The CarboMedics prosthetic valve (Sulzer CarboMedics,

Transcription

1 Long-Term Result of 1144 CarboMedics Mechanical Valve Implantations Chang Hyun Kang, MD, Hyuk Ahn, MD, Kyung Hwan Kim, MD, and Ki-Bong Kim, MD Department of Thoracic and Cardiovascular Surgery, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, South Korea Background. It has been reported that the CarboMedics mechanical valve has acceptable valve-related complication rates. The aim of this study was to evaluate the long-term performance of the CarboMedics valve. Methods. Between August 1988 and September 1999, we implanted 1144 CarboMedics valves in 850 patients (aortic, 179; mitral, 385; double-valve, 234; tricuspid, 52). Mean patient age was years. Follow-up was completed in 95.3% and the median follow-up period was 7.9 years (6753 patient-years). Patients were divided into four groups according to implanted valve location and number (aortic, mitral, double, and tricuspid valve groups). Results. The overall hospital mortality rate was 3.4%, and the mortality rates in these groups were 1.7% in the aortic, 2.6% in the mitral, 4.7% in the double, and 9.6% in the tricuspid valve. The tricuspid valve group revealed a significantly higher mortality rate than the aortic and mitral valve groups (p < 0.05). Overall 10-year survival was 87.1% 2.6%, 88.9% 1.7%, 82.4% 2.9%, and 77.5% 7.0% in the aortic, mitral, double, and tricuspid valve groups, respectively. Age and tricuspid valve replacement were identified as significant risk factors of long-term survival by multivariate analysis (p < 0.05), and 99.4% 0.6%, 98.2% 0.8%, 99.2% 0.8%, and 87.6% 0.5% in the aortic, mitral, double, and tricuspid valve groups were free of valve thrombosis at 10 years. The tricuspid valve group demonstrated a significantly higher rate of valve thrombosis (p < 0.05). Conclusions. This long-term study of patients that received CarboMedics valve implantation demonstrates that the CarboMedics prosthetic valve has an acceptable incidence of valve-related complications. However, it should be noted that tricuspid valve replacement indicated a higher level of overall mortality and valve thrombosis. (Ann Thorac Surg 2005;79: ) 2005 by The Society of Thoracic Surgeons The CarboMedics prosthetic valve (Sulzer CarboMedics, Austin, TX) was first introduced in 1986 for clinical applications, and we have used this valve since 1988 as a primary mechanical prosthesis for diseased valve replacement. This prosthetic valve had several unique characteristics when it was first introduced. In particular, the sewing ring was reinforced by titanium, which prevented structural deterioration and allowed it to be easily detected by simple chest x-ray, in addition, the valve leaflet was rotatable, and the sewing ring was coated with Biolite carbon (Sulzer CarboMedics). Many clinical studies have reported the good hemodynamic performance and favorable clinical outcomes of the CarboMedics prosthetic valve [1 9], and comparisons with the St. Jude prosthetic valve produced similar clinical results [10, 11]. However, few studies have described the results of this valve in a comprehensive valve replacement patient group. The authors have used the CarboMedics mechanical valve in all positions, including the tricuspid position. Therefore, we undertook to evaluate the long-term survival of CarboMedics mechanical valve replacement Accepted for publication Dec 21, Address reprint requests to Dr Ahn, Department of Thoracic & Cardiovascular Surgery, Seoul National University Hospital, Seoul National University College of Medicine, 28 Yongon-Dong, Chongro-Gu, Seoul , Korea; ahnhyuk@snu.ac.kr. according to valve position, and to evaluate valve-related complications during long-term follow-up. Material and Methods Patients Patients who had undergone valve replacement with a CarboMedics prosthetic valve between August 1988 and February 1999 were included in this study. The study group comprised 850 patients who were enrolled. During the study period, 1405 patients underwent valve replacement at Seoul National University Hospital. A bioprosthesis was used in 85 (6.1%) patients, mechanical valves other than the CarboMedics valve in 470 (33.5%), and the CarboMedics valve was used in 60.5% of all valve replacement patients. The mean patient age was years, and there were 399 men and 451 women in the patients group. A total of 1144 CarboMedics valves were implanted