Prevalence of CHD rises sharply with age, peaking in the 65–74 year old age group.2 In addition about 14% of men and 8% of women will have angina in this age group.3 The usual age for first myocardial infarction (MI) