HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Gary L. Grunkemeier Charles A.S. Marrin Permission Requests Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Birkmeyer, N. J.O. Articles by O’Connor, G. T. Search for Related Content PubMed PubMed Citation Articles by Birkmeyer, N. J.O. Articles by O’Connor, G. T.

Ann Thorac Surg 2000;70:1946-1952
© 2000 The Society of Thoracic Surgeons

---

Original article: cardiovascular

Prosthetic valve type for patients undergoing aortic valve replacement: a decision analysis¹

^a Department of Surgery, Dartmouth Medical School, Hanover, New Hampshire, USA
^b Department of Medicine, Dartmouth Medical School, Hanover, New Hampshire, USA
^c Center For the Evaluative Clinical Sciences, Dartmouth Medical School, Hanover, New Hampshire, USA
^d Veterans Affairs Medical Center, White River Junction, Vermont, USA
^e Medical Data Research Center, Providence Health System, Portland, Oregon, USA

Accepted for publication June 9, 2000.

Address reprint requests to Dr Birkmeyer, Surgical Outcomes Assessment Program, Department of Surgery, Dartmouth-Hitchcock Medical Center, 1 Medical Center Dr, Lebanon, NH 03756
e-mail: nancyj.birkmeyer{at}dartmouth.edu

	Abstract

Top Footnotes Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Background. In two large, randomized, clinical trials long-term survival after aortic valve replacement (AVR) was similar for patients receiving tissue and mechanical aortic heart valve prostheses. Higher bleeding rates among patients with mechanical valves, who must receive permanent oral anticoagulation to prevent thromboembolism, were offset by higher reoperation rates for valve degeneration among patients with tissue valves. Because the average age of patients undergoing AVR and clinical practices have changed considerably since the randomized clinical trials were conducted, we performed a decision analysis to reassess the optimal valve type for patients undergoing AVR.

Methods. We used a Markov state-transition model to simulate the occurrence of valve-related events and life expectancy for patients undergoing AVR. Probabilities of clinical events and mortality were derived from the randomized clinical trials and large follow-up studies.

Results. Although the two valve types were associated with similar life expectancy in 60-year-old patients (mean age of patients in the randomized clinical trials), tissue valves were associated with greater life expectancy than mechanical valves (10.7 versus 11.1 years) in 70-year-old patients (currently mean age of AVR patients). For 70-year-old patients, the effects of major bleeding complications (24%) with mechanical valves substantially outweighed those of reoperation for valve failure (12%) with tissue valves at 12 years. Of the clinical practice changes assessed, the recommended valve type was most sensitive to changes in bleeding rates with anticoagulation. However, bleeding rates would have to be 68% lower than those reported in the European randomized clinical trial to affect the recommended valve type for 70-year-old patients. Reoperation rates would have to be five times higher, and mortality rates at reoperation would have to be four times higher to affect the recommended valve type for 70-year-old patients.

Conclusions. Although mechanical valves are preferred for AVR patients less than 60 years old, most patients currently undergoing AVR are elderly and would benefit more from tissue valves.

	Introduction

Top Footnotes Abstract Introduction Material and methods Results Comment Acknowledgments References

 
The choice between mechanical and tissue prosthetic valves for patients undergoing aortic valve replacement (AVR) involves careful consideration of the risks and benefits of each type of prosthesis. Although mechanical prostheses have excellent durability, they require life-long anticoagulation therapy to minimize risks of thrombosis and embolism. Anticoagulation is associated with increased risks of bleeding. Although tissue valves do not require anticoagulation, they degenerate over time, necessitating reoperation for valve re-replacement. Two large, randomized clinical trials, each initiated in the 1970s, compared the outcomes of patients receiving each valve type [1, 2]. In these trials, higher bleeding rates observed among patients receiving mechanical valve prostheses were offset by higher rates of reoperation for prosthetic valve failure among patients receiving tissue valves, resulting in equivalent survival rates for patients receiving each valve type.

Since these studies were conducted, changes in the characteristics of patients undergoing AVR have taken place that may affect the relative risks and benefits of tissue and mechanical valves. With the declining incidence of rheumatic valve disease, AVR is now performed primarily for degenerative valve disease in increasingly older patients [3–5]. In elderly patients, mechanical valves may be less favorable because risks of anticoagulation-related bleeding increase with age [6–15]. In addition, tissue valve failure, which typically occurs 10 to 15 years after valve implantation, may be of lesser concern for older patients who have lower valve failure rates and shorter life expectancies [16].

Several important changes in clinical practices since the randomized clinical studies were conducted may also impact the relative risks and benefits of the valve types. Following recommendations for lower intensity oral anticoagulation, made possible by the standardization of measurement and dosing with the international normalized ratio, and the use of specialized clinics to monitor anticoagulation, would result in decreased bleeding risks associated with mechanical valves [17]. Changes have also taken place that may improve the performance of tissue valves. Newer generation tissue valves, including porcine valves treated with agents to retard calcification and homografts made available through heart transplantation, may have improved durability compared to older types of tissue valve prostheses [18]. In addition, risks of mortality with AVR have declined [19], as have risks of mortality with cardiac surgery in general [20–22].

In this study, we used a Markov state-transition model to assess the effects of increasing age and changes in clinical practices on the choice between tissue and mechanical valves for patients undergoing AVR.

	Material and methods

Top Footnotes Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Decision analysis
Decision analysis is a quantitative method for synthesizing data from numerous sources to evaluate treatment alternatives [23]. All decision analyses involve the following steps: the alternative therapeutic strategies and potential outcomes are specified in a decision model, the probability that each outcome will occur is estimated using the best available data, and the model is analyzed to calculate the expected value of each treatment alternative.

Markov model
In this study, we used a Markov state-transition model to simulate the occurrence of valve-related events and to estimate life expectancy (LE) for patients undergoing aortic valve replacement (Fig 1). This type of decision model is used for analyzing clinical problems involving risks that change or that can occur repeatedly over time [23, 24]. With this type of model, a large hypothetical cohort of patients is followed over time. With each cycle of the model, the cohort is redistributed among specified health states according to transition probabilities that are estimated from the medical literature. The model continues to run until all of the members of the cohort have died or reached the "dead" health state. Life expectancy is then calculated for each therapeutic strategy as the sum of the average cycles spent in living health states.

View larger version (15K): [in this window] [in a new window]   Fig 1. Schematic of decision model structure. All patients are initially in the event-free state. With each 1-year cycle of the model, they can move to a different state or remain in the same state according to chance events specified under "Chance Events Each Cycle." (Event-Free = alive without reoperation or major bleeding; Alive Following Reoperation = alive following reoperation for prosthetic valve failure; Alive Following Bleeding = alive following a major (requiring hospitalization or blood transfusion) bleeding event; Dead = dead from any cause.)

Assumptions
We incorporated several important simplifying assumptions into the model for this decision analysis. First, we only explicitly modeled events that have been proven to differ in incidence by valve type. Events that were associated with valve replacement but did not differ by valve type in the randomized clinical trials and other studies (perioperative mortality with primary AVR, prosthetic valve endocarditis, and thromboembolism [1, 2, 25–27]) were not explicitly modeled, but were captured in the valve-related excess mortality rates described below. Thus, our model estimates LE for survivors of primary aortic valve replacement who have no other indication (eg, atrial fibrillation) or contraindication (eg, women of childbearing age) to anticoagulation. Our model assumes that patients with mechanical valves who experience major bleeding will continue to receive oral anticoagulation therapy because studies have shown that risks of thromboembolism in patients with mechanical valves who are not anticoagulated exceed the risks of recurrent bleeding in those that are [28–30]. Finally, we made the simplifying assumption that patients undergoing reoperation for valve failure receive mechanical valves at reoperation. In this way, our analysis focuses on the implications of the initial valve decision, not subsequent decisions. Using a more complex, Monte Carlo simulation model, we tested the effect of this assumption. Because the results of this analysis were essentially identical to those of our baseline analysis, they are not reported here.

Model input variables
All model input variables and the sources of this information are described in Table 1. Where applicable, specific formulas for calculating model parameters are given in the Appendix.

Risks related to valve failure
Risks of tissue valve failure increase over time and are accelerated in younger patients. We modeled tissue valve failure based on a study that fit several parametric models to actuarial tissue valve failure curves and determined that the Weibull model provided the best fit to the data [16]. In this model the cumulative hazard of tissue valve failure is a function of patient age at valve implantation and the number of years since valve implantation. The instantaneous risk of tissue valve failure is equal to the first derivative of the cumulative hazard. Mechanical valves have a constant, low rate of failure [1, 2].

