| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1996;61:621-628
© 1996 The Society of Thoracic Surgeons
Department of Cardiac Surgery, Katholieke Universiteit Leuven, Leuven, Belgium
Accepted for publication September 27, 1995.
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
Methods. Determinants of early and late (more than 90 days) deaths and event-free survival were studied for combined valve operations and coronary artery bypass grafting in 741 patients using multivariate analysis.
Results. Ninety-day survival probability was 89% (95% confidence interval, 87% to 92%). Preoperative risk factors for early death were age, female sex, renal failure, New York Heart Association class IV or V, and mitral insufficiency. The operative risk factor was the duration of aortic cross-clamping. Five- and 10-year survival probabilities were 74% (95% confidence interval, 71% to 78%) and 43% (95% confidence interval, 36% to 50%), respectively. Preoperative risk factors for late death were age, preoperative renal failure, New York Heart Association class IV or V, vessel disease, and nonsinus rhythm. Five- and 10-year event-free survival probabilities were 57% (95% confidence interval, 53% to 61%) and 23% (95% confidence interval, 17% to 28%), respectively. Preoperative risk factors for non–event-free survival were age, female sex, reduced left ventricular function, mitral regurgitation, and pacemaker rhythm.
Conclusion. The demographic factors of age and female sex; the comorbid condition of renal failure; the cardiac conditions of advanced New York Heart Association class, left ventricular function, mitral regurgitation, vessel disease, and cardiac rhythm; and the operative condition of ischemia time are the most important predictors of clinical outcome after combined valve operations and coronary artery bypass grafting.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
In this study, we analyzed our results at the Katholieke Universiteit Leuven for the years 1980 to 1992 and describe the determinants of early, late, and event-free survival after combined valve operations and CABG in 741 patients, using multivariate analysis. Preliminary results of a subset of 420 patients operated on from 1984 to 1990 were reported earlier [9].
|
Material and Methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
.
|
For the valve replacements, a biologic valve was used in 39% of cases and a mechanical valve in 61%. When the aortic valve was replaced, the distribution of biologic and mechanical valves was 43% and 57%, respectively. When the mitral valve was replaced, we used predominantly a mechanical valve (72%). In multiple valve replacement, biologic valves were used in 38% of instances and mechanical valves in 59%; in 2 patients (3%), the tricuspid valve was replaced by a biologic valve in combination with a mechanical valve in the mitral position. The following types of biologic valves were used: Ionescu Low Profile (Shiley Inc, Irvine, CA), Mitroflow (Mitral Medical of Canada, Ltd, Richmond, Canada), and Carpentier-Edwards Pericardial (Baxter Healthcare Corp, Santa Ana, CA). The mechanical valves used included the Björk Shiley Convex-Concave (Shiley Inc), Björk Shiley Monostrut (Shiley Inc), St. Jude Medical (St. Jude Medical Inc, St. Paul, MN), Medtronic-Hall (Medtronic Inc, Minneapolis, MN), and CarboMedics (CarboMedics Inc, Austin, TX).
Pharmacologic or mechanical support was initiated during weaning according to standard clinical practice, and all patients were admitted to the intensive care unit. They were transferred to the ward when respiratory function was sufficient, no more inotropic support was needed, and cardiac rhythm and noncardiac organ function had stabilized.
All patients were initially given long-term anticoagulation therapy with fenprocoumon (Marcoumar; Roche, Brussels, Belgium). In the case of biologic valve prosthesis or valvuloplasty, coumarin administration was discontinued between 6 weeks and 6 months postoperatively when sinus rhythm was stable and no other indication for anticoagulation therapy was present. All of the patients were given platelet-inhibiting drugs.
Variables Registered
Preoperative, operative, and in-hospital postoperative variables were found in the patient files for patients operated on before 1990. Thereafter, results were directly entered in the database and double checked from the patient file after discharge.
Follow-up information was obtained in most instances by a regular follow-up report provided by the referring cardiologist; in other cases we used a questionnaire mailed to the patient, telephone review, or examination in our offices. Follow-up of hospital survivors had a median duration of 38.3 months (range, 1 to 157.4 months) and was 98% complete.
PREOPERATIVE VARIABLES.
Preoperative data included the following: age, sex, associated diseases (insulin-dependent diabetes mellitus, dialysis-dependent renal failure, respiratory failure requiring artificial ventilation), previous cardiac operations, New York Heart Association (NYHA) class, valve lesion (stenosis, regurgitation, mixed), cause of valve disease (rheumatic, degenerative, ischemic, congenital, others), extent and symptoms of coronary artery disease, left ventricular ejection fraction, left ventricular regional wall motion, end-diastolic left ventricular pressure, cardiac rhythm, rhythm disturbances, and arrhythmias. Coronary artery disease was defined as 50% or greater narrowing of diameter in a coronary artery. The extent of coronary artery disease was determined by the number of diseased arteries (single, double, or triple vessel); left main stenosis was categorized separately but was regarded as at least a double-vessel disease.
Left ventricular function was scored according to the available data of ejection fraction (EF), segmental wall motion, and left ventricular end-diastolic pressure (LVEDP). Segmental wall motion was assessed from the preoperative left ventricular angiogram and analyzed according to Leighton and associates [11]. Possible scores were: (1) normal, at EF greater than 0.50 or, when EF was not available, LVEDP less than 15 mm Hg in combination with normal segmental wall motion; (2) mild impairment, at EF between 0.40 and 0.50 or, when not available, mild impairment of segmental wall motion in combination with LVEDP less than 20 mm Hg; (3) moderate impairment, at EF between 0.30 and 0.40 or moderate impairment of segmental wall motion combined with LVEDP between 15 and 25 mm Hg; and (4) severe impairment, at EF less than 0.30 or severely impaired segmental wall motion combined with LVEDP greater than 25 mm Hg. The preoperative variables are listed in Table 2.
|
FOLLOW-UP VARIABLES.
Death was classified as described above. All postoperative events were registered according to the published guidelines [12] as valve-related complications (transient ischemic attack, cerebrovascular accident, peripheral embolism, anticoagulation therapy–related hemorrhage, endocarditis), ischemic complications (myocardial infarction, return of angina), or NYHA functional class III and IV. Data were gathered about reinterventions (percutaneous transluminal coronary angioplasty, repeat CABG, transplantation).
Statistical Methods
Three points of interest were studied separately. Early mortality rate was defined as death within 90 days after operation, both in the hospital and after discharge. The late mortality rate included all later deaths. For event-free survival, the first event in a patient (death of a major complication) was taken as the end point. The influence of preoperative and intraoperative variables on these three outcome measures was assessed by univariate and multivariate analyses.
For univariate analysis of the 90-day mortality rate, we used a Pearson 
2 test or Fisher's exact test (depending on appropriateness of the 2 test) to analyze differences between groups. Normal continuous data were compared using Student's t test, and abnormal data by the Wilcoxon rank-sum test. Logistic regression was performed for the multivariate analysis of 90-day mortality rate. Furthermore, after censoring all patients surviving more than 90 days, we applied the Cox proportional hazards model separately on the mortality data for the first 90 days after operation to be able to compare the risk ratios of the predictive factors for 90-day mortality with those for late mortality.
The Kaplan-Meier method was used to estimate the late and event-free survival probabilities. The Wilcoxon test statistic for censored data was calculated to compare survival curves between different groups. For the multivariate analysis of late and event-free survival, we used the Cox proportional hazards model.
The analysis was stratified by surgeon. Age and type of diseased valve (aortic, mitral, tricuspid valve, or combinations) were fixed in the model. All preoperative and intraoperative factors having p values less than 0.10 by univariate analysis were subsequently entered as possible risk factors. Several equivalent models could be produced; however, the different predictive factors in the models were highly correlated and often gave redundant information. Therefore, we present the model with the most meaningful indices for clinicians. Factors were considered significant predictors at p less than 0.05.
Different goodness-of-fit tests—mainly graphic tests—were used to check the proposed model. Graphic tests were: (1) log-log survival curves to test the proportional hazards assumption of each individual determinant; (2) Kay's residuals [13] to test the appropriateness of the full model; and (3) Arjas' method [14] to check the agreement between the predicted and the observed mortality rates. Furthermore, the S statistic as proposed by Moreau and co-workers [15] and its p value were calculated. No major deviations were seen.
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
.
|
). Two phases could be discerned: an early one (up to 3 months postoperatively) with high mortality rates, and a late one (more than 3 months after operation) when mortality rates were more stable. Therefore, predictive factors for death were analyzed separately for these two time periods and not for overall hospital mortality.
|
.
|
Therefore, the model consisted of the following independent risk factors: age, female sex, renal failure, NYHA class IV or V, mitral regurgitation, and aortic cross-clamp time (Table 5). Next we searched for the influence of factors related to the valve prosthesis used. The type of prosthesis (biologic/mechanical) was not found to be an independent factor. Our finding in the univariate analysis that the biologic Carpentier-Edwards aortic valve prosthesis bears a higher risk is related to its use at an advanced age, as evidenced by its lack of significance when placed together with age in a Cox regression model. Regarding the size of the prosthesis, we found in the univariate analysis that an aortic valve prosthesis smaller than size 23 and a mitral valve prosthesis smaller than size 29 bear a greater risk. This was not confirmed by multivariate analysis, probably because females received smaller valves and female sex was already an independent risk factor included as such in the model.
|
Using this model, we predicted 90-day survival in 3 hypothetical patients. The first patient was a 65-year-old male subjected to aortic valve replacement for aortic valve stenosis and three grafts for three-vessel disease. He was in NYHA functional class II and had no renal failure; valve replacement and coronary grafting were performed within 60 minutes of aortic cross-clamping. The second patient was a 75-year-old woman who underwent aortic valve replacement for aortic valve insufficiency combined with three grafts for three-vessel disease. Her NYHA functional class was IV, she had normal kidney function, and the aorta was cross-clamped for 90 minutes during the operation. The third patient was also a 75-year-old woman undergoing mitral valve replacement for severe mitral insufficiency and a triple bypass for three-vessel disease. She was in NYHA functional class IV, had atrial fibrillation, and had normal kidney function. The aortic cross-clamp time was 120 minutes. Patient 1 had a 3-month expected survival of 98%, patient 2 of 79%, and patient 3 of 39%.
LATE DEATHS.
A univariate analysis was performed on clinical preoperative indices. The same variables were considered as for the 90-day follow-up. Significant factors included age, preoperative NYHA functional class, left ventricular function score, preoperative cardiac rhythm, vessel disease, preoperative renal failure, the type of valve lesion, the surgeon, the type of valve procedure, the duration of aortic cross-clamping, and the duration of cardiopulmonary bypass. The results are presented in Table 4
.
Next we performed a multivariate analysis using the proportional hazards regression technique. We started with stratification according to the surgeon and placed age in the model. Then we examined the importance of the type of valve involved. Solitary aortic valve replacement was taken as baseline, and we found by modeling that mitral valve involvement yielded a higher risk; involvement of both the aortic and mitral valves gave even a higher risk. The valve lesion then was examined, using aortic valve stenosis as baseline. Results showed that mitral valve insufficiency (and, equally well, mitral valve insufficiency plus stenosis) was an independent risk factor for late survival. Next, other covariables (which were significant factors when tested univariately) were added and tested. It was found that vessel disease, nonsinus rhythm, and renal failure were independent determinants. The NYHA functional class IV or V was also significant, but when this covariable was added to the model, mitral insufficiency (or mitral insufficiency plus stenosis) lost its significance. These last two factors were obviously interrelated. The final model for late mortality rate included age, NYHA functional class IV or V, vessel disease, nonsinus rhythm, and renal failure (see Table 5).
This model was used to predict late deaths in the same patients as presented before when we assessed the 90-day mortality rate. For patient 1, this value was 93% at 5 years; for patient 2, it was 66%; and for patient 3, it was 61%, provided that they had already survived the first 90 days.
DETERMINANTS OF EVENT-FREE SURVIVAL.
In a univariate analysis, the following factors were found to be significantly related to postoperative events: age, sex, preoperative NYHA functional class, preoperative left ventricular function score, cardiac rhythm, renal failure, valve involvement (aortic, mitral, combined), cause of valve disease, valve procedure (solitary, multiple), surgeon, and duration of cardiopulmonary bypass (see Table 4).
A multivariate analysis was performed using the proportional hazards regression technique. To build the model, we stratified according to the surgeon; then age, aortic insufficiency, and mitral insufficiency were fixed in the model first; then covariables that were statistically significant in the univariate analysis were added and tested. The following independent predictors resulted: age, mitral insufficiency, female sex, moderately or severely depressed preoperative left ventricular function, and pacemaker rhythm (see Table 5).
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
When looking at the Kaplan-Meier survival estimates in our group of combined procedures, one recognizes that the steepest drop in survival occurred during the first 3 months (to 89%). The decrease in survival then declined more gradually to 43% at 10 years. The event-free probability showed a similar course, although the rate of decline was faster in both phases. Therefore, in the search for risk factors determining the mortality rate, we analyzed both phases separately: an ``early'' phase (ie, the first 90 days) and a ``late'' phase (ie, beyond 90 days postoperatively).
Independent risk factors for early death could be divided into risk factors related to patient conditions, comorbid conditions, cardiac conditions, and operative conditions. As for patient-related conditions, advanced age and female sex were found. This is not surprising, because both are known risk factors that are consistently found when determining the hospital mortality rate after adult cardiac operations in general [16–18]. The same holds true for the comorbid condition of renal failure. Renal dysfunction has recently been shown to be an important risk factor for operative death in patients who have this condition [16–21]. Renal dysfunction was defined either as elevated creatinine levels or as dependence on dialysis, and we found it to have a risk ratio as high as 3.4. Although in all other studies, the reported odds ratio was also high, the incidence of dialysis dependency in the cardiac operative population is low [17], and therefore it is unlikely to change the results of any population-based comparisons [21].
As cardiac conditions predicting early death, we found advanced NYHA functional class and mitral valve insufficiency to be independent risk factors. In our model, NYHA functional class could be replaced by the left ventricular function score, which was based on a combination of EF, regional wall motion, and LVEDP. Although NYHA functional class is subject to observer bias, we preferred this index because angiographic and hemodynamic data may not always be available and because the NYHA functional class can be considered a reflection of left ventricular function. Factors related to left ventricular function such as ejection fraction or left ventricular dysfunction estimates are consistently found to be independent risk factors for hospital death in cardiac or coronary operations [16–19, 22]. Although all the above risk factors for early death seem to be nonspecific for combined valve and coronary operations, mitral insufficiency might well be specific. Higgins and colleagues [19] identified mitral insufficiency as a risk factor associated with perioperative mortality and morbidity in patients undergoing CABG, with an odds ratio of 2.53 for mortality and of 2.59 for morbidity. Kirklin and associates [23] pointed out that ischemic mitral valve disease is a significant risk factor for early death after primary combined valvular and coronary artery operations.
Other cardiac conditions such as vessel disease and cardiac rhythm disturbances were not recognized as independent predictors of early death in our study. Also in Kirklin's study [23], only left main disease was found to be an independent risk factor for early death when coronary morphology was considered.
As mentioned earlier, we also analyzed the risk factors influencing late death, that is, in the period beyond 90 days postoperatively. Of the patient-related conditions, only age remained as an independent predictor of late death; female sex was no longer a determinant. In addition, the comorbid condition of renal failure remained as a risk factor for late death. Concerning the cardiac conditions, advanced NYHA functional class persisted as an independent risk factor for late death. Most investigators studying either aortic or mitral valve replacement combined with CABG found some index of cardiac function that predicted the outcome: either preoperative NYHA class [3], LVEDP [2], wall motion score [1], EF [2], cardiothoracic index, or signs of heart failure [24]. Once the early postoperative phase had passed, mitral valve insufficiency was no longer an independent risk factor for death. This was also reported by Davis and associates [25], studying valvular disease in the elderly. These authors found that among patients undergoing mitral valve replacement, NYHA functional class, ischemic valvular disease, and increased wedge pressure were predictive of operative death, but only poor left ventricular function was a predictor of poor long-term survival [25]. Other cardiac-related conditions, however, became important for late survival: vessel disease and cardiac rhythm. Perhaps progression of coronary artery disease associated with the preferential use of venous grafts is responsible for the influence of coronary status on late survival [26]. In the report of Czer and colleagues [2], predictors of late death after mitral valve replacement and coronary revascularization were left main coronary artery disease or severe triple-vessel disease, besides depressed left ventricular function and high end-diastolic volume at the time of catheterization. Miller and associates [27] documented that long-term survival after mitral valve replacement was compromised by the presence of substantial coronary artery disease and that the adverse effect of coronary disease was not completely ameliorated by bypass grafting. On the other hand, Andrade and co-workers [1] denied this and stated that once coronary artery disease was corrected in patients undergoing mitral valve replacement, survival did not differ from that of patients with isolated mitral valve replacement. In aortic valve replacement combined with myocardial revascularization, Lytle and associates [28] demonstrated that the late survival rate was unfavorably influenced by the presence of double-vessel disease and impaired left ventricular function.
In the present study, we found that preoperative nonsinus rhythm was an independent predictor of late death. Aranki and associates [29] have described atrial fibrillation or heart block as a predictor of operative death in elderly patients undergoing aortic valve replacement with or without CABG.
The operation-related condition—the duration of aortic cross-clamping—was a predictor not of late death, but of early death. This was also recognized for aortic valve replacement with concomitant CABG by Ståhle and colleagues [8], who even suggested restricting the number of distal anastomoses to shorten the aortic cross-clamp time.
Concerning the analysis of risk factors influencing event-free survival, we studied the postoperative course as one phase and did not divide it into an early and a late phase as we did for survival probability. The main reason was that the majority of events during the first 90 days postoperatively were cardiac deaths, so that a separate analysis of early event-free survival would not provide additional information. We also did not differentiate in the analysis between specific events and in fact considered event-free survival as the probability of being completely symptom free. We documented that less than 60% of the patients undergoing combined valve operations and CABG lived 5 years without experiencing an event. Independent risk factors for non–event-free survival included most of the risk factors for early and late death: age, female sex, impaired left ventricular function, mitral valve insufficiency, and nonsinus rhythm, or, more specifically, pacemaker rhythm. Some factors found in the analysis of early and late death, however, were not found to be independent risk factors for non–event-free survival. Renal failure was one of them, but, as mentioned before, the incidence of this feature was low in our population. In addition, the duration of aortic cross-clamping during the operation did not play a significant role in event-free survival, as only early death is related to it. Surprisingly, coronary vessel disease, which determines late death, was also not an independent predictor of event-free survival. This is in contrast to the findings of Lytle and associates [24] in patients undergoing mitral valve replacement combined with CABG, but the same authors could not detect vessel disease as an independent predictor of event-free survival in patients undergoing aortic valve replacement and revascularization [30].
In summary, we conclude that patient-related conditions (age and female sex), the comorbid condition of renal failure, cardiac conditions (advanced NYHA functional class, impaired left ventricular function, mitral valve insufficiency, coronary vessel disease, and nonsinus rhythm), and the operative condition of myocardial ischemia time are the most important predictors of clinical outcome after combined valve operations and CABG.
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. J. Jokinen, M. J. Hippelainen, T. Hanninen, A. K. Turpeinen, and J. E.K. Hartikainen Prospective assessment of quality of life of octogenarians after cardiac surgery: factors predicting long-term outcome Interactive CardioVascular and Thoracic Surgery, October 1, 2008; 7(5): 813 - 818. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Jokinen, M. J. Hippelainen, O. A. Pitkanen, and J. E.K. Hartikainen Mitral Valve Replacement Versus Repair: Propensity-Adjusted Survival and Quality-of-Life Analysis Ann. Thorac. Surg., August 1, 2007; 84(2): 451 - 458. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Jones, D. Lovell, G. W. Cran, and S. W. MacGowan Impact of coronary artery bypass grafting on survival after aortic valve replacement Interactive CardioVascular and Thoracic Surgery, June 1, 2006; 5(3): 327 - 330. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Doenst, J. Ivanov, M. A. Borger, T. E. David, and S. J. Brister Sex-specific long-term outcomes after combined valve and coronary artery surgery. Ann. Thorac. Surg., May 1, 2006; 81(5): 1632 - 1636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. J. Flameng, P. Herijgers, S. Dewilde, and E. Lesaffre Continuous retrograde blood cardioplegia is associated with lower hospital mortality after heart valve surgery J. Thorac. Cardiovasc. Surg., January 1, 2003; 125(1): 121 - 125. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. A. Thistlethwaite, M. Mo, M. M. Madani, R. Deutsch, D. Blanchard, D. P. Kapelanski, and S. W. Jamieson Operative classification of thromboembolic disease determines outcome after pulmonary endarterectomy J. Thorac. Cardiovasc. Surg., December 1, 2002; 124(6): 1203 - 1211. [Abstract] [Full Text]
|
|
|
|
 |
|
 R. G. Seipelt, F. A. Schoendube, J. F. Vazquez-Jimenez, H. Doerge, M. Voss, and B. J. Messmer Combined mitral valve and coronary artery surgery: ischemic versus non-ischemic mitral valve disease Eur. J. Cardiothorac. Surg., August 1, 2001; 20(2): 270 - 275. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Fasol, F. Lakew, B. Pfannmuller, M. J. Slepian, and E. Joubert-Hubner Papillary muscle repair surgery in ischemic mitral valve patients Ann. Thorac. Surg., September 1, 2000; 70(3): 771 - 777. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Iung VALVE DISEASE: Interface between valve disease and ischaemic heart disease Heart, September 1, 2000; 84(3): 347 - 352. [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Eagle, R. A. Guyton, R. Davidoff, G. A. Ewy, J. Fonger, T. J. Gardner, J. P. Gott, H. C. Herrmann, R. A. Marlow, W. C. Nugent, et al. ACC/AHA guidelines for coronary artery bypass graft surgery: A report of the American College of Cardiology/ American Heart Association task force on Practice Guidelines (Committee to revise the 1991 Guidelines for Coronary Artery Bypass Graft Surgery) J. Am. Coll. Cardiol., October 1, 1999; 34(4): 1262 - 1347. [Full Text] [PDF]
|
|
|
|
|
|
E. Weinschelbaum, P. Stutzbach, M. Oliva, J. Zaidman, A. Torino, and E. Gabe MANUAL DEBRIDEMENT OF THE AORTIC VALVE IN ELDERLY PATIENTS WITH DEGENERATIVE AORTIC STENOSIS J. Thorac. Cardiovasc. Surg., June 1, 1999; 117(6): 1157 - 1165. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Cohen, T. E. David, J. Ivanov, S. Armstrong, and C. M. Feindel THE IMPACT OF AGE, CORONARY ARTERY DISEASE, AND CARDIAC COMORBIDITY ON LATE SURVIVAL AFTER BIOPROSTHETIC AORTIC VALVE REPLACEMENT J. Thorac. Cardiovasc. Surg., February 1, 1999; 117(2): 273 - 284. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. T. Bajzer, W. J. Stewart, D. M. Cosgrove, S. J. Azzam, K. L. Arheart, and A. L. Klein Tricuspid valve surgery and intraoperative echocardiography: Factors affecting survival, clinical outcome, and echocardiographic success J. Am. Coll. Cardiol., October 1, 1998; 32(4): 1023 - 1031. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
