Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Should warfarin be routinely prescribed for the first three months after a bioprosthetic valve replacement?

Three Part Question

In patients undergoing a [bioprosthetic valve replacement] is [anticoagulation superior to antiplatelet therapy] for the prevention of [thromboembolism]?

Clinical Scenario

You are consenting a 64-year old lady for AVR. She is quite keen to go for a bioprosthesis as her mother was on warfarin in the past and it had 'never agreed with her'. She is then quite disappointed when you tell her that she will actually have to be on warfarin for 3-months after the operation, and asks what would happen if she didn't take it. You can't quote her a figure of increased risk and therefore resolve to look up the answer.

Search Strategy

Medline 1966–May 2006
[exp Warfarin/ OR exp Anticoagulants/ OR OR antithrombo$.mp OR OR vitaminK antagonist$.mp] AND [ OR OR OR] AND [exp Mitral valve/ OR exp Aortic Valve/ OR exp Heart Valve Prosthesis/ OR exp Heart Valve Prosthesis Implantation/ OR valve OR OR]-limit-to-humans

Search Outcome

A total of 620-abstracts were found. All major guidelines were also included and their reference lists searched. 15 papers were deemed to represent the best evidence on the topic

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Butchart et al,
Guideline on behalf of the European Society of CardiologyReview based on cohort studies (Level 2a)Recommendation for Warfarin after a bioprosthesis.For the first 3 months, in all patients with bioprostheses or mitral valve repair involving the use of a prosthesic annuloplasty ring. Although there is widespread use of aspirin as an alternative to anticoagulation for the first 3 months in patients with no other indications for anticoagulation, there are no randomized studies to support the safety of this strategy.Only 2 references given in support of warfarin for bioprostheses [CTS-net;Gherli]
Recommendation for the initiation of anticoagulationUntil data from randomized trials are available, intravenous unfractionated heparin with aPTT 1.5-2.0 until a therapeutic INR is achieved with oral anticoagulation.
Target INR for bioprosthesisTarget INR is 2.5 unless AF, LA>50mm, EF<35%, MVR, TVR, PVR, a mitral gradient or spontaneous echo contrast in which case INR target is 3.0.
ACCP consensus conference,
2001 and 2004,
6th and 7th Consensus conference From the American College of Chest PhysiciansReview based on cohort studies ( level 2a)Recommendation after bioprosthesisWe recommend that patients with bioprosthetic valves in the mitral position be treated for the first 3 months after valve insertion with oral anticoagulants (grade 1C+ recommendation).
Recommended INRWe recommend a target INR of 2.5 (range, 2.0 to 3.0) during the first 3 months after operation in patients with bioprosthetic valves in the mitral or aortic position (grade 1A recommendation based on an investigation that used an INR of 2.0 to 2.3).
Bonow et al for the ACC/AHA,
Guideline for the American College of Cardiology / American Heart Association, from a group of 90 american and Canadian specialists.Review based on cohort studies ( level 2b)Recommendations for bioprosthesisWarfarin is usually recommended although in several centres only aspirin is used. Risk is particularly high in the first few days after surgery and heparin should be started as soon at the risk of bleeding is reduced.5 references given
Specific recommendationsAVR or MVR plus no risk factors – Aspirin 80-100mg/day

AVR plus AF, LV dysfunction, previous Thromboembolism, hypercoagulable condition. Warfarin INR 2-3 (Grade 1)