Risk of perioperative mortality with reoperation was estimated using a multivariate logistic regression model based on primary data from the Northern New England Cardiovascular Disease prospective heart valve registry [19, 34]. This model is based on data from all patients (n = 2,542) undergoing AVR between 1989 and 1996 in Maine, New Hampshire, or Vermont. Risks of mortality increase with patient age (relative risk, 3.4% per year) and for patients undergoing reoperation (relative risk, 2.2). Because published mortality rates with reoperation for valve failure vary, we tested the effects of both higher and lower mortality with reoperation in sensitivity analyses.

Risks related to bleeding
In our model, major bleeding was defined as bleeding requiring hospitalization or blood transfusion. Risks of major bleeding were modeled by fitting exponential functions to age-stratified results from a large study that analyzed risk factors for bleeding complications in oral anticoagulant therapy [15]. Patients with tissue valves have bleeding rates similar to the population at large with risks increasing with patient age [1, 2]. The probability that major bleeding resulted in death was modeled by fitting an exponential function to age-stratified data from a trial comparing rates of bleeding during antithrombotic therapy with either warfarin or aspirin in patients with atrial fibrillation (see Appendix for formulas) [6].

Sensitivity analyses
To test the stability of our results to variation in the estimates of particular model parameters (rates of anticoagulant-related bleeding, tissue valve failure, and perioperative mortality with reoperation for valve failure), we performed two–way sensitivity analyses. In these analyses, we assess the effects of patient age while varying the other parameter over a reasonable range to determine its affect on valve recommendations. Because most of these model parameters are functions of age or other factors, we applied variable relative risks to these functions in sensitivity analysis.

	Results

Top Footnotes Abstract Introduction Material and methods Results Comment Acknowledgments References

 
In Figure 2, we compare LE with each valve type for patients of different ages. Mechanical valves were favored for 50-year-old patients (21.4 versus 21.0 years LE) (Fig 1). Tissue valves were favored for 60 (15.8 versus 15.9 years LE), 70 (10.7 versus 11.1 years LE), and 80-year-old patients (6.4 versus 6.7 years LE).

View larger version (24K): [in this window] [in a new window]   Fig 2. Life expectancy by age and valve type. is the difference between tissue and mechanical valves in life expectancy.

View larger version (14K): [in this window] [in a new window]   Fig 3. Two-way sensitivity analysis of the effects of anticoagulant-related bleeding and patient age at implantation on the recommended valve type. (RR = relative risk where RR = 1 is the baseline estimate; RR = 1.5 is 50% higher than the baseline estimate, and RR = 0.5 is 50% lower than the baseline estimate.)

View larger version (14K): [in this window] [in a new window]   Fig 4. Two-way sensitivity analysis of the effects of reoperation for tissue valve failure and patient age at implantation on the recommended valve type. (RR = relative risk where RR = 1 is the baseline estimate, RR = 1.5 is 50% higher than the baseline estimate, and RR = 0.5 is 50% lower than the baseline estimate.)

View larger version (13K): [in this window] [in a new window]   Fig 5. Two-way sensitivity analysis of the effects of mortality with reoperation and patient age at implantation on the recommended valve type. (RR = relative risk where RR = 1 is the baseline estimate, RR = 1.5 is 50% higher than the baseline estimate, and RR = 0.5 is 50% lower than the baseline estimate.)

	Comment

Top Footnotes Abstract Introduction Material and methods Results Comment Acknowledgments References

There are two randomized [1, 2] and three nonrandomized studies [25–27] that compared survival and valve-related events among patients receiving tissue and mechanical valve prostheses. The basic recommendations derived from both the randomized and nonrandomized studies were the same with similar rates of mortality and valve-related events among AVR patients with mechanical and tissue valves. We believe that by avoiding bias and confounding inherent in nonrandomized studies, the randomized clinical trials represent the best evidence to date regarding the relative risks and benefits of tissue and mechanical valves. However, bleeding rates among both mechanical and tissue valve recipients were much higher in the United States than in the European randomized clinical trial. This is likely due to a relatively intense anticoagulation regimen and the frequent use of anticoagulation among patients with tissue valves in the US study. Because the anticoagulation regimen in the European study more closely resembles current recommendations, we used the bleeding rates from that study in our model. Estimates of age-specific bleeding and reoperation rates derived from more recent studies were used to refine baseline estimates to account for the effects of patient age on valve-related events.

