Published online before print April 12, 2004, doi: 10.1161/01.CIR.0000125853.51637.C8
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2004;109:2092-2096.)
© 2004 American Heart Association, Inc.
Clinical Investigation and Reports |
Decision Guidelines for Prophylactic Replacement of Björk-Shiley Convexo-Concave Heart Valves
Impact on Clinical Practice
From the Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, (M.J.v.G., Y.v.d.G.), and Center for Clinical Decision Sciences, Department of Public Health, Erasmus Medical Center, Rotterdam (E.W.S.), the Netherlands.
Correspondence to Y. van der Graaf, MD, PhD, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, PO Box 85500, Str. 6.131, 3508 GA Utrecht, The Netherlands. E-mail Y.vanderGraaf{at}jc.azu.nl
Received September 18, 2003; revision received January 22, 2004; accepted February 2, 2004.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Background— Because of risk of outlet strut fracture, prophylactic replacement should be considered for Björk-Shiley convexo-concave (BScc) valve recipients. We assessed the effects of epidemiological and decision-analytic guidelines on actual BScc valve replacement.
Methods and Results— We performed a retrospective cohort study including all 2263 Dutch BScc patients with a mean follow-up of 11.3 years (range, 0 to 23 years). Outcomes were outlet strut fracture, mortality, and BScc valve replacement. For the surviving patients in 1992 (n=1330), we calculated the expected differences in life expectancy (LE) with and without BScc valve replacement according to decision guidelines developed in 1992. Differences in LE were compared with actual replacements. During 8 years of follow-up, there were 494 deaths (40%), and 11 patients had suffered outlet strut fracture. Of 1330 patients, 96 (10%) had undergone BScc valve replacement, particularly in years after introduction of initial and updated guidelines. One hundred seventeen patients (9%) had an estimated gain in LE after BScc valve replacement. These patients were more likely to undergo replacement than patients with an estimated loss of LE (hazard ratio, 6.6; 95% CI, 4.4 to 10; P<0.0001). A loss in LE after reoperation was predicted for 8 of 11 patients who experienced outlet strut fracture after guidelines were available.
Conclusions— Valve replacement for BScc heart valve patients was largely in concordance with guidelines in the Netherlands. Individualized guidelines that are based on high-quality epidemiological data and are updated and implemented rigorously can influence clinical practice in complex decision problems.
Key Words: follow-up studies • prevention • prosthesis • surgery • valves
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
The Björk Shiley convexo-concave (BScc) valve was introduced in 1979 as a successor of the Björk Shiley spherical valve. In 1986, the BScc valve was withdrawn from the market after repeated reports of outlet strut fracture. By this time, >86 000 patients had undergone valve replacement with a BScc valve. For these patients, prophylactic replacement had to be considered to avert possible catastrophic events. From a decision-analytic point of view, prophylactic replacement of the BScc valve is indicated when long-term fracture risks outweigh short-term reoperation risks.1 Epidemiological studies have been successful in determining individual risk factors for valve fracture, long-term survival, and operative mortality.2–5 Great efforts have been made to achieve adherence to the published decision guidelines in the Netherlands. In 1992, the Dutch BScc Follow-up Study, a retrospective cohort study, identified 4 risk factors for outlet strut fracture that were relatively easy to determine in the BScc patient: opening angle, valve size, mitral position, and young age.5 This publication led to worldwide notification by registered mail of all cardiologists and cardiothoracic surgeons. Subsequently, a formal decision model was introduced for the prophylactic replacement of BScc valves using the results of the Dutch BScc Follow-Up Study.6 In 1994, a more detailed model was developed with manufacturing characteristics that became available from Shiley Inc.7,8 Throughout the years, the decision guidelines were repeatedly refined according to new follow-up results.9,10 After closure of the Second Dutch BScc Follow-Up Study in 1998, a mailing was sent to all Dutch cardiothoracic centers with detailed individualized risk estimates for each of their patients. Figure 1 illustrates the time path with relevant information on decision guidelines and support.
