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

What is the optimal level of anticoagulation in adult patients receiving warfarin following implantation of a mechanical prosthetic mitral valve?

Three Part Question

In patients with [mechanical prosthetic mitral valve] receiving long-term [warfarin therapy] what is the [optimal therapeutic INR target]?

Clinical Scenario

You have inserted a mechanical prosthetic mitral heart valve into a 60-year-old man. He subsequently has a GI bleed while on warfarin so you decide to review the literature to confirm optimal therapeutic INR which minimises long-term thromboembolic and haemorrhagic complications.

Search Strategy

Medline 1966 to 2006 November Week 3 using OVID interface.
EMBASE 1980 to 2006 Week 50.
[exp Cardiopulmonary Bypass/OR OR exp Thoracic Surgery/OR Coronary art$ OR Cardiopulmonary OR exp Cardiopulmonary Bypass/OR exp Cardiovascular Surgical Procedures/OR exp Thoracic Surgical Procedures/OR exp Coronary Artery Bypass/OR cardiac OR exp Heart Transplantation/OR mitral OR exp Mitral Valve/OR exp Heart Valve Prosthesis/OR exp Heart Valve Prosthesis Implantation/OR mitral valve OR exp Bioprosthesis/OR mitral valve] AND [ OR exp International Normalized Ratio/OR OR exp Warfarin/] AND [Thrombosis/OR valve OR OR Thromboembolism/OR stroke. mp. OR exp Cerebrovascular Accident/OR exp Cerebral Hemorrhage/OR exp Subarachnoid Hemorrhage/OR exp Hemorrhage/OR OR exp Gastrointestinal Hemorrhage/OR OR postoperative OR exp Postoperative Complications/OR exp Prosthesis Failure/OR prosthetic valve] AND [limit to adults].

Search Outcome

A total of 894 papers on Medline and 1235 on Embase were found using the above search. All major guidelines were included, and their reference lists searched. Twelve papers were considered to represent the best evidence on the topic and are summarised in the table

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Stein et al,
Sixth American College of Chest Physicians (ACCP) consensus conference on antithrombotic therapy in patients with mechanical or biological prosthetic heart valvesConsensus report based on case series, non-randomized and randomized trials (Level 2a)Recommendations for optimal INR range to prevent thromboembolic and haemorrhagic complications post-mechanical MVRTarget INR range 2.5-3.5 for bi-leaflet and tilting disc valves

Higher INR recommended for caged ball valves

In patients with additional risk factors for thromboemboli (e.g. poor EF, enlarged LA, AF), or who suffer systemic embolisation, combine warfarin therapy with aspirin 80-100 mg/day
Salem et al,
Seventh American College of Chest Physicians (ACCP) consensus conference on antithrombotic and thrombolytic therapyConsensus report based on case series, non-randomized and randomized trials (Level 2a)Recommendations for optimal INR range to prevent thromboembolic and haemorrhagic complications post-mechanical MVRTarget INR 3.0 (range 2.5-3.5) for bi-leaflet and tilting disc valves

Target INR 3.0 (range 2.5-3.5) combined with aspirin 75-100 mg for caged-ball/disc valves

Add Aspirin 75-100mg/day in patients with additional risk factors (e.g. poor EF, enlarged LA, AF, MI), or in those with systemic embolisation despite adequate INR
Bonow et al,
American College of Cardiology (ACC) / American Heart Association (AHA) Practice GuidelinesReview based on cohort studies, case series (Level 2a)Recommendations for optimal INR range to prevent thromboembolic and haemorrhagic complications post-mechanical MVRRecommended Target INR 2.5-3.5 (Class I recommendation)

Tilting-disc valves may be more thrombogenic and could justify INR  to 3.0-4.5 although this also 's risk of bleeding

Addition of aspirin 80-100 mg/day to INR 2.0-3.5 's risk of thromboembolism & CV mortality – partic in high-risk patients with known thrombo-embolic risk
Bonow et al,
American College of Cardiology (ACC) / American Heart Association (AHA) Practice GuidelinesReview based on cohort studies, case series (Level 2a)Recommendations for optimal INR range to prevent thromboembolic and haemorrhagic complications post-mechanical MVRTarget INR 2.5-3.5 (Class I; Level C evidence recommendation) for MVR with mechanical valve