One of the prior nonrandomized studies focused on effects of patient comorbidity of the choice of valve prosthesis for patients undergoing AVR [27]. This study found that patients with renal disease (preoperative creatinine, >2 mg/dL), chronic pulmonary disease in patients over the age of 60 years, coronary artery disease (>75% occlusion of a coronary artery), or impaired ventricular function (ejection fraction, <40%), have less than 10 years LE and should receive tissue valves. Freedom from reoperation at 10 years exceeded 95% for each of these groups.

In contrast to the recommendations from our decision analysis, the majority of patients undergoing AVR receive mechanical valves, including elderly patients. According to the 1988 National Health Interview Survey, approximately two-thirds of US patients received mechanical valves [35]. In the United Kingdom, mechanical valve usage increased from 54% to 70% from 1986 to 1989 [36]. The results of a 1985 survey by valve manufacturers indicate that worldwide, 69% of valves implanted are mechanical and 31% are tissue [37]. Mechanical valve usage is lowest in the southern hemisphere (57%) and highest in Europe (78%).

In considering the discordance between our results and current practices, it is important to review several limitations of our analysis. Estimates of valve-related events used in our model are based largely on the two randomized clinical trials of patients undergoing AVR in the 1970s. Although changes in patient age and clinical practices since that time have likely affected rates of valve-related events, through sensitivity analysis we have shown that these changes have only modest effects on age-specific recommendations for valve type.

Another limitation of our analysis is its use of LE as its primary outcome measure. Patient preferences and quality of life issues may also be important in clinical decisions about valve prosthesis. Anticoagulation with warfarin requires frequent office visits and phlebotomies for monitoring as well as activity and dietary restrictions. The potential for valve-related events (bleeding and reoperation) may cause substantial anxiety among patients and serious morbidity when they occur. We decided not to incorporate these factors into our model for a number of reasons. First, patient preferences regarding many of these clinical outcomes are known to vary widely. For example, a study of the effect of stroke and stroke prophylaxis on quality of life found wide variability in patients’ value for life after a moderate stroke (mean utility, 0.39; 10^th to 90th percentile, 0 to 0.99) and for life with warfarin therapy (mean utility, 0.987; 10^th to 90th percentile, 0.95 to 1.0). Second, although there are published sources for some of the utilities that we would need to incorporate patient preferences in our model including patient utilities for anticoagulation and different types of stroke, estimates for other important variables including reoperation for valve failure are lacking. Third, under circumstances such as this where treatment alternatives differ with regard to the timing of risks of morbidity and mortality (delayed until the time of valve failure for tissue valves) patient attitudes toward risk should also be accounted for [38]. For these reasons, we decided not to include patient preferences in our model and instead suggest that patient preferences be considered on an individual basis, particularly for patients for whom the valve types would be expected to offer equivalent life expectancy.

We believe that the common use of mechanical valves in elderly patients may reflect problems in clinical decision making. In particular, clinicians may underestimate bleeding risks of anticoagulation and overestimate LE and risk of tissue valve failure in elderly patients. In addition, clinicians may not incorporate patient preferences about chronic anticoagulation and potential clinical outcomes into the choice of heart valve prosthesis. On the basis of our findings and the current demographic profile of patients with aortic valve disease, we believe that more patients undergoing AVR should be receiving tissue valves.

	Acknowledgments

Top Footnotes Abstract Introduction Material and methods Results Comment Acknowledgments References

 
This study was supported by grants from the American Heart Association (93011160), the Agency for Health Care Policy and Research (T32 HS00070), and the American Federation for Aging Research (P 99155). Dr John Birkmeyer is a recipient of a Career Development Award from the Health Services Research and Development Program of the Department of Veterans Affairs.

	Footnotes

Top Footnotes Abstract Introduction Material and methods Results Comment Acknowledgments References

 
¹ The views expressed herein do not necessarily represent the views of the Department of Veterans Affairs or the United States government.