|
Reducing uncertainty and managing risks with decision models and guidelines have become an integral part of current clinical practice. Although it has been shown that decision guidelines can improve quality of care, there is limited knowledge of their effectiveness in daily practice.11,12
In the present study, we assessed the impact of epidemiological analyses and formal decision guidelines for the prophylactic reoperation of BScc heart valve patients. Here, we compared differences in estimated life expectancy with actual decisions in the Dutch BScc cohort.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Patients
This retrospective cohort study included all 2263 Dutch BScc valve recipients, as described in detail previously.5,13 For the present analyses, all Dutch BScc patients were identified and followed up until December 2001. The 2263 patients received 60° (n=2254) and 70° (n=279) valves from 1979 to 1985. Of these, 1122 patients (50%) underwent aortic valve replacement, 800 (35%) had mitral valve replacement, and 341 (15%) underwent double valve replacement.
Follow-Up
We obtained follow-up data from the municipal registries and data on reoperations from all Dutch centers for cardiothoracic surgery. Information on the cause and mode of death was collected from the patient’s general practitioner or from clinical records. Mean duration of follow-up was 11.3 years (range, 0 to 23 years). Sixty-nine patients (3%) were lost to follow-up.
Outcome events during follow-up were outlet strut fracture, replacement of the BScc valve, and death as a result of all causes. Outlet strut fracture was defined as a separation of both legs of the minor strut of the valve, which was ascertained after emergency valve replacement operation or at necropsy.
Risk Calculations
For decision-making purposes, calculations were performed for patients who were alive and still had their valve in 1992. That year was used as the point of reference because it corresponds to the publication of the results of the First Dutch BScc Follow-Up Study in February 1992.5 For these patients, we estimated the loss of life expectancy as a result of fracture using a decision-analytic model based on the 1992 data.6 This loss can be compared directly with the loss of life expectancy caused by surgical mortality. If the loss of life expectancy caused by surgical mortality was estimated to be higher than the loss of life expectancy resulting from fracture, surgery was not beneficial in terms of life expectancy. If the estimated surgical mortality was lower than the loss in life expectancy, surgery was expected to result in a gain in life expectancy.
Data Analyses
We constructed Kaplan-Meier curves for the cumulative incidence of outlet strut fracture, a BScc replacement, and survival of all 2263 patients. We estimated valve replacement rates for each year on the basis of the total number of patients at risk for that year. A smoothed nonparametric curve was constructed with the Lowess algorithm.14
The differences in life expectancy with and without valve replacement estimations were categorized in 6 groups and subsequently compared with the number of BScc valve replacements during 8 years of follow-up. Cox regression analysis was used to determine whether patients with an estimated gain in life expectancy after valve replacement were more likely to undergo replacement of their BScc valve than patients with an estimated loss of life expectancy. In addition, characteristics, individual risks, and outcome of patients who suffered an outlet strut fracture after 1992 were assessed.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Among the initial cohort of 2263 patients, 154 died within 30 days, 1193 died during follow-up, and 230 underwent replacement of a BScc valve. In total, 52 patients suffered an outlet strut fracture, which was fatal in 34 (65%). Figure 2 shows cumulative survival, BScc replacements, and outlet strut fractures. The replacement rates per year are depicted in Figure 3. An increase in replacements was observed during several periods of follow-up. The first increase in replacements occurred in 1980 and 1981, and 2 other increases were noted in 1992 and 1998, which followed the notification of the first and second Dutch BScc Follow-Up Study.
|
|
Cohort From 1992
On February 1, 1992, a total of 1330 BScc patients were alive and still had their valve in situ. Baseline patient characteristics and valve characteristics are presented in Table 1. After 8 years of follow-up, 494 deaths had occurred (40%). Reoperation with replacement of a BScc valve was performed in 96 patients (10%). In 3 patients, this was an emergency replacement after outlet strut fracture. Eleven patients suffered an outlet strut fracture, which concerned a mitral valve in 9 and an aortic valve in 2 cases. Mortality was 78% (7 of 9) and 100% (2 of 2) for patients with mitral and aortic strut fractures, respectively.
|
Patients with an estimated gain in life expectancy after valve replacement (117, 9%) were more likely to actually undergo a BScc replacement than patients with an estimated loss of life expectancy (1213, 91%) (hazard ratio, 6.6; 95% CI, 4.4 to 10; P<0.0001). Table 2 shows the number of BScc replacements and the replacement rate for the different categories in expected gain in life expectancy after BScc valve replacement.
|
Most patients (8 of 11) who suffered an outlet strut fracture were those for whom a loss of life expectancy after BScc valve replacement was estimated (Table 3). Two patients had a considerable estimated gain in life expectancy after BScc valve replacement (11% and 28%). Inquiry into the rationale for a conservative strategy (nonreplacement) in these 2 patients revealed that both patients had received negative advice about replacement from their cardiothoracic surgeon because of supposedly outweighing surgical mortality risks. According to the decision model, these mortality risks were relatively low (5% and 1%).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Results of this study with >20 years of follow-up show that decision guidelines positively influenced replacement of fracture-prone valves in Dutch BScc valve recipients. Most outlet strut fractures that occurred after guidelines were available could not have been prevented with the available epidemiological knowledge.
The number of replacements in our study population showed several striking increases. The first peak of replacements was observed during the early years of implantation of the BScc valve. This peak may be explained by recurrence of endocarditis in a relatively large group of patients5 that often requires repeated surgery, especially in prosthetic valve recipients.15–17 The second and third replacement peaks coincide with the identification and publication of risk factors for outlet strut fracture in 1992 and the nationwide mailing of individual risk estimates to all Dutch cardiothoracic centers in 1998, respectively. The consequences of disclosing such important information are now shown empirically.
Several important factors can be identified for the relative success of the decision guidelines. First, media attention was substantial when information became available on the increased risk of mechanical failure of 60° BScc valves in 1992. In addition, because identified risk factors were credible and easy to comprehend, the 1992 results could lead to straightforward recommendations about prophylactic reoperation. Second, no alternatives were available for managing the complex decision-making scenarios that subsequently arose for BScc patients and their physicians. Further assistance in decision making was offered when age thresholds for prophylactic valve replacements were introduced in 1993. Third, when guidelines were extended with the latest follow-up data in 1998, individualized risk estimations were put at the disposal of all Dutch cardiothoracic centers. In this way, the decision guidelines were effective in identifying most high-risk valves and in making proper recommendations, including prevention of unnecessary reoperations. Finally, confidence in the decision guidelines was apparently high. For the different guidelines, essential components include (1) surgical mortality after prophylactic replacement, (2) age-specific annual risk of death, (3) annual risk of strut fracture, and (4) mortality and morbidity after strut fracture. The accuracy of the guidelines is determined by the validity and precision of these components. The complete model was not prospectively validated, but the robustness of the different risk estimates was substantiated in several studies.3,18,19 Estimation of surgical mortality after prophylactic replacement is the most uncertain component because data are sparse. However, we recently developed and validated a prediction rule that was able to give individualized estimates of the risk of mortality after reoperation.20
Despite all efforts, the decision guidelines probably could not serve as a basis for rational decisions for all Dutch BScc patients because they ignore specific individual circumstances and personal values of patients. Thus, a number of patients did not receive a reoperation although they had a relatively high estimated gain of life expectancy after valve replacement. For these patients, the prospect of a repeated heart surgery may have been even more disturbing than a possible valve fracture. Conversely, some BScc valve replacements were in patients with an obvious loss of life expectancy after valve replacement. Naturally, it is to be expected that other serious valve conditions could have urged valve replacement, but in these cases, the guidelines could have been left out of the decision-making process. Another explanation might be that for these patients the anxiety of living with a valve that might fail prevailed, even though risks were small. However, in a study that assessed the psychological distress among BScc valve patients compared with Sorin valve recipients, no significant increase in distress was measured despite knowledge of the risk of valve fracture.21