MVR plus risk factor warfarin INR 2.5-3.5
Lowe et al (SIGN),
Scotland, UK
Scottish Intercollegiate Guidelines Network National Clinical GuidelineGuideline based on cohort studies (Level 2a)Recommendation for bioprosthesisIn patients with isolated aortic or tricuspid bioprosthetic valves with no other risk factors, warfarin is recommended for three months only, in the absence of atrial fibrillation or a history of systemic embolism.
Mitral Bioprosthetic valves.patients with mitral bioprosthetic valves should receive: warfarin for 3-6 months (target INR 2.5, range 2.0-3.0)
High risk patients with bioprosthetic valvesPatients with bioprosthetic heart valves who have additional thrombotic risk factors should receive long term prophylaxis with warfarin (INR 2.5, range 2.0-3.0).
British Society of Haematologists,
Guideline from the British Society of HaematologistsGuideline based on cohort studies (Level 2b)Recommendations for aortic bioprosthesesOral anticoagulants are not required for valves in the aortic position in patients in sinus rhythm, although many centres anticoagulate patients for 3-6 months after any tissue valve implant.
Recommendation for mitral bioprosthesisPatients with mitral bioprosthesis should receive oral anticoagulants ( INR 2.5) for 3 months
Survey of Anticoagulation management for bioprosthetic aortic valve. 726 CTSnet member surgeons replied 40% from USA 30% from EuropeSurvey (level 3b)Awareness of ACC guidelines79% were aware of guidelines
Expectation that a patient without comorbidities would require warfarin60% No warfarin 40% Yes, will need warfarin
Does warfarin increase hospital stay63% Yes ( 45% of surgeons believed that it would prolong discharge by 2 days)
Are antiplatelets an acceptable alternative to warfarin60% Yes 15% No 24% Sometimes
Warfarin is no longer the standard of care for tissue aortic valves63% yes 27% no
Vaughan and Waterworth et al,
A Survey of anticoagulation practice among UK cardiothoracic surgeons 52% (97/185) repliesSurvey (Level 3b)Warfarin use for Tissue AVR53% (51/97) of consultants never use warfarin for AVR

47% (46/97) use <= 3 months warfarin
Only a 52 % response rate. Only UK surgeons
Warfarin for Tissue MVR33% (28/86) never use warfarin for MVR

63% (54/86) use <= 3 months of warfarin
Guideline adherence16% (16/97) follow the guidelines for anticoagulation after tissue AVR. 28% (24/86) follow MVR guidelines
Sundt et al,
Retrospective analysis of 1151 patients who underwent AVR bioprosthesis between 1993 and 2000 Anticoagulation 624pts. No anticoagulation 527pts (410p-78% received antiplatelets) Complications obtained by questionnaires sent to patients (87% reponse rate)Cohort study (Level 2b)Incidence of CVAAnticoagulation 2.4%, no Anticoagulation 1.9%, P=0.32Largest study in the literature. Nonrandomized study. Incomplete follow-up study No benefit to early anticoagulation after BAVR.
Mediastinal bleeding requiring exploration5.0% anticoagulation, 7.4% antiplatelets gp
All Bleeding complications1.1% anticoagulation, 0.8% antiplatelet group
Gherli et al,
249 Patients undergoing Biological AVR from 2001 to 2002. Received one of two therapies based on surgeons preference. Warfarin Group 141 pts. Warfarin for 3 months ASA Group: 108 pts. AspirinProspective cohort study (Level 2b)BleedingNo differenceUnderpowered study to exclude hypothesis that anticoagulation may be of benefit.
SurvivalNo difference
Total Cerebral Ischemic events2.8% 4/141 ASA group, 7.4% 8/108 Warfarin group
Major bleeding2.1% ASA group, 3.7% warfarin group
Death2.1% ASA group, 4.6% warfarin group
Heras et al,
816 patients undergoing bioprosthetic replacement of Aortic or/and Mitral valve from 1975 to 1982. -AVR 424pts. -MVR 326pts. -Both valves 66pts. Anticoag.=Heparin or/and warfarin to INR 3.0-4.5 67% mitral and 33% aortic replacements diacharged on warfarin Hancock, Carpentier-Edwards, and Ionescu-Shiley valves used. Follow up data obtained by questionnaireRetrospective Cohort Study (level 2b)Thromboembolisms' rate at 1-10,11-90, and >90 days postop.1-10 days 5 events AVR without warfarin, 0 events AVR with warfarin, 3 events MVR without warfarin, 2 events MVR with warfarin

11-90 days 3 events all valves without warfarin, 6 events all valves with warfarin

>90 days Event rate per year is calculated as 50% per year in the first 10 days for AVR without warfarin in the 424 patients included
Liberal definition of Thromboembolism leading to a rate of 55% of patients recorded as having an event in the first year
Total thromboembolic events2.5%/yr with warfarin, 3.9%/yr without warfarin
Major bleeding.-111patients bled.(2.3%/year). (With/Without anticoagulation), In AVR 6.2% / 1.6%, In MVR 2.7% / 1.1%, In double 4.5%/ 1.9%
Moinuddeen et al,
Retrospective review of 185 patients undergoing tissue AVR from 1987 to 1996 109 had anticoagulation for 3 months 76 pts were not anticoagulatedRetrospective cohort study ( level 2b)Post-operative CVA18% (21/109) anticoagulation gp

18% (14 /76) no anticoagulation gp
Aortic valve replacements only considered here.
Bleeding complications9.2% (10/109) anticoagulation gp

9.2% (7/76) no anticoagulation gp
Survival at 7 years62% anticoagulation gp

67% no anticoagulation gp
Mistiaen et al,
Retrospective analysis of 500 patients with a median age of 73 who received a Carpentier Edwards Pericardial tissue valve from 1986 to 2001 analysed for thrombotic complications Follow up by questionnaire to cardiologists Mean follow up 4.2 yrsRetrospective cohort study (level 2b)Post operative CVA in patients without AF23% (7/30) with warfarin, 6.4% (12/185) with ASA, 6.2% (10/159) no treatmentRetrospective analysis Poor outcome data collection method
Multivariate analysis for risk factors for thromboembolismPrior CVA RR 4.8, p=0.0016, Warfarin usage RR 3.0, p=0.0028, Endocarditis RR 5.6 p=0.006, Hospital thromboembolism RR 6.1 p=0.016
Yao et al,
279 patients undergoing either biological or mechnical MVR from 1973 to 1998. 154 biological, and 125 mechanical. Anticoagulation at discretion of surgeonCohort study (level 2b)15 year freedom from thromboembolismBiological valve 72.2% +/- 3.7%

Mechanical valve 93.5%+/- 3.6% P<0.01
Very small numbers in the anticoagulated Mitral bioprosthesis group
10 year freedom from thromboembolism for bioprostheses71.2%+/-4.8% without anticoagulation (114pts)