	Appendix

 
Valve failure risks
The probability of tissue valve failure is modeled as a Weibull function in which the cumulative hazard of tissue valve failure is equal to: , where , and years = the number of years since valve implantation. The instantaneous risk of tissue valve failure is equal to the first derivative of the cumulative hazard: [16].

Reoperation risks
The probability of death with reoperation is modeled as a multivariate logistic regression equation: , where , where age = patient age at reoperation and reoperation = 1 [19, 34].

Bleeding risks
The probability of bleeding is modeled as an exponential function 0.014 x 10^{(0.033 x age)} for patients with mechanical valves and 0.004 x 10^{(0.033 x age)} for patients with tissue valves, where age = patient age [15], The probability that major bleeding results in death is also modeled as an exponential function 2.29 x 10^{(0.015 x age)}, where age = patient age [6].

	References

Top Footnotes Abstract Introduction Material and methods Results Comment Acknowledgments References

 

1. Bloomfield P., Wheatley D., Prescott R., Miller H. Twelve year comparison of a Bjork-Shiley mechanical heart valve with porcine bioprostheses. N Engl J Med 1991;324:573-579.[Abstract]
2. Hammermeister K., Sethi G., Henderson W., et al. A comparison of outcomes in men 11 years after heart-valve replacement with a mechanical valve or bioprosthesis. N Engl J Med 1993;328:1289-1296.[Abstract/Free Full Text]
3. Selzer A. Changing aspects of the natural history of valvular aortic stenosis. N Engl J Med 1987;317:91-98.[Medline]
4. Passik C., Ackerman D., Pluth J., Edwards W. Temporal changes in the causes of aortic stenosis: a surgical pathologic study of 646 cases. Mayo Clin Proc 1987;62:119-123.[Medline]
5. Normand J., Loire R., Zambartas C. The anatomical aspects of adult aortic stenosis. Eur Heart J 1988;9:31-36.
6. Stroke Prevention in Atrial Fibrillation Investigators. Bleeding during antithrombotic therapy in patients with atrial fibrillation. Arch Intern Med 1996;156:409-416.[Abstract]
7. Byeth R., Quinn L., Landefeld C. Prospective evaluation of an index for predicting the risk of major bleeding in outpatients treated with warfarin. Am J Med 1998;105:91-99.[Medline]
8. Fihn S., Callahan C., Martin D., McDonell M., Henikoff J., White R. The risk for and severity of bleeding complications in elderly patients treated with warfarin. Ann Intern Med 1996;124:970-979.[Abstract/Free Full Text]
9. Landefeld C., Goldman L. Major bleeding in outpatients treated with warfarin: incidence and prediction by factors known at the start of outpatient therapy. Am J Med 1989;87:144-152.[Medline]
10. Landefeld C., Beyth R. Anticoagulant-related bleeding. Clinical epidemiology, prediction, and prevention. Am J Med 1993;95:315-328.[Medline]
11. Launbjerg J., Egeblad H., Heaf J., Nielsen N., Fuglehom A., Ladefoged K. Bleeding complications to oral anticoagulant therapy: multivariate analysis of 1010 treatment years in 551 outpatients. J Intern Med 1991;229:351-355.[Medline]
12. Palareti G., Leali N., Coccheri S., et al. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study. Lancet 1996;348:423-428.[Medline]
13. Palareti G., Manotti C., D’Angelo A., et al. Thrombotic events during oral anticoagulant treatment: results of the inception cohort, prospective, collaborative ISCOAT study. Thromb Haemost 1997;78:1438-1443.[Medline]
14. Steffensen F., Kristensen K., Ejlersen E., Dahlerup J., Sorensen H. Major haemorrhagic complications during oral anticoagulant therapy in a Danish population-based cohort. J Intern Med 1997;242:497-503.[Medline]
15. van der Meer F., Rosendaal F., Vandenbroucke J., Briet E. Assessment of a bleeding risk index in two cohorts of patients treated with oral anticoagulants. Thromb Haesmost 1996;76:12-16.[Medline]
16. Grunkemeier G., Jamieson W., Miller D., Starr A. Actuarial versus actual risk of porcine structural valve deterioration. J Thorac Cardiovasc Surg 1994;108:709-718.[Abstract/Free Full Text]