Addition of Aspirin 75-100mg/day recommended for all patients with mechanical valve (Class I; Level B evidence recommendation)

Consider Clopiogrel (75mg/day) in pts unable to tolerate aspirin
Butchart et al,
Guidelines from the Consensus Committee of the European Society of Cardiology (ESC) Working Groups on Valvular Heart Disease, Thrombosis and Rehabilitation / ExerciseConsensus report based on cohort studies (Level 2a)Recommendations for optimal INR range to prevent thromboembolic and haemorrhagic complications post-mechanical MVRIdentify need to combine patientRecognises paucity of clinical information, particularly related to thromboembolic complications Recommendation to adjust INR to intra-cardiac conditions and prosthesis thrombogenicity. Recommendations may need to be revised in individual patient if associated with pathological bleeding Recommend careful evaluation of newly introduced prostheses
Baglin et al,
Guidelines from the British Committee for Standards in Haematology (BCSH)Guidelines based on cohort studies (Level 2a)Recommendations for optimal INR range to prevent thromboembolic and haemorrhagic complications post-mechanical MVRTarget INR 3.0 for bileaflet and tilting disc valves

Target INR 3.5 for older generation (caged-ball) valves
Overall similar to the ACCP recommendations although no comments on combination with aspirin or dipyridamole
Lowe et al,
Scottish Intercollegiate GuidelinesGuidelines based on available cohort & RCT's (Level Ib-IV)Need for anticoagulation therapyPatients with mechanical heart valves require long-term warfarin therapy (level IIa, IIb, III evidence)Varying levels of evidence Significant amount of evidence remains "expert opinion"
Recommendations for first generation valves (Starr-Edwards, Bjork-Shiley)INR 3.0-4.5 (level IV evidence)
Recommendations for second generation valves (St.Jude, Medtronic, Monostrut)INR 2.5-3.5 (level IIa, IIb, III evidence)

Consider additional anti-platelet therapy - aspirin. Dypyridamole – for patients who suffer systemic embolisation despite adequate, intensity warfarin therapy (level Ib evidence)
Butchart et al,
345 post-Medtronic Hall MVR (single tilting disc) patients 183 anticoagulated to mean INR of 2.51 vs. 162 anti-coagulated to mean INR of 3.07Cohort study (Level 2b)Event free survival for valve thrombosis100% vs. 100% event-free survival @ 3 yrsMedtronic Hall valve only Not randomized. Patients with lower INR's had surgery 5-10 years before patients with higher INR's – could be confounding factor in apparent outcome in terms of changing patient characteristics and level of anticoagulation over time.
Event-free survival for embolic (neurological, lmb or visceral ischaemia) & haemorrhagic (all bleeding) events80% (all events); 93% (serious events) vs. 89% (all events); 98% (serious events) event free survival @ 3 yrs

Concluded optimal INR 3.0 post-MVR

All comparisons lower INR vs. higher INR
Cannegieter et al,
1608 patients with mechanical heart valves attending anticoagulation clinic - 60% AVR - 30% MVR - 10% AVR+MVR 123254 INR measurements over 6 yearsRetrospective cohort (Level 2b)Total incidence of adverse thromboembolic & bleeding events in different INR ranges

ction, peripheral embolism, valve thrombosis

Episodes of major bleeding, intracranial & spinal bleeding, major extracranial bleeding
Adverse outcome results expressed as number of events / 100 pt-yrs for various INR ranges

INR 1.0-1.4: adverse events 27/100 pt-yrs (95%CI 3.3-99)

INR 2-2.4: adverse events 7.5/100 pt-yrs (95%CI 3.6-12.6)

INR 2.5-4.9: adverse events 2/100 pt-yrs (95%CI 1.0-3.8)

INR 5.0-5.5: adverse events 4.8/100 pt-yrs (95%CI 2.6-7.7)

INR >6.5: adverse events 75/100 pt-yrs (95%CI 54-101)

77/164 (47%) bleeding events associated with trauma / other underlying pathology

INR 2.5-4.9 also lowest risk when adverse outcome limited to death / stroke leading to handicap

No signif difference in incidence at different INR levels (i.e. same proportion of warfarin-related bleeding at high & low INR's)

Study biased towards AVR and tilting disc valves No specific sub-group analysis of mitral valves other than to acknowledge their higher thrombogenicity
Optimal level of anticoagulationOptimal level of anticoagulation with lowest risk of complications = 2.5-4.9
Tiede et al,
Literature review of cohort studies, case seriesSystematic review (Level 2a)Prevention of adverse thrombo-embolic and haemorrhagic eventsTarget INR for tilting-disc valves 3.3-3.7

Target INR all other mechanical valves 2.8-3.2
Generally recommended to combine with aspirin 80 mg/day in patients<70 years unless contraindications. Increase INR target by 0.5-1.0 and consider dipyridamole in patients with thrombo-embolic disease
Vink et al,
9808 patients from 30 studies followed up over 45,360 patient-years Comparison of thromboembolic & bleeding complications in pts with mean INR >3.0 vs. mean INR <3.0Meta-analysis (Level 2a)Incidence of valve thrombotic eventsValve thrombosis - 2.06 vs. 3.44 (RR=0.60; 95%CI 0.47-0.76; p<0.0001)Mean age of recipients at implantation 55 years – outcomes may not extrapolate to older population Study looked at all types of mechanical valve replacement. Careful control of definitions of "thromboembolism" & "bleeding events" Patients excluded if on anti-platelet therapy More studies / patient-years' data for lower INR group Included studies are cohort studies without controls Studies based on an "intention to treat" with no info on compliance / achieved intensity of anticoagulation which may confound incidence of bleeding events Studies included both older & newer generation valves
Incidence systemic embolic eventsSystemic embolism -15.91 vs. 20.12 (RR=0.79; 95%CI 0.74-0.84; p<0.0001)

Overall incidence of thromboembolism -17.11 vs. 23.13 (RR=0.74; 95%CI 0.70-0.78; p<0.0001)
Incidence of bleeding eventsRisk of bleeding event - 12.94 vs. 11.96 (RR=1.08; 95%CI 1.00-1.16; p=ns)

Total number of thromboembolic & bleeding events – 29.76 vs. 35.33 (RR=0.84; 95%CI 0.79-0.89; p<0.0001)

Recommended target INR > 3.0

All comparisons higher vs. lower INR reported as events/1000pt-yrs
Pengo et al,
205 patients with valve replacement followed for 1.5 to 4.5 years. - 53% AVR - 35% MVR - 12% AVR+MVR - 104 patients with INR target 3 (MVR 38%) vs. 101 patients with INR target 4 (MVR 32%)RCT (Level 1b)Incidence of Major bleeding4 vs. 11 events (1.2 vs. 3.8/100 pt-yrs); p=0019Largest proportion of bleeding complications were epistaxis (42%) and bruising (17%). Thrombo-embolic episodes associated with AF. All thromboembolic episodes were TIA's with full recovery No subgroup analyses for different valve sites
Incidence of Minor bleeding85 vs. 123 (26 vs. 43/100 pt-yrs); p=0.001
Incidence of systemic thrombo-embolism6 vs. 6 events (1.8 vs. 2.1/10 pt-yrs); p=ns
Incidence of vascular death3vs. 3 events (09. vs. 1/100pt-yrs); p=ns