100% with anticoagulation (22pts) P=0.049


The most recent guidelines have come from the European Society of Cardiologists in 2005. They recommend that due to the absence of studies that demonstrate the safety of omitting anticoagulation for 3 months post-bioprosthesis, warfarinisation with a target INR of 2.5 or 3.0 in higher risk patients should be given. The ACCP guidelines from 2001 and updated in 2004 recommend warfarin for 3 months for mitral bioprostheses, giving this a grade 1C+ recommendation, and in the aortic position they also recommend warfarin but give this a grade 2C recommendation. They recommend targeting an INR of 2.0–3.0 (Grade 1C). The AHA/ACC guidelines published in 1998 recommended that the highest risk of thromboembolism is in the days immediately post-operatively and thus heparin should be started followed by warfarin for 3 months. However, for patients with a bioprosthesis with no risk factors such as AF, EF<30%, thromboembolism, hypercoagulable state, aspirin with a dose of 80–100 mg/day may be given. This is a grade 1 recommendation. In 1998 the Scottish Intercollegiate Guideline Network recommended warfarin for 3 months for aortic bioprostheses (Grade C) and 3–6 months of warfarin for mitral bioprostheses (Grade A). They recommend an INR target of 2–3. The British Society of Haematology produced guidelines in 1998 (that were unchanged in an update in 2005) recommending that mitral bioprostheses receive 3–6 months of anticoagulation. However, they did not recommend warfarin for aortic bioprostheses although they acknowledged that many institutions did do this. In contrast to the above guidelines that advise 3 months of warfarin therapy, two large surveys have shown that this is not routine practice for aortic valves. In 2004 CTSNet ( performed a survey, with 726 respondents worldwide and found that while 80% of surgeons were aware of current guidelines, 60% did not routinely give 3 months of warfarin. In addition, only 15% of surgeons felt that antiplatelets were not an acceptable alternative to warfarin and over 60% of surgeons thought that warfarin was no longer the standard of care for tissue aortic valves. In 2005 Vaughan and Waterworth performed a survey of UK surgeons. They found that 53% of consultants never use warfarin for bioprosthetic aortic valves, and 33% do not anticoagulate tissue valve replacements. Only 16% of surgeons followed ACCP guidelines. Turning to the original papers, most recently Sundt et al. from the Mayo Clinic in 2005 published a retrospective review of their practice with 1151 patients undergoing tissue AVR, half of whom were anticoagulated. In the 90-days post-surgery, 2.4% of those who were anticoagulated had a stroke compared to 1.9% of patients who were not anticoagulated. There was no difference in bleeding rates or reopening rates. They conclude that while they showed no significant benefit, they also showed no harm due to bleeding rates and acknowledged the underpowered nature of their study. Gherli et al. in 2004 published a study advocating the use of aspirin for tissue aortic valve replacements. They reported that in a study of 249 patients (of whom 108 patients received warfarin according to surgeon's preference) there was no difference in stroke rate. There was also no difference in bleeding rates. Of note only 4 aspirin and 8 warfarin patients had a stroke. Heras et al. reporting from the Mayo Clinic in 1995 provide much of the evidence quoted by the ACCP guidelines. They quote a rate of 50% thromboembolic events per year in the first 10 days after AVR without warfarin but none with warfarin, and for overall follow up for tissue MVR the event rate of 2.5%/year was significantly lower than 3.9%/year without warfarin. However, the significance of the data pertaining to AVRs has been called into question by authors from the same institution. Sundt et al. stated that of the 424 patients who had a tissue AVR only 5 patients had a thromboembolic event in the first 10 days, and thereafter none of the AVR data demonstrated a significant difference. Moinudeen et al. reported in 185 patients that the rate of CVA/TIAs was 18% in both the aspirin and warfarin groups after a mean 5-year follow up. The bleeding rate was not significantly different. They concluded that warfarin was not required for AVR although again this study is too small to exclude the benefit of warfarin in this situation. Mistiaen et al. in 2004 analysed 500 elderly patients receiving a Carpentier–Edwards valve and found on multivariate analysis that use of warfarin actually increased the risk of thromboembolism with a risk ratio of 3.0 after 4-year follow up. While this was a study of 500 patients, only 30 patients in sinus rhythm actually received long-term warfarin, to form this high risk group (7 had a CVA). Yao et al. in 2003 reported that the 10-year freedom from thromboembolism for biological MVRs was 100% if they were anticoagulated for the whole period but only 71% if anticoagulation was not given. Of note there were only 22 patients in the anticoagulation group. The ACCP guidelines quote the paper by Turpie et al. from 1988 to demonstrate that 5% (2/40) of patients with an INR=2.5–4.0 and 5.1% (2/39) with an INR=2.0–2.3 had a thromboembolus after tissue MVR, but the bleeding rate was lower in the low INR group. This paper did not have a 'no-anticoagulation' arm. The study by Ionescu et al. from 1982 is also quoted as evidence in favour of anticoagulation for tissue MVR. In this series of patients from 1971–1981, 5.9% (4/68) who did not receive anticoagulants and 0% (0/182) of patients who received warfarin had an ischemic event during the first 3 months.

Clinical Bottom Line

Patients post-bioprosthetic MVR should have warfarin for 3 months with an INR of 2–3. For patients post-bioprosthetic AVR without additional risk factors, antiplatelet therapy alone is adequate.