17. Tiede D., Nishimura R., Gastineau D., Mullany C., Orszulak T., Schaff H. Modern management of prosthetic valve anticoagulation. Mayo Clin Proc 1998;73:665-680.[Medline]
18. Turina J., Hess O., Turina M., Krayenbuehl H. Cardiac bioprostheses in the 1990’s. Circulation 1993;88:775-779.[Free Full Text]
19. Birkmeyer N., Marrin C., Morton J., et al. Decreasing mortality for aortic and mitral valve replacement in Northern New England. Ann Thorac Surg 2000;70:432-437.[Abstract/Free Full Text]
20. Hannan E., Kilburn H., Racz M., Shields E., Chassin M. Improving the outcomes of coronary artery bypass surgery in NY State. J Am Med Assoc 1994;271:761-766.[Abstract]
21. O’Connor G., Plume S., Olmstead E., et al. A regional study to improve the hospital mortality associated with coronary artery bypass graft surgery. J Am Med Assoc 1995;275:841-846.
22. Petersen E., Jollis J., Bebchuk J., et al. Changes in mortality after myocardial revascularization in the elderly: the national Medicare experience. Ann Intern Med 1994;121:919-927.[Abstract/Free Full Text]
23. Weinstein M., Fineberg H. Clinical decision analysis. Philadelphia: WB Saunders, 1980.
24. Beck J., Pauker S. The Markov process in medical prognosis. Med Decis Making 1983;3:419-458.
25. Holper K., Wottke M., Lewe T., et al. Bioprosthetic and mechanical valves in the elderly: benefits and risks. Ann Thorac Surg 1995;60:S443-S446.
26. Myken P., Caidahl K., Larsson P., Larsson S., Wallentin I., Berggtren H. Mechanical versus biological valve prosthesis: a ten-year comparison regarding function and quality of life. Ann Thorac Surg 1995;60:S447-S452.
27. Peterseim D., Cen Y., Cheruvu S., et al. Long-term outcome after biologic versus mechanical aortic valve replacement in 841 patients. J Thorac Cardiovasc Surg 1999;117:890-897.[Abstract/Free Full Text]
28. Cannegieter S., Rosendaal F., Wintzen A., Van der Meer F., Vandenbroucke J., Briet E. Optimal oral anticoagulant therapy in patients with mechanical heart valves. N Engl J Med 1995;333:11-17.[Abstract/Free Full Text]
29. Wong J., Webb R., Pauker S. Double trouble. A patient with two prosthetic valves and two episodes of intracranial bleeding. Med Decis Making 1987;7:174-193.
30. Wernly J., Crawford M. Choosing a prosthetic heart valve. Cardiol Clin 1998;16:491-504.[Medline]
31. The National Center For Health Statistics. Vital Statistics of the United States. In: Mortality, 1991. Washington DC: U.S. Public Health Service, 1991.
32. Abdelnoor M., Nitter H., Trettli S. Relative survival of patients after heart valve replacement. Eur Heart J 1990;11:23-28.[Abstract/Free Full Text]
33. Lindblom D., Lindblom U., Qvist J., Lundstrom H. Long-term relative survival rates after heart valve replacement. J Am Coll Cardiol 1990;15:566-573.[Abstract]
34. O’Connor N., Tosteson A., Marrin C., Olmstead E., O’Connor G. A prospective regional study of aortic valve replacement. Circulation 1994;90:1-282.[Free Full Text]
35. Garver D., Kaczmarek R., Silverman B., Gross T., Hamilton P. The epidemiology of prosthetic heart valves in the United States. Tex Heart Inst J 1995;22:86-91.[Medline]
36. Taylor K. Heart valve surgery in the United Kingdom: present practice and future trends. Br Heart J 1991;66:335-336.[Free Full Text]
37. Rabago G. A worldwide overview of valve usage. In: Rabago G., Cooley D., eds. Heart valve replacement & future trends in cardiac surgery. Mount Kisco, NY: Futura Publishing, 1987:3-9.
38. Cher D, Miyamoto J, Lenert L. Incorporating risk attitude into Markov-process decision models: importance for individual decision making. 1997;17:340–50.

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Gary L. Grunkemeier Charles A.S. Marrin Permission Requests Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Birkmeyer, N. J.O. Articles by O’Connor, G. T. Search for Related Content PubMed PubMed Citation Articles by Birkmeyer, N. J.O. Articles by O’Connor, G. T.