of 1isk factors for excess mortality after aortic valve replacement. J Am Coli Cardiol 1995; 26:1280-86 Shapira OM , Aida GS, Zelingher J, et a!. Enhanced blood conservation and improved clinical outcome after valve surgery using heparin-bonded cardiopulmonary bypass circuits. J Cardiac Surg 1996; 11:307-17 Aldea GS, Doursounian M, O'Gara P, et al. Heparin-bonded circuits with a reduced anticoagulation protocol in primary CABG: a prospective randomized study. Ann Thorac Surg 1996; 62:410-18 Borkon AM, Soule LM, Baughman KL, et a!. Aortic valve replacement in the elderly patient. Ann Thorac Surg 1988; 46:270-77 Jamieson WRE, Burr LH, Munro AI, et a!. Cardiac valve replacement in the elderly: clinical performance of biological prostheses. Ann Thorac Surg 1989; 48:173-85 Chambers LW, MacDonald LA, Tugwell P, et a!. The McMaster H ealth Index Questionnaire as a measure of quality of life for patients with rheumatoid disease. J Rheumatol 1982; 9:780-84 Bergner M, Bobbitt RA, Carter \VB, et a!. The Sickness Impact Profile: development and final revision of a health status measure. Med Care 1981; 19:787-805 Wan TT, Livieratos B. Interpreting a general index of subjective well-being. Milbank Mem Fund Q Health Soc 1978; 56:531-56 Stewart AL, Hays RD, Ware JE. The MOS short-form health survey: reliability and validity in patient population. Med Care 1988; 26:724-35 Lavie CJ, Milani RV. Effects of cardiac rehabilitation and exercise training programs in patients ?.75 years of age. Am J Cardia! 1996; 78:675-77 Bubien RS, Knotts-Dolson SM, Plumb VJ. Effect of radiofrequency catheter ablation on health-related quality of life and activities of daily living in patients \vith recurrent arrhythmias. Circulation 1996; 94:1585-91 Hlatky MA, Rogers WJ, Johnstone I, eta!. Medical care costs and quality of life after randomization to coronary angioplasty or coronary bypass surgery. N Eng! J Med 1997; 336:92-99
Clinical Investigations