Hypothesis: Renal insufficiency (RI) is associated with an increased risk of morbidity and mortality following coronary artery bypass graft (CABG) operations, particularly among patients who are dependent on dialysis.
Design and setting: A retrospective analysis of data collected at a tertiary care center. Patients: One hundred eighty-four consecutive patients with RI who underwent CABG surgery between 1992 and 2004. This group consisted of 152 patients with serum creatinine levels of [greater than or equal to] 1.7 mg/dL (group I) and 32 kidney transplant recipients (group II). Of the patients in group I, 90 were dialysis-free (subgroup IA) and 62 were dialysis-dependent (subgroup IB). Main outcome measures: Demographics, perioperative data, and outcomes for each of the three groups were evaluated and compared.
Results: Fifty-four percent of the patients were in New York Heart Association classes III and IV, 36% had unstable angina, and 21% had left main coronary disease. The mean ejection fraction was 38%. The median postoperative length of stay in the hospital was 10 days. Of the patients in group IB, 8% required reexploration for bleeding compared to 3% in groups IA and II (p < 0.05). Dialysis was needed postoperatively in five patients in group IA and two patients in group II (5.7%). The raw operative mortality rate was 7.6% and was higher in group IB (9.7%) compared to groups IA and II (6.7% and 6.2%, respectively; p < 0.05). The actuarial 5-year survival rate was higher in group II compared to group I (79% vs 59%, respectively; p < 0.05). The difference in survival rates was more apparent between groups II and IB (79% vs 57%, respectively; p < 0.005).
Conclusions: CABG is associated with an increased rate of perioperative complications and mortality in patients with RI. Dialysis dependence is a major risk factor for patients undergoing CABG surgery. However, with acceptable surgical results, dialysis patients should not be denied CABG surgery. A survival advantage is demonstrated among patients with previous kidney transplants compared to those patients who are dependent on dialysis.
Key words: coronary bypass; dialysis; renal insufficiency; renal transplant.
Abbreviations: CABG = coronary artery bypass graft; CPB = cardiopulmonary bypass; IABP = intraaortic balloon pump; NYHA = New York Heart Association; PD = peritoneal dialysis; RI = renal insufficiency
Coronary artery disease remains a major cause of morbidity and mortality in patients with renal insufficiency (RI). Patients with decreased renal function who undergo coronary artery bypass graft (CABG) operations carry a significant operative risk and require prolonged hospital care. With the increase in the aging population, more patients with RI are being referred for CABG surgery. Dialysis-dependent patients have accelerated atherosclerosis and, as a result, have decreased survival when they receive a diagnosis of coronary artery disease. Those that present for CABG surgery and require the use of cardiopulmonary bypass are a challenging group since they are prone to have a bleeding diathesis as well as fluid and electrolyte imbalances during the perioperative period.
The aims of this study were (1) to define the preoperative demographics and intraoperative characteristics of patients with RI undergoing CABG surgery, (2) to analyze and compare the early results of CABG surgery in patients who are dialysis-free or dialysis-dependent and those who had undergone prior kidney transplantation, and (3) to evaluate the impact of RI on the intermediate and long-term outcomes of patients in this population.
MATERIALS AND METHODS
We retrospectively analyzed the medical records of 184 consecutive patients with RI who underwent CABG surgery over a 12-year period between January 1992 and June 2004 at our center. These consisted of 152 patients with serum creatinine levels of [greater than or equal to] 1.7 mg/dL (group I) and 32 kidney transplant recipients who underwent CABG surgery at a mean period of 17 months (range, 1 to 73 months) following their renal transplants (group II). Of the patients in group I, 90 patients were dialysis-free (subgroup IA) and 62 were dialysis-dependent (subgroup IB). The following demographic and clinical parameters for patients were analyzed: age; sex; race; risk factors; cardiac history; New York Heart Association (NYHA) class; coronary arteries involved; ventricular function; and cardiac medications received. The procedural data collected included the duration of cardiopulmonary bypass (CPB), the aortic cross-clamp time, the number and types of grafts used, the concomitant procedures, the need for inotropes and intraaortic balloon pump (IABP), and blood use. The postoperative data collected included the duration of mechanical ventilatory support, length of ICU stay, length of postoperative hospital stay, postoperative complications, hospital readmissions, operative mortality, and survival. The data were collected using the hospital records and the institutional data reported to the Society of Thoracic Surgeons database. Follow-up was accomplished by reviewing our computerized databases of thoracic surgical and abdominal organ transplants as well as the hospital and outpatient clinic records, and by communicating with the referring primary physician or specialist. The National Social Security Death Index (www.ancestry.com/search/rectype/ vital/ssdi/main.htm) was used to confirm the date of death among the nonsurvivors. Mortality data were available on all patients. The Social Security Death Index was queried in June of 2004, and patients who were not found in the index were assumed to be alive (ie, free from all-cause mortality) at that time.
Coronary revascularization was performed through a median sternotomy using CPB with membrane oxygenation and moderate systemic hypothermia (temperature range, 28 to 32[degrees]C). Warm antegrade and retrograde blood cardioplegia was used for myocardial preservation in all cases. Anticoagulation was achieved prior to CPB with IV heparin (300 to 400 U per kg body weight) to achieve activated dotting times of > 400 s. All patients received aprotinin consisting of a 2-million KIU IV loading dose, a 2-million KIU dose in the pump prime volume, and 500,000 KIU per hour during surgery as a continuous IV infusion. Desmopressin acetate (0.3 [micro]g/kg IV) was used selectively after the termination of CPB when clotting was not achieved following the reversal of heparin with protamine. All patients received antibiotic prophylaxis consisting of 1.5 g of cefuroxime IV 30 min before the incision was made and another 1.5-g dose after separation from the CPB machine. Patients who were allergic to penicillin and/or cephalosporin received 1 g of vancomycin prior to the incision being made.
Renal Management Protocol
The renal management protocol for the dialysis-dependent patients consisted of maintenance hemodialysis (HD) on the day before surgery, large-volume hemofiltration intraoperatively toward the end of CPB, and postoperative serum potassium management with IV infusions of dextrose-insulin-bicarbonate or with kayexalate (25 to 50 g) administered either orally or as a rectal enema. Maintenance HD was resumed 12 to 24 h after the procedure to correct hyperkalemia, metabolic acidosis, and/or signs of fluid overload (pulmonary capillary wedge pressure, > 20 mm Hg) and when signs of pulmonary function deterioration were noted. Patients who were receiving peritoneal dialysis (PD) preoperatively were switched to HD the day before undergoing surgery through a percutaneous venous access and resumed PD after hospital discharge. Patients who did not have stable hemodynamics perioperatively and/or were receiving large doses of pressors and who did not tolerate HD were placed on continuous venovenous hemofiltration until they had been weaned from pressors and their hemodynamics had stabilized.
Numeric variables are reported as the mean [+ or -] SD or as median values whenever indicated. Categoric patient and hospital characteristics were compared using the [chi square] test, and continuous characteristics were compared using the Student t test. Univariate logistic regression analysis was performed to identify, among the perioperative variables evaluated, the risk factors for mortality at 1 year. Actuarial survival, which is reported as freedom from all-cause mortality, was determined using the Kaplan-Meier method, and a statistical comparison between the groups was calculated by the log rank test. Statistical analysis was performed using a statistical software package (SAS, version 8.2; SAS Institute; Cary, NC). Differences in survival were considered to be statistically significant when the p value was < 0.05.
Table 1 shows a summary of the demographics in the two subgroups with RI (subgroups IA and IB) and in group II. The entire patient cohort consisted of 115 men (63%) and 68 women (37%). The mean age for the patients was 62 years. Almost half of the patients were African-American (48%), 27% were white, and 21% were Hispanic. African Americans constituted 52% of the nontransplanted patients (group I) and 28% of those who had received transplants (group II). In contrast, whites constituted only 24% of the nontransplanted patients but were the leading group among those who had received transplants (38%). Among the risk factors, hypertension was the most prevalent (90%), followed by diabetes (66%), cigarette smoking (65%), hypercholesterolemia (58%), peripheral vascular disease (55%), chronic obstructive lung disease (52%), and cerebrovascular disease (20%). The preoperative risk factors analyzed were less prevalent among patients in group II. Of the patients in subgroup IB, 57 (92%) received HD and 5 (8%) received PD preoperatively. The mean duration of dialysis was 41 months (range, 2 to 98 months).
Table 2 shows the cardiac history and profile as well as the intraoperative data in each of the three groups. Twenty-one percent of the patients had significant left main coronary artery stenosis (ie, > 50% diameter narrowing). The overall mean left ventricular ejection fraction was 38%. More patients in the dialysis-free subgroup (subgroup IA) had a history of myocardial infarction or heart failure. Likewise, more patients in subgroups IA and IB were in NYHA classes III to IV compared to group II (63% and 50% vs 38%, respectively). Over half of the patients (59%) in subgroup IA had undergone urgent or emergent operations compared to 26% and 28% in subgroup IB and group II, respectively (p < 0.05). The mean number of coronary grafts was three; 76% of the patients had received at least one internal mammary artery graft. Twenty-nine patients (16%) had undergone concomitant valve operations consisting of mitral valve repair or replacement in 16 patients (9%), aortic valve replacement in 9 patients (5%), and tricuspid valve annuloplasty in 4 patients (2%). Pressors were used perioperatively in 36% of the patients, and an IABP was used in 11% of the patients. The mean hemoglobin level at hospital admission was 12 mg/dL. Blood product transfusions were administered perioperatively in 66% of the patients, and significantly more were administered in subgroups IA and IB than in group II.
Table 3 shows the postoperative data and the operative morbidity and mortality of the patients in each of the three groups. The poststernotomy reexploration rate for bleeding complications was 4.9% (subgroup IB, 8%; subgroup IA, 3%; subgroup II, 3%; p < 0.05). Twenty-four percent of the patients in subgroup IB required > 24 h of mechanical ventilation compared to 10% and 9% of those in subgroup IA and in group II, respectively (p < 0.05). Postoperative cerebrovascular events were more common among group I patients (10% vs 3% in group II patients; p < 0.05). Postoperative HD was needed in seven patients in subgroup IA and group II (5.7%) [subgroup IA, five patients (5.6%); group II, two patients (6.2%)]. Dialysis was required on a permanent basis in four patients in subgroup IA (4%) and in one patient in group II (3%). The latter patient had experienced chronic rejection in the renal allograft prior to her CABG surgery (preoperative serum creatinine level, 2.5 mg/dL). The operative mortality rate among the seven patients in subgroup IA and group II who required postoperative dialysis was 14% (with one postoperative death from sepsis). There were two sternal wound infections in group I (one patient in subgroup IA and one patients in subgroup IB) and none in group II. Leg infections were more common among patients in group II (12%) compared to those in group I (8%), but the difference was not statistically significant. Likewise, septic complications occurred in 6% of group II and in 3% of group I (not statistically significant). Twenty-four patients (13%) required hospital readmission within 30 days of hospital discharge. The causes for hospital readmission were medical management of congestive heart failure in nine patients, management of postoperative atrial fibrillation in four patients, pericardial effusion requiting drainage in three patients, over-anticoagulation in three patients, pneumonia in three patients, uncontrolled insulin-dependent diabetes in one patient, and digoxin toxicity in one patient.
There were 14 early deaths for an overall raw operative mortality rate of 7.6%. The causes of early death were cardiac-related in 10 patients, sepsis with multisystem organ failure in 2 patients, pneumonia in 1 patient, and stroke in 1 patient. Table 4 shows the operative mortality rates by procedure type (ie, isolated CABG surgery vs combined CABG/valve surgery) and surgical era (from 1992 to 1996 vs from 1997 to 2004). Of the 155 patients who underwent isolated CABG operations (ie, with no concomitant valve operations), the operative mortality rate was 7.1%. The operative mortality rate among the 29 patients who had concomitant CABG and valve operations was 10.3%. A significant reduction in mortality rate for patients after isolated CABG surgery was observed among patients in all three groups who underwent surgery in the later surgical era compared to those in the earlier era. The mortality rate for patients who underwent isolated CABG surgery from 1997 to 2004 was 3.5%, and was 8.1% among patients in subgroups IA and IB with no operative deaths occurring among the patients in group II. Table 5 shows the risk factors for the mortality at 1 year after surgery among the variables analyzed. Six variables stood out as independent risk factors.