The thrombogenic characteristics of prosthetic mechanical heart valves are well recognised [Friedli, Cannegieter] necessitating postoperative anticoagulation therapy. However, it is important to achieve an anticoagulation balance which prevents both adverse thromboembolic events [Butchart 2004, Rosendaal] and bleeding [Levine]. There can also be variability of warfarin effects which when outwith a predetermined 'target' INR range is associated with higher risks of valve related thromboembolic or haemorrhagic mortality [Butchart 2004, Rosendaal]. Finally, prosthetic valves themselves vary in their thrombogenic properties [Vink, Kuntze, Butchart 1991]. Several major guidelines have been published which attempt to identify an 'optimal' target INR for patients with mechanical mitral valves. However, with a scarcity of large randomised trials, most guidelines are based on cohort studies and case series, a fact acknowledged by the groups themselves. The sixth and seventh ACCP guidelines recommend a target INR of 2.5–3.5 in patients with tilting disc and bi-leaflet mechanical valves in the mitral position. For older generation valves (caged ball or disc) or in patients with additional risk factors, the addition of low dose aspirin 75–100 mg/day is recommended. Previous concerns about increased bleeding events in patients receiving both aspirin and vitamin-K antagonists are somewhat allayed by studies demonstrating significant decreases in thromboembolic events with combination therapy without significant differences in bleeding events [Vaughan]. Comparable results have been reported in other studies [Turpie]. However, as emphasised in these guidelines the benefits of anti-platelet therapy combined with warfarin needs to be evaluated on the basis of total daily aspirin dose and variability of INR within and between groups of patients. Dose-dependent bleeding risk with aspirin combined with warfarin is also reported in the 1998 AHA/ACC guidelines which were revised in 2006 [Bonow, 1998, 2006]. In parallel with the ACCP guidelines these recommend an INR target of 2.5–3.5 for newer generation mechanical mitral valves. A target of 3.0–4.5 should be considered in patients with valves of increased thrombogenicity. However, whilst the 1998 guideline suggests supplementation with aspirin in those patients with additional thromboembolic risk factors, by 2006 it has become a Class I Level B recommendation with Clopidogrel being advised for those unable to take Aspirin. The most recent European guidelines have come from the European Society of Cardiologists (ESC) (Bonow), the British Committee for Standards in Haematology (BCSH) (Baglin) and The Scottish Intercollegiate Guidelines Network (SIGN) (Lowe). In common with the American guidelines, these support a target INR of 2.5–3.5 for second generation prosthetic mechanical mitral valves. However, the European and UK based guidelines are more restrained regarding supplementary aspirin with the BCSH making no comment and the ESC and SIGN advocating a cautious approach to addition of an antiplatelet agent, recommending use in high-risk cases only after thorough risk factor identification and warfarin optimisation. However, despite these slight differences between USA and European guidelines no studies have examined whether these produce clinically different outcomes. As acknowledged by all guidelines there are few good RCTs or cohort studies available. This may in part reflect ethical considerations at a time when expert committees are producing contemporary evidence-based guidelines. In 1991, Butchart et al. published a cohort study comparing mean target INRs of 2.5 and 3.0 in patients who had received a Medtronic-Hall mitral prosthesis. Embolic and haemorrhagic event-free survival was greater with a higher INR and they concluded that an INR above 3 was appropriate. However, there were possible confounding factors with this study, not least that patients with lower INRs had surgery 5–10 years earlier than those with higher INRs. n 1995 Cannegieter et al. studied adverse haemorrhagic or thromboembolic events in patients who had undergone valve replacement reporting an optimal INR range of 2.5–4.9. A literature review by Tiede et al. in 1998 recommended an INR of 3.3–3.7 for tilting disc and 2.8–3.2 for other valves . A more recent meta-analysis including nearly 10,000 patients from 30 studies concluded that an INR over 3.0 was associated with lower risks of thromboembolic complications without increased bleeding risk (Vink). However, their study patients were relatively young and excluded if on anti-platelet therapy, and they did not provide data on either the achieved INR levels or INR variability. Unfortunately these three studies included all types of valve and valve position with little if any subgroup analysis. Finally, an RCT by Pengo et al. found no significant difference in systemic thromboembolism and vascular death between patients with target INRs three and four, although 'major' and 'minor' haemorrhage were significantly greater in the higher INR group. Once again this study included all valve types with no subgroup analysis.

Clinical Bottom Line

In patients with newer generation mechanical prosthetic mitral valves the INR should be maintained at 2.5–3.5. For older type valves the INR should be 3.5–4.5. Some patients with no contraindications to anti-platelet therapy at increased risk of thromboembolic episodes may benefit from additional low-dose aspirin therapy (75–100 mg/d). To decrease the risk of thromboembolic and haemorrhagic events associated with anticoagulation therapy it is imperative to ensure a low variability in the level of anticoagulation.