  1. E.G. Butchart, C.Gohlke-Barwolf, M.J.Antunes et al. Working Groups on Valvular Heart Disease, Thrombosis, and Cardiac Rehabilitation and Exercise Physiology, European Society of Cardiology. Recommendations for the management of patients after heart valve surgery, European Heart Journal 26 (2005) 2463-2471.
  2. P.D. Stein, J.S.Alpert, H.I.Bussey et al. Antithrombotic therapy in patients with mechanical and biological prosthetic heart valves.[erratum appears in Chest 2001 Sep;120(3):1044]. Chest. 119(1 Suppl):220S-227S, (2001).
  3. D.N. Salem, P.D.Stein, A.Al-Ahmad et al. Antithrombotic therapy in valvular heart disease--native and prosthetic: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. 126(3 Suppl):457S-482S, (2004).
  4. R.O. Bonow, B.Carabello, Leon et al. ACC/AHA Guidelines for the Management of Patients With Valvular Heart Disease. Executive Summary. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Management of Patients With Valvular Heart Disease). Journal of Heart Valve Disease. 7(6):672-707, (1998).
  5. R.O. Bonow, B.Carabello,, Jr. et al. Guidelines for the management of patients with valvular heart disease: executive summary. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Management of Patients with Valvular Heart Disease). Circulation. 98(18):1949-84, (1998).
  6. Lowe et al. Scottish Intercollegiate Guidelines Network Royal College of Physicians of Edinburgh [Online] Antithrombotic Therapy. SIGN Guideline Network. (1999). (Accessed 07/11/06)
  7. British Society of Haematologists. Guidelines on oral anticoagulation. Third Edition, British Journal of Haematology 101 (1998) 387.
  8. Cardiothoracic Surgery Network (CTS-net) Valve technology center. Anticoagulation therapy after aortic Tissue valve replacement. CTSnet (2004).[Accessed May 2006]
  9. P. Vaughan and P.D.Waterworth. An audit of anticoagulation practice among UK cardiothoracic consultant surgeons following valve replacement/repair. Journal of Heart Valve Disease. 14(5):576-82, (2005).
  10. T.M. Sundt, K.J.Zehr, J.A.Dearani et al. Is early anticoagulation with warfarin necessary after bioprosthetic aortic valve replacement? Journal of Thoracic & Cardiovascular Surgery. 129(5):1024-31, (2005).
  11. T. Gherli, A.Colli, C.Fragnito et al. Comparing warfarin with aspirin after biological aortic valve replacement: a prospective study. Circulation. 110(5):496-500, (2004).
  12. M. Heras, J.H.Chesebro, V.Fuster et al. High Risk of Thromboemboli Early After Bioprosthetic Cardiac Valve Replacement. Journal of the American College of Cardiology 25 (1995) 1111-1119.
  13. K. Moinuddeen, J.Quin, R.Shaw et al. Anticoagulation is unnecessary after biological aortic valve replacement. Circulation. 98(19 Suppl):II95-8; discussion II98-9, (1998).
  14. W. Mistiaen, C.P.Van, P.Muylaert et al. Thromboembolic events after aortic valve replacement in elderly patients with a Carpentier-Edwards Perimount pericardial bioprosthesis. Journal of Thoracic & Cardiovascular Surgery. 127(4):1166-70, (2004).
  15. H. Yao, T.Miyamoto, S.Mukai et al. Long-term results of mitral valve replacement: biological xenograft versus mechanical valves. Journal of Artificial Organs 6 (2003) 30-36.
  16. A.G. Turpie, J.Gunstensen, J.Hirsh et al. Randomised comparison of two intensities of oral anticoagulant therapy after tissue heart valve replacement. Lancet. 1(8597):1242-5, (1988).
  17. M.I. Ionescu, D.R.Smith, S.S.Hasan et al. Clinical durability of the pericardial xenograft valve: ten years experience with mitral replacement. Annals of Thoracic Surgery. 34(3):265-77, (1982).