Our study illustrates that decision guidelines may have a practical impact when physicians are uncertain about appropriate practice and when epidemiological data can provide a solution.12 Methodological standards for evidence-based guideline development should be consulted to warrant a meticulous construction and validation process.22 However, not every medical dilemma that needs some sort of prognostication can comply with these standards. Lack of time and alternatives may force physicians and epidemiologists to come up with guidelines whose effectiveness in clinical practice remains to be proven. Furthermore, the development of decision guidelines is a continuous process that does not stop with guideline publication. To prevent outdated recommendations, extensive communication with physicians and regular updates of guidelines are necessary.23 The controllable development phase of guidelines is usually followed by an unpredictable implementation phase that includes the physician’s appreciation of the decision guidelines and the patient’s view of the uncertainty of his or her future. Clearly, the patient’s perspective and choice should dominate the outcome of the decision-making process, yet this outcome may not always correspond with recommendations. In addition, discrepancies between guidelines and the physician’s clinical judgment are major impediments for successful guideline implementation.24 It is therefore essential to develop a credible instrument that can reduce uncertainty for both patients and physicians.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
In the present study, we demonstrated that the introduction of guidelines for BScc valve replacement had a positive effect on actual decision making in clinical practice. Mainly patients with outweighing fracture risks underwent BScc valve replacement, and an unknown number of unnecessary reoperations were prevented. High-quality epidemiological data prove essential to manage uncertainty for physicians and patients involved in complex medical decisions. However, guidelines should be constantly updated and publicized to warrant a meticulous decision-making process.
|
Acknowledgments |
|---|
The independent supervisory panel, which took care of the Björk-Shiley research interests of Pfizer Inc, and the patients assembled in the class action, has supported our research.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
1. Birkmeyer JD, Marrin CA, O’Connor GT. Should patients with Bjork-Shiley valves undergo prophylactic replacement? Lancet. 1992; 340: 520–523.[CrossRef][Medline] [Order article via Infotrieve]
2. Ericsson A, Lindblom D, Semb G, et al. Strut fracture with Bjork-Shiley 70 degrees convexo-concave valve: an international multi-institutional follow-up study. Eur J Cardiothorac Surg. 1992; 6: 339–346.[Abstract]
3. Lindblom D, Bjork VO, Semb BK. Mechanical failure of the Bjork-Shiley valve: incidence, clinical presentation, and management. J Thorac Cardiovasc Surg. 1986; 92: 894–907.[Abstract]
4. Omar RZ, Morton LS, Halliday DA, et al. Outlet strut fracture of Bjork-Shiley convexo concave heart valves: the UK Cohort Study. Heart. 2001; 86: 57–62.
5. van der Graaf Y, de Waard F, van Herwerden LA, et al. Risk of strut fracture of Björk-Shiley valves. Lancet. 1992; 339: 257–261.[CrossRef][Medline] [Order article via Infotrieve]
6. van der Meulen JH, Steyerberg EW, van der Graaf Y, et al. Age thresholds for prophylactic replacement of Björk-Shiley convexo-concave heart valves: a clinical and economic evaluation. Circulation. 1993; 88: 156–164.
7. Kallewaard M, Algra A, van der Graaf Y. Welder identity, weld date, and the risk of outlet strut fracture in Björk-Shiley convexo-concave valves: the Dutch Cohort Study. Heart. 1996; 76: 510–512.
8. Steyerberg EW, van der Meulen JH, van Herwerden LA, et al. Prophylactic replacement of Björk-Shiley convexo-concave heart valves: an easy-to-use tool to aid decision-making in individual patients. Heart. 1996; 76: 264–268.
9. Kallewaard M, Algra A, Defauw J, et al. Which manufacturing characteristics are predictors of outlet strut fracture in large sixty-degree Björk-Shiley convexo-concave mitral valves? The Bjork-Shiley Study Group. J Thorac Cardiovasc Surg. 1999; 117: 766–775.
10. Steyerberg EW, Kallewaard M, van der Graaf Y, et al. Decision analyses for prophylactic replacement of the Björk-Shiley convexo-concave heart valve: an evaluation of assumptions and estimates. Med Decis Making. 2000; 20: 20–32.
11. Grimshaw JM, Russell IT. Effect of clinical guidelines on medical practice: a systematic review of rigorous evaluations. Lancet. 1993; 342: 1317–1322.[CrossRef][Medline] [Order article via Infotrieve]
12. Woolf SH, Grol R, Hutchinson A, et al. Clinical guidelines: potential benefits, limitations, and harms of clinical guidelines. BMJ. 1999; 318: 527–530.
13. Kallewaard M, Algra A, Defauw J, et al. Long-term survival after valve replacement with Björk-Shiley CC valves: Björk-Shiley Study Group. Am J Cardiol. 2000; 85: 598–603.[CrossRef][Medline] [Order article via Infotrieve]
14. Cleveland WS. The Elements of Graphing Data. Summit, NJ: Hobart Press; 1994.
15. Mansur AJ, Dal Bo CM, Fukushima JT, et al. Relapses, recurrences, valve replacements, and mortality during the long-term follow-up after infective endocarditis. Am Heart J. 2001; 141: 78–86.[CrossRef][Medline] [Order article via Infotrieve]
16. Ivert TS, Dismukes WE, Cobbs CG, et al. Prosthetic valve endocarditis. Circulation. 1984; 69: 223–232.
17. Tornos MP, Permanyer-Miralda G, Olona M, et al. Long-term complications of native valve infective endocarditis in non-addicts a 15-year follow-up study. Ann Intern Med. 1992; 117: 567–572.
18. Blot WJ, Omar RZ, Kallewaard M, et al. Risks of fracture of Björk-Shiley 60 degree convexo-concave prosthetic heart valves: long-term cohort follow up in the UK, Netherlands and USA. J Heart Valve Dis. 2001; 10: 202–209.[Medline] [Order article via Infotrieve]
19. Oxenham H, Bloomfield P, Wheatley DJ, et al. Twenty year comparison of a Björk-Shiley mechanical heart valve with porcine bioprostheses. Heart. 2003; 89: 715–721.
20. van Gorp MJ, Steyerberg EW, Kallewaard M, et al. Clinical prediction rule for 30-day mortality in Bjork-Shiley convexo-concave valve replacement. J Clin Epidemiol. 2003; 56: 1006–1012.[CrossRef][Medline] [Order article via Infotrieve]
21. Kallewaard M, Defauw J, van der Graaf Y. Psychological distress among recipients of Björk-Shiley convexo-concave valves: the impact of information. Heart. 1997; 78: 577–580.
22. Laupacis A, Sekar N, Stiell IG. Clinical prediction rules: a review and suggested modifications of methodological standards. JAMA. 1997; 277: 488–494.
23. Shekelle PG, Ortiz E, Rhodes S, et al. Validity of the Agency for Healthcare Research and Quality clinical practice guidelines: how quickly do guidelines become outdated? JAMA. 2001; 286: 1461–1467.
24. Garfield FB, Garfield JM. Clinical judgment and clinical practice guidelines. Int J Technol Assess Health Care. 2000; 16: 1050–1060.[CrossRef][Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  G. Gerosa, R. Carta, M. Montisci, L. Leoni, S. Iliceto, G. Rizzoli, and F. di Marco How to deal with recipients of valves prone to structural failure in the 2000s: Padua experience with the TRI Technologies valve. Ann. Thorac. Surg., September 1, 2006; 82(3): 858 - 864. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. J. Blot, M. A. Ibrahim, T. D. Ivey, D. E. Acheson, R. Brookmeyer, A. Weyman, J. Defauw, J. K. Smith, and D. Harrison Twenty-Five-Year Experience With the Bjork-Shiley Convexoconcave Heart Valve: A Continuing Clinical Concern Circulation, May 31, 2005; 111(21): 2850 - 2857. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||