  1. Stein PD, Alpert JS, Bussey HI et al. Antithrombotic therapy in patients with mechanical and biological prosthetic heart valves. Chest 2001; 119:220S–227S.
  2. Salem DN, Stein PD, Al-Ahmad A et al. Antithrombotic therapy in valvular heart disease – native and prosthetic: The Seventh ACCP conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126:457S–482S.
  3. Bonow RO, Carabello B, de Leon AC 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 Circulation 1998; 98:1949–1984.
  4. Bonow RO, Carabello BA, Chatterjee K et al. ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: executive summary. (Writing Committee to Revise the 1998 Guidelines for the Management of Patients ... Circulation 2006; 114:450–527.
  5. Butchart EG, Gohlke-Barwolf C, Antunes MJ et al. European Society of Cardiology. Recommendations for the management of patients after heart valve surgery. Eur Heart J 2005; 26:2463–2471.
  6. Baglin TP, Keeling DM, Watson HG. Guidelines on oral anticoagulation (warfarin): third edition – 2005 update. Br J Haematol 2005; 132:277–285.
  7. Lowe G, Belch J, Burton C et al. Antithrombotic therapy. Scottish Intercollegiate Guidelines Network (SIGN), 1999; Publication number 36.
  8. Butchart EG, Lewis PA, Bethel JA et al. Adjusting anti-coagulation to prosthesis thrombogenicity and patient risk factors. Recommendations for the Medtronic Hall valve. Circulation 1991; 84:(Suppl 3), 61S–69S.
  9. Cannegieter SC, Rosendaal FR, Wintzen AR et al. Optimal oral anticoagulant therapy in patients with mechanical heart valves. N Engl J Med 1995; 333:11–17.
  10. Tiede DJ, Nishimura RA, Gastineau DA et al. Modern management of prosthetic valve anticoagulation. Mayo Clin Proc 1998; 73:665–680.
  11. Vink R, Kraaijenhagen RA, Hutten BA et al. The optimal intensity of vitamin K antagonists in patients with mechanical heart valves. A meta-analysis. J Am CollCardiol 2003; 42:2042–2048.
  12. Pengo V, Barbero F, Banzato A et al. A comparison of a moderate with a moderate-high intensity oral anticoagulant treatment in patients with mechanical heart valve disease. Thromb Haemostasis 1997; 77:839–844.
  13. Friedli B, Aerichide N, Grondin P et al. Thromboembolic complications of heart valve prostheses. Am Heart J 1971; 81:702–708.
  14. Cannegieter SC, Rosendaal FR, Briet E. Thromboembolic and bleeding complications in patients with mechanical heart valve prostheses. Circulation 1994; 89:635–641.
  15. Butchart EG, Payne N, Li HH et al. Better anticoagulation control improves survival after valve replacement. J Thoracic Cardiovasc Surg 2002; 123:715–723.
  16. Rosendaal FR, Cannegieter SC, van der Meer FJ et al. A method to determine the optimal intensity of oral anticoagulant therapy. Thrombosis and Haemostasis 1993; 69:236–239.
  17. Levine MN, Raskob G, Landefeld S et al. Hemorrhagic complications of anticoagulant treatment. Chest 2001; 119:108S–121S.
  18. Kuntze CEE, Ebels T, Eijgelaar A et al. Rates of thromboembolism with three different mechanical heart valve prostheses: randomised study. Lancet 1989; 8637:514–517.
  19. Butchart EG, Lewis PA, Bethel JA et al. Adjusting anticoagulation to prosthesis thrombogenicity and patient risk factors. Circulation 1991; 84:Suppl_III61S–69S.
  20. Vaughan P, Waterworth PD. An audit of anticoagulation practice among UK cardiothoracic consultant surgeons following valve replacement/repair. J Heart Valve Dis 2005; 14:576–582.
  21. Turpie AGG, Gent M, Laupacis A et al. A comparison of aspirin with placebo in patients treated with warfarin after heart valve replacement. N Engl J Med 1993; 329:524–529.