during the study period. Four hundred thirty-four Carbo- Medics prosthetic valves (37.9%) were used in the aortic position, 658 (57.5%) in the mitral position, and 52 (4.5%) in the tricuspid position. Single valve replacement was performed in 577 patients (67.9%), double valve replacement in 252 (29.6%), and triple valve replacement in 21 (2.5%). The patients were divided into four subgroups according to replaced valve positions and numbers: isolated aortic valve replacement group (AVR group; n 179); isolated mitral valve replacement group (MVR group; n 385); double 2005 by The Society of Thoracic Surgeons /05/$30.00 Published by Elsevier Inc doi: /j.athoracsur

2 1940 KANG ET AL Ann Thorac Surg CARBOMEDICS PROSTHETIC VALVE 2005;79: Table 1. Preoperative Clinical Data Variable AVR (n 179) MVR (n 385) DVR (n 234) TVR (n 52) p Value Age (year) NYHA class I 22.6% a 8.0% 5.2% 7.7% II 39.1% 33.1% 41.9% 28.2% 0.05 III 33.1% 49.2% 41.3% 41.0% 0.05 IV 5.3% 9.6% 11.6% 23.1% a Atrial fibrillation 12.8% a 76.1% 69.7% 75.0% Previous surgery 16.8% b 36.4% 33.3% 53.8% b LVESD (mm) b b LVEDD (mm) c c c c EF (%) LA size (mm) a a Indicates significantly different from the other three groups. b Indicates significantly different from the MVR and DVR groups and from each other. c Indicates significantly different between all four groups. AVR aortic valve replacement; DVR double valve replacement; EF ejection fraction; LA left atrium; LVEDD left ventricular end-diastolic dimension; LVESD left ventricular end-systolic dimension; MVR mitral valve replacement; NYHA New York Heart Association; TVR tricuspid valve replacement. valve replacement group (DVR group; n 234); and tricuspid valve replacement group (TVR group; n 52). Double valve replacement group represents concomitant AVR and MVR. The TVR group includes isolated TVR (n 13), concomitant MVR and TVR (n 18), and triple vale replacement (n 21). Procedures Operations were performed under moderate hypothermia. Myocardial protection was maintained by using cardioplegic solution and by local cooling with iced slush. Blood antegrade intermittent cardioplegic solution delivery was the most frequently used method. To monitor myocardial temperature, a small metal-tipped temperature probe was inserted into the apical septum of the left ventricle, and myocardial temperature was maintained at around 15 C during aortic cross clamping. Valve sutures were placed using the mattressed suture technique with 2-0 Ethibond (Ethicon, Summerville, NJ) reinforced by Teflon pledget or spaghetti. Nick s procedure was performed for selected patients with a small aortic annulus. The left atrial auricle was excised or obliterated in the cases of atrial fibrillation or intraatrial thrombus. Antiarrhythmic procedures have been performed concomitantly with valve replacement since The Cox-Maze and pulmonary vein isolation procedures were performed most frequently. Anticoagulation Heparin or low-molecular-weight heparin was used from the immediate postoperative period if there was no evidence of active bleeding, and warfarin was used after beginning oral feeding. The prothrombin time was accommodated according to the site and the number of replaced valves. The international normalized ratio (INR) was maintained between 1.5 and 2.5 in the isolated AVR group, between 2.0 and 3.0 in the MVR or DVR group, and between 2.5 and 3.5 in the TVR group. Dipyridamole or aspirin was added for those with risk factors of thromboembolism. Follow-up was completed by attending surgeons or cardiologists at intervals of 3 months. More frequent follow-up was required for the patients Table 2. Etiology of Valvular Disease Cause AVR MVR DVR TVR p Value Rheumatic 58 (32.4%) a 209 (54.3%) 130 (55.6%) 16 (30.8%) a Redo operation 26 (14.5%) b 134 (34.8%) 77 (32.9%) 28 (53.8%) b Degenerative 35 (19.6%) c 22 (5.7%) 11 (4.7%) 2 (3.8%) Congenital 38 (21.2%) a 6 (1.6%) 2 (0.9%) 5 (9.6%) a Endocarditis 22 (12.3%) c 9 (2.3%) 10 (4.3%) 1 (1.9%) Others 0 (0%) 5 (1.3%) 4 (1.7%) 0 (0%) 0.05 a Indicates significantly different from the MVR and DVR groups, but no significance between the AVR and TVR groups. b Indicates significantly different from the MVR and DVR groups, and also significantly different from each other. c Indicates significantly different from the other three groups. AVR aortic valve replacement; DVR double valve replacement; MVR mitral valve replacement; TVR tricuspid valve replacement.

3 Ann Thorac Surg KANG ET AL 2005;79: CARBOMEDICS PROSTHETIC VALVE 1941 Table 3. Additional Procedures Procedures AVR MVR DVR TVR Vavular procedure AVP MVP/MAP TVP/TAP Anti-arrhythmic procedure Cox-Maze PVI Other procedures CABG VSD/ASD closure Nick s procedure Others ASD atrial septal defect; AVP aortic valvuloplasty; AVR aortic valve replacement; CABG coronary artery bypass graft; DVR double valve replacement; MAP mitral annuloplasty; MVP mitral valvuloplasty; MVR mitral valve replacement; PVI pulmonary vein isolation; TAP tricuspid annuloplasty; TVP tricuspid valvuloplasty; TVR tricuspid valve replacement; VSD ventricular septal defect. who revealed fluctuant INR level by in-hospital pharmacists according to the anticoagulation service program at our hospital by 2 weeks or 1 month for a brief period. Fig 1. Overall survival. (AVR aortic valve replacement [ ]; DVR double valve replacement [ ];MVR mitral valve replacement [ ];TVR tricuspid valve replacement [ ].) Follow-Up Data Data were gathered by reviewing medical records, by telephone contact, and by consulting the national database of death statistics (Korean National Statistical Office). Telephone contact was done directly with the patient in principle. However, information was also gathered from the family members if the patient had already expired. The information about the survival of all the patients was completely gathered by the national database of death statistics. However, we regarded the data with unknown cause of death and obscure late clinical course as an incomplete follow-up. The data collection period was from July 2001 to Dec and during this period complete data was collected on 810 patients. The completeness of follow-up was 95.3%, the mean period of follow-up was years (0.1 year to 13.2 years), and the cumulative follow-up was patient-years (PY). Statistical Analysis The terminologies and definitions in this report were described by the methods suggested by Edmund and colleagues [12]. Continuous variables were described as means standard deviation, and discrete variables as numbers or ratios. Comparisons of continuous variables were made by one-way ANOVA. Turkey s multiple comparison test was used for the intergroup analysis. Pearson s Chi-square test or Fisher s exact test were used to compare discrete variables. The Kaplan-Meier statistics were used to calculate survival rates and freedom from valve-related complications, and the log-rank test was used to determine intergroup statistical difference. Risk factor analysis for the Table 4. Causes of Early and Late Mortality Causes AVR MVR DVR TVR Total Early mortality Low cardiac output Left ventricle rupture Bleeding Arrhythmia Infection Others Late mortality Congestive heart failure Valve-related death Noncardiac death Sudden/unknown AVR aortic valve replacement; MVR mitral valve replacement; DVR double valve replacement; TVR tricuspid valve replacement. Fig 2. Freedom from thromboembolism. (AVR aortic valve replacement [ ]; DVR double valve replacement [ ]; MVR mitral valve replacement [ ];TVR tricuspid valve replacement [ ].)

4 1942 KANG ET AL Ann Thorac Surg CARBOMEDICS PROSTHETIC VALVE 2005;79: Table 5. Valve-Related Morbidity AVR MVR DVR TVR Thromboembolism Valve thrombosis Hemorrhage PVE Nonstructural dysfunction Structural dysfunction Reoperation AVR aortic valve replacement; DVR double valve replacement; MVR mitral valve replacement; PVE prosthetic valve endocarditis; PY patient-year; TVR tricuspid valve replacement. long-term survival was performed using Cox s proportional hazard model. The variables included in the Cox s proportional hazard model were age, sex, New York Heart Association (NYHA) functional class, etiology of valvular heart disease, the need for a redo operation, preoperative atrial fibrillation, position of the valve, and the administrations of tricuspid valvuloplasty and antiarrhythmic surgery. Univariate and multivariate analysis was performed including all the previously listed risk factors. Statistical analysis was performed using SPSS statistical software (version 11.0; SPSS Inc, Chicago, IL), and a statistically significant difference was defined as a p value of less than Results Preoperative Clinical Characteristics and Etiology Preoperative clinical data and causes of valve replacement are summarized in Tables 1 and 2. The AVR group demonstrated relatively favorable preoperative functional status and various etiologies. However, the TVR group revealed more advanced preoperative status and high proportion of reoperation. Redo surgery due to previously implanted prosthetic valve failure was performed in 275 patients Fig 3. Freedom from