The mean follow-up period was 63 months (median, 58 months; range, 0 to 137 months). The 1-year, 3-year, and 5-year Kaplan-Meier survival rates for the entire patient cohort were 86%, 77%, and 62%, respectively. There was a 5-year survival advantage for patients who had undergone renal transplantation (group II) compared to patients who were dialysis-dependent (subgroup IB). For these two groups, the 5-year survival rates were 79% and 57%, respectively (p < 0.005). At the time that the National Social Security Death Index was queried in June 2004, there were 70 deaths (38%). These consisted of 33 deaths in subgroup IA (37%), 25 deaths in subgroup IB (40%), and 12 deaths in group II (38%). The causes of death among the 12 group II patients were identified through our computerized institutional renal transplant database. One death was attributed to a cardiac cause. The remaining 11 deaths were non-cardiac-related and included fulminant pneumonia with respiratory failure (4 patients), sepsis with multisystem organ failure (3 patients), malignancy (2 patients), and stroke (2 patients).
Cardiac disease continues to be a major cause of death in patients with RI and may be responsible for as much as 60% of all causes of death in renal patients in the United States. (1) Previous studies (2,3) have demonstrated an adverse operative outcome with surgical revascularization in patients who have impaired renal function, particularly those who are dialysis-dependent. Despite this increased risk, long-term studies (4) have demonstrated a benefit of surgical revascularization over percutaneous coronary intervention in both overall survival and angina-free survival. Still, patients who are dialysis-dependent continue to have a higher restenosis rate compared to the general population of patients undergoing revascularization, particularly those undergoing percutaneous coronary intervention. (5)
In our study, a survival advantage was noted at 5 years among patients who had previously received kidney transplants and had undergone CABG surgery over those patients who had RI and were dependent on dialysis. Impaired renal function has been shown to be an important predictor of postoperative mortality. (6,7) Patients with RI, particularly those who are receiving dialysis, are unable to cope with the fluid shifts in the different body compartments that are induced by CPB. (8) Furthermore, other comorbid conditions caused by atherosclerosis tend to increase the operative mortality rate in these patients. In our report, dialysis-dependent patients had a higher proportion of composite risk factors compared to those who were dialysis-free or those who had received kidney transplants. Advanced age, diabetes, and peripheral vascular disease have all been found to be significant independent predictors of decreased survival. (9)
Previous studies have shown that the presence of congestive heart failure (NYHA class IV) and older age are the most important predictors of hospital mortality. In our study, about two thirds of the dialysis-free patients (63%) and half of the dialysis-dependent patients were in NYHA classes III and IV compared to only 38% of those in the group of patients who had received renal transplants. The latter group of patients was more capable of handling fluid shifts and the fluid overload relating to congestive heart failure. Most patients with RI as a result of a hypertensive nephropathy demonstrate ventricular hypertrophy and subsequent subendocardial ischemia secondary to hypertension. (10) Similarly, diabetic and uremic patients develop a polyneuropathy and as such they tend to have silent angina despite significant atherosclerosis in their coronary circulation. (10,11)
Postoperative bleeding has long been recognized as a common problem in dialysis-dependent patients and, to a lesser extent, in patients with RI who are dialysis-free. In our study, five of the nine patients who required reexploration for postoperative bleeding were dialysis-dependent (8% of that subgroup). This compares favorably to the 7% reexploration rate reported by Franga et al (12) in a similar group of dialysis-dependent patients. Bleeding in these patients may be attributed to the qualitative platelet dysfunction as a result of the impaired release of von Willebrand factor. Over the past 5 years, we have been quite satisfied with the routine use of aprotinine and desmopressin acetate in dialysis-dependent patients and in patients with impaired renal function who are undergoing CABG surgery. We have had only four cases (3.2%) of postoperative bleeding that required returning to the operating room for sternal reexploration among the 126 renal patients who underwent surgery over this more recent period.
The overall raw operative mortality rate of 7.6% for patients who underwent isolated CABG surgery and combined CABG/valve surgery compares favorably to the operative mortality rates of 11 to 20% reported in previous studies. (12-15) The higher mortality rate of 9.7% in the dialysis-dependent patients may partly be a result of the decreased ability to fight infection due to decreased leukocyte chemotaxis combined with leukopenia and accelerated atherosclerosis. (16) Shroyer et al (17) analyzed the Society of Thoracic Surgeons database and found that an increase in the preoperative serum creatinine concentration to [greater than or equal to] 2.0 mg/dL was significantly associated with increased rates of postoperative death (hazard ratio, 2.01). Our operative mortality rate of 6.7% for the dialysis-free patients with RI also compares favorably to the 7% rate reported by Nakayama et al. (18) The low operative mortality rate of 6.2% in CABG surgery patients who had received renal transplants and the improved 5-year survival rate of 79% as well as encouraging data showing a reversal of left ventricle dysfunction following renal transplantation should provide a more aggressive stance in providing renal allografts to patients with end-stage renal disease.
The use of arterial grafts, particularly the internal mammary artery, has been shown to improve late outcomes in patients with renal disease who are undergoing CABG surgery. (18) In our study, only 76% of the revascularized patients had an arterial conduit, as over half of the patients were operated on either urgently or emergently. However, these numbers could be further improved on so as to provide arterial conduits to most patients when possible.
In our series, only 7 of 122 patients who were not receiving dialysis prior to undergoing their CABG operation (ie, subgroup IA and group II) required dialysis postoperatively (5.7%). Of those, only five patients required dialysis on a permanent basis (4.1%). The operative mortality rate in this cohort of patients who did not need to receive dialysis before surgery and who required postoperative dialysis (one of seven patients; 14%) compares favorably to the mortality rate of 10 to 20% that has been reported before. Advances in off-pump CABG surgery have shown (19) a tendency for a reduced likelihood of acute renal failure in patients with RI who are undergoing CABG surgery. However, there remains a tendency for possible incomplete revascularization with operations performed without the utilization of CPB.
Several limitations of the study need to be addressed. The study design was a retrospective observational one, although most of the data analyzed were prospectively collected and entered into the Society of Thoracic Surgery Network database. Our study was a single-center study that was carried out by a surgical group practicing in an academic training center, and that might have influenced some of the operative data, specifically, the times for aortic cross-clamping and CPB. It is likely that patient selection, the choice of procedures, and perioperative management may differ from those of other centers. Our institution is a referral center, and as such some of the patients were followed up by their cardiologists at their local hospitals and not at our outpatient center, which might have influenced some of the follow-up data. Furthermore, an increase in the serum creatinine level of > 50% of the normal value was chosen as a marker of renal dysfunction for simplicity of measurement rather than the creatinine clearance rate, which might be a more objective measure of renal function. The causes of death beyond the 30-day postoperative period and the first hospitalization were not analyzed among all the groups, specifically in subgroups IA and IB. Although the survival data obtained from the National Social Security Death Index database are quite accurate, with a reported sensitivity of 97% and a specificity of 99%, some patients who have died recently have not yet been entered into the database. (20) Finally, the improved outcome of CABG surgery in patients who had received kidney transplants could have been due to a selection bias. These patients usually have fewer comorbid conditions and are generally more compliant compared to those who are dialysis-dependent, many of whom would not have qualified for a transplant due to multiple comorbidities, substance abuse, and compliance. Immunosuppression therapy in patients who had received renal transplants did not have a significant impact on the rate of local or systemic infections compared to patients who had not received renal transplants and was not shown to be a risk factor for mortality at 1 year following CABG surgery.
In conclusion, preoperative RI increases the mortality and morbidity of patients undergoing surgical coronary revascularization. Its effects on patient outcomes and survival extend beyond the initial in-hospital stay and the 30-day postoperative period. A survival advantage was demonstrated among patients who had received kidney transplants who underwent CABG surgery compared to those who were dialysis-dependent. An important finding of our study is the fact that in the subgroups of patients with renal failure (subgroups IA and IB), particularly among patients who are dialysis-dependent (subgroup IB), a significant proportion of the deaths occurred outside the traditional 30-day period. Our results demonstrate the poorer outcome of dialysis-dependent patients compared to patients who are dialysis-free or those who have undergone renal transplants. Still, the long-term outcome of the dialysis-dependent patients who underwent myocardial revascularization is superior to the 19% annual mortality rate that has been reported before. (21)
(1) Burton BT, Krueger KK, Bryan FA Jr. A national registry of long term dialysis patients. JAMA 1971; 218:718-712
(2) Ko W, Kreiger KH, Isom OW. Cardiopulmonary bypass procedures in dialysis patients. Ann Thorac Surg 1993; 55: 677-684
(3) Owen CH, Cummings RG, Sell TL, et al. Coronary artery bypass grafting in patients with dialysis-dependent renal failure. Ann Thorac Surg 1994; 58:1729-1733
(4) Koyanagi T, Nishada H, Kitamura M, et al. Comparison of clinical outcomes of coronary artery bypass grafting and percutaneous coronary angioplasty in dialysis patients. Ann Thorac 1996; 61:1793-1796
(5) Khan JK, Rutherford BD, McConahay DR, et al. Short and long term outcome of percutaneous transluminal coronary angioplasty in chronic dialysis patients. Am Heart J 1990; 119:484-489
(6) Mangano CM, Diamondstone LS, Ramsey JG, et al. Renal dysfunction after myocardial revascularization: risk factors, adverse outcomes, and hospital resource utilization: the Multicenter Study of Perioperative Ischemia Research Group. Ann Intern Med 1998; 128:194-203
(7) Browner WS, Li J, Mangano DT. In hospital and long term mortality in male veterans following non cardiac surgery. JAMA 1992; 268:228-232
(8) Nakayama Y, Sakata R, Ura M, et al. Coronary artery bypass grafting in dialysis patients. Ann Thorac Surg 1999; 68:1257-1261
(9) Herzog CA, Ma JZ, Collins AJ. Poor long term survival after acute myocardial infarction among patients on long term dialysis. N Engl J Med 1998; 339:799-795
(10) Parfrey PS, Harnett JD, Barre PE. The natural history of myocardial disease in dialysis patients. J Am Soc Nephrol 1991; 2:2-12
(11) Foley RN, Parfrey PS. Cardiac disease in chronic uremia: clinical outcomes and risk factors. Adv Ren Replace Ther 1997; 4:234-248
(12) Franga DL, Kratz JM, Crumbley AJ, et al. Early and long term results of coronary artery bypass grafting in dialysis patients. Ann Thorac Surg 2000; 70:813-819
(13) Batiuk TD, Kurtz SB, Oh JK, et al. Coronary artery bypass operation in dialysis patients. Mayo Clin Proc 1991; 66:45-53
(14) Rostand SG, Rutsky EA: Cardiac disease in dialysis patents. In: Nissenson AR, Fine RN, Gentile DE, eds. Clinical dialysis. 2nd ed. Norwalk, CT: Appleton-Lange, 1990; 409-446
(15) Blum U, Skupin M, Wagner R, et al. Early and long term results of cardiac surgery in dialysis patients. Cardiovasc Surg 1994; 2:97-100
(16) Brenner BM, Lazarus JM. Chronic renal failure. In: Wilson JD, Braunwald E, Isselbacher KJ, et al, eds. Harrisons principles of internal medicine. New York, NY: McGraw-Hill, 1991; 1150-1157
(17) Shroyer AL, Grover FL, Edwards FH. Coronary artery bypass risk model: the society of thoracic surgeon's adult cardiac national data base. Ann Thorac Surg 1998; 1995:879-884
(18) Nakayama Y, Sakata R, Ura M, et al. Long term results of coronary artery bypass grafting in patients with renal insufficiency. Ann Thorac Surg 2003; 75:496-500
(19) Ascione R, Nason G, Al-Ruzzeh S, et al. Coronary artery revascularization with or without cardiopulmonary bypass in patients with preoperative non dialysis-dependent renal insufficiency. Ann Thorac Surg 2001; 72:2020-2025
(20) Williams BC, Demitrack LB, Fries BE. The accuracy of the National Death Index when personal identifiers other than the social security numbers are used. Am J Public Health 1992; 82:1145-1147
(21) Trachiotis GD, Hanumara D, McKenna L, et al. Surgical revascularization after acute myocardial infarction in patients with end-stage renal disease. Eur J Cardiothorac Surg 2004; 26:671-675
Manuscript received November 11, 2004; revision accepted February 17, 2005.
Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal.org/misc/reprints.shtml).
Correspondence to: Malek G. Massad, MD, Division of Cardiothoracic Surgery, The University of Illinois at Chicago, 840 S Wood St, CSB Suite 417 (MC 958), Chicago, IL 60612; e-mail: firstname.lastname@example.org
* From the Division of Cardiothoracic Surgery, Department of Surgery; The University of Illinois at Chicago, Chicago, IL. Presented at the 2004 Annual Meeting of the Western Surgical Association, November 7-10, 2004, Las Vegas, NV.
COPYRIGHT 2005 American College of Chest Physicians
COPYRIGHT 2005 Gale Group