valve thrombosis. (AVR aortic valve replacement [ ]; DVR double valve replacement [ ]; MVR mitral valve replacement [ ];TVR tricuspid valve replacement [ ].) (32.4%). Mean duration between first surgery and redo surgery was years ( years). We performed 52 TVR procedures in this series. The most common cause of TVR was reoperation. Among 28 reoperations, 8 patients were redo TVR due to bioprosthesis failure and 3 patients showed tricuspid insufficiency in spite of the previous tricuspid annuloplasty. The next common cause was rheumatic tricuspid stenoinsufficiency (n 16), which indicated leaflet thickening, and commissural and subvalvular apparatus fusion. Five patients were congenital deformity, like Ebstein s anomaly, in which the repair itself was very demanding. Valve Size and Concomitant Procedures The most frequently used sizes of prosthetic valve were as follows: 21, 23, and 25 mm in the AVR group; 29 and 31 mm in the MVR group; and 33 mm in the TVR group. Concomitant operations are listed in Table 3; 83 patients (9.8%) received antiarrhythmic surgery, which included a pulmonary isolation procedure in 63 (7.4%), the Cox-Maze procedure in 18 (2.1%), right atrium separation in 1 (0.1%), and cryoablation of the accessory conduction pathway in 1 patient (0.1%). Mean cardiopulmonary bypass time and aortic cross clamp time were minutes ( minutes) and minutes ( minutes), respectively. Early and Late Mortality Early mortality within 30 postoperative days occurred in 29 patients (3.4%; Table 4). The early mortality rates by patient group were 1.7% in the AVR group, 2.6% in the MVR group, 4.7% in the DVR group, and 9.6% in the TVR group. The early mortality rate of TVR group was significantly higher than those of the AVR and MVR groups (p 0.05). No statistically significant early mortality difference was evident between the AVR, MVR, and DVR groups. Late mortalities occurred in 79 patients (9.3%). The overall 5- and 10-year survival of all patients were 90.9% 1.0% and 86.4% 1.4%, respectively. The 10-year survival by patient group were 87.1% 2.6% in the AVR group, 88.9% 1.7% in the MVR group, 82.4% 2.9% in the DVR group, and 77.5% 7.0% in the TVR group (Fig 1). Multivariate analysis identified the following as risk factors that affected long-term survival; age (odd ratio 1.06, 95% confidence interval [CI] , p 0.001) and tricuspid valve replacement (odd ratio 2.35, 95% CI , p 0.034). Valve-Related Complications Fifty-one thromboembolic events occurred in 45 patients and 8 patients died due to thromboembolic events (Fig 2; Table 5). Thromboembolism locations were as follows: brain in 44 (86.3%), gastrointestinal organ in 3 (5.9%), kidney in 2 (3.9%), and the extremities in 2 (3.9%). All mortality cases were related to thromboembolism in the brain. Seventeen valve thrombosis events occurred in 13 patients (Fig 3). There was no mortality associated with valve thrombosis. Valve thrombosis locations were as

5 Ann Thorac Surg KANG ET AL 2005;79: CARBOMEDICS PROSTHETIC VALVE 1943 Fig 4. Freedom from anticoagulation-related hemorrhage. (AVR aortic valve replacement [ ]; DVR double valve replacement [ ]; MVR mitral valve replacement [ ];TVR tricuspid valve replacement [ ].) follows: aortic valve in 1, mitral in 7, combined aortic and mitral in 1, and tricuspid in 8. Reoperations were required for 8 events and the other 9 events were managed by thrombolytic therapy. The 10-year freedom from valve thrombosis was achieved by 99.4% 0.6% in the AVR group, 98.2% 0.8% in the MVR group, 99.2% 0.8% in the DVR group, and by 87.6% 0.5% in the TVR group. The TVR group had a statistically significant higher rate of valve thrombosis than the other three patient groups (p 0.001). Anticoagulation-related hemorrhage occurred 50 times in 40 patients (Fig 4). The locations of hemorrhage were subdural hemorrhage in 18 (36.0%), intracerebral hemorrhage in 9 (18.0%), gastrointestinal hemorrhage in 9 (18.0%), and hemorrhage at other sites in 14 (28.0%). Eight patients died due to intracranial hemorrhage. Fig 5. Freedom from PVE. (AVR aortic valve replacement [ ]; DVR double valve replacement [ ];MVR mitral valve replacement [ ]; PVE prosthetic valve endocarditis; TVR tricuspid valve replacement [ ].) Prosthetic valve endocarditis occurred 22 times in 20 patients (Fig 5). Reoperations were performed in 3 patients and other patients were managed by medical treatment. There were 6 mortalities (30.0%) including two operative deaths. Initial treatment of prosthetic valve endocarditis was intensive antibiotics therapy in our hospital during study period. Surgical intervention was considered when the progression of hemodynamic instability or resistant to medical treatment had occurred. Nonstructural dysfunction requiring treatment occurred in 5 patients and all were managed by reoperation. The cause of nonstructural dysfunction was paravalvular leakage in 4 patients and pannus overgrowth in 1 patient. Behçet s disease had been confirmed postoperatively in 1 patient and he underwent reoperations twice due to recurrent paravalvular leakage. However, he died due to left ventricular dysfunction after the second reoperation. No structural dysfunction occurred in any CarboMedics mechanical valve during follow-up. Reoperations were performed in 16 patients. The causes of reoperation were valve thrombosis in 8 (50.0%), paravalvular leakage in 4 (25.0%), prosthetic valve endocarditis in 3 (18.8%), and pannus overgrowth in 1 patient (6.3%). Five patients died after reoperation, indicating a reoperation operative mortality of 31.3%. The 10-year freedom from reoperation was achieved by 99.4% 0.6% in the AVR group, 96.6% 1.2% in the MVR group, 97.3% 1.5% in the DVR group, and 95.5% 3.2% in the TVR group; differences between patient groups were not significant (p 0.05). Comment The early mortality rate of CarboMedics mechanical valve replacement has been reported to be between 1.2% and 9% by previous studies [1 9]. Differences in mortality rates between studies are mainly attributable to patient preoperative status and the natures of concomitantly performed operations. Soga and colleagues. [2] reported a 1.2% surgical mortality rate and suggested that this low mortality rate may have been due to a relatively low proportion of NYHA functional class III and IV, a low incidence of concomitant coronary artery bypass surgery, and a low incidence of emergency surgery. Other studies have reported higher early mortality rates due to older age, and higher incidences of ischemic heart disease, left ventricular dysfunction, and reoperation [3, 13]. In our study, the patients were relatively young and had a low incidence of ischemic heart disease and the early mortality rate was 3.4%, which is comparable to the other studies. The most common causes of death after mechanical valve replacement are congestive heart failure and valverelated complications. Although heart failure and myocardial infarct had been reported to be the most common causes of death in a previous report [4], in the present study, valve-related complications was the most common cause of late death and congestive heart failure was the next. The long-term survival of multiple valve replacement has been reported to be lower than that of single valve replacement [1, 4]. However, the authors could not

6 1944 KANG ET AL Ann Thorac Surg CARBOMEDICS PROSTHETIC VALVE 2005;79: find any difference in long-term survival between the single valve replacement group (composed of the AVR and MVR groups) and the double valve replacement group. On the other hand, TVR was identified as a significant risk factor by multivariate analysis. Tricuspid valve replacement has been reported to be associated with high early mortality and poor long-term result [14]. In the present study, we identified two categories of TVR patients, those who demonstrated tricuspid disease combined with mitral valve disease and those with isolated tricuspid disease, including Ebstein s anomaly. Moreover, both of these categories are expected to have poor short-term and long-term results because of preexisting right ventricular dysfunction. In addition, tricuspid valve replacement reveals a tendency toward high valve thrombosis and requires high levels of anticoagulation, which might result in a high risk of anticoagulation-related hemorrhage. In the present study, we found higher rates of valve thrombosis and of hemorrhagic events in the TVR group. However, differences in the incidences of anticoagulation-related hemorrhage were not significant versus the other valve replacement groups. It was our preference to use the mechanical prosthetic valve in the tricuspid position instead of the bioprosthetic valve because our patients were relatively younger than the western groups. We experienced 8 cases of redo TVR due to the bioprosthesis failure in this series and identified acceptable durability of the CarboMedics mechanical valve without structural or nonstructural dysfunction in tricuspid position. Valve replacement by mechanical valvular prosthesis has the advantage of long-term durability. However, it could also eventually result in valve-related complications including thoromboembolism and hemorrhage. The overall rate of valve-related complication has been reported to be about 50% [4], the rate of thromboembolism to be 0.5% 1.7% per PY [3 8], and aortic prosthetic valve locations were found to exhibit the least thromboembolism [4, 7]. However, some have reported higher rates of thromboembolism after AVR [2], and suggested relatively low intensity anticoagulation in the AVR group as a reason. In our study, although not statistically significant, the AVR group demonstrated a relatively high rate of thromboembolism under the low intensity anticoagulation. Valve thrombosis usually develops within the first postoperative year, because during this period endothelialization of the sewing ring is incomplete. One report claimed a higher rate of valve thrombosis for the CarboMedics prosthetic valve versus the St. Jude prosthetic valve [15]. However, another study reported comparable rates for the two valves [11]. The valve thrombosis rate of the CarboMedics valve has been reported to be 0.1% 0.4% per PY. In the present study, the valve thrombosis rate in the TVR group was found to be significantly higher than in the other groups. The valve thrombosis rate in the tricuspid valve position has been reported to be around 1 %/PY, which is significantly higher than those of other sites [16]. The cause of a higher valve thrombosis rate in the tricuspid position is multifactorial. It was suggested that a low pressure in the tricuspid position promotes fibrous tissue overgrowth and subsequent surface irregularities [17], and cardiac output, heart rate, arrhythmia, and ventricular function are associated with valve thrombosis [18]. To summarize, this study demonstrates that the Carbo- Medics mechanical valve is a reliable and safe prosthetic valve with a low level of valve-related complications and excellent long-term survival. However, relatively high overall mortality and valve thrombosis were observed in the tricuspid position, which we believe were probably due to tricuspid valve disease rather than the prosthetic valve. References 1. Bernal JM, Rabasa JM, Gutierrez-Garcia FG, Morales C, Nistal JF, Revuelta JM. The Carbomedics valve: experience with 1049 implants. Ann Thorac Surg 1998;65: Soga Y, Okabayashi H, Nishina T, et al. Up to 8-year follow-up of valve replacement with Carbomedics valve. Ann Thorac Surg 2002;73: Rödler SM, Moritz A, Schreiner W, End A, Dubsky P, Wolner E. Five-year follow-up after heart valve replacement with the CarboMedics bileaflet prosthesis. Ann Thorac Surg 1997;63: Dalrymple-Hay MJ, Pearce R, Dawkins S, et al. A singlecenter experience with 1,378 CarboMedics mechanical valve implants. Ann Thorac Surg 2000;69: Fiane AE, Gerian OR, Svennevig JL. Up to eight years follow-up of 997 patients receiving the CarboMedics prosthetic heart valve. Ann Thorac Surg 1998;66: Copeland III JG, Sethi GK. Four-year experience with the Carbomedics valve: the North American Experience. Ann Thorac Surg 1994;58: De Luca L, Vitale N, Giannolo B, Caferella G, Piazza L, Cotrufo M. Midterm follow-up after heart valve replacement with CarboMedics bileaflet prosthesis. J Thorac Cardiovasc Surg 1993;106: Craver J. CarboMedics prosthetic heart valve. Eur J Cardiothorac Surg 1999;15:S Nistal JF, Hurle A, Revuelta JM, Gandarillas M. Clinical experience with the CarboMedics valve: early results with a new bileaflet mechanical prosthesis. J Thorac Cardiovasc Surg 1996;112: Butterfield M, Fisher J, Davies GA, Spyt TJ. Comparative study of hemodynamic function of the CarboMedics valve. Ann Thorac Surg 1991;52: Lim KH, Caputo M, Ascione R, et al. Prospective randomized comparison of CarboMedics and St. Jude Medical bileaflet mechanical heart valve prosthesis: an interim report. J Thorac Cardiovasc Surg 2002;123: Edmund LH, Clark RE, Cohn LH, Grunkemeier GL. Miller DC, Weisel RD. Guidelines for reporting morbidity and mortality after cardiac valvular operations. Ann Thorac Surg 1996;62: Czer LS, Chaux A, Matloff JM et al. Ten-year experience with the St. Jude Medical valve for primary valve replacement. J Thorac Cardiovasc Surg 1990;100: Ratnatunga CP, Edwards MB, Dore CJ, Taylor KM. Tricuspid valve replacement: UK Heart Valve Registry mid-term results comparing mechanical and biological prostheses. Ann Thorac Surg 1998;66: Rosengart TK, O Hara M, Lang SJ, et al. Outcome analysis of 245 CarboMedics and St. Jude valves implanted at the same institution. Ann Thorac Surg 1998;66: Van Nooten GJ, Caes F, Taeymans Y, et al. Tricuspid valve replacement: postoperative and long-term results. J Thorac Cardiovasc Surg 1995;110: Horstkotte D, Burckhardt D. Prosthetic valve thrombosis. J Heart Valve Dis 1995;4: Ryder SJ, Bradley H, Brannan JJ, Turner MA, Bain WH. Thrombotic obstruction of the Bjork-Shiley valve: the Glasgow experience. Thorax 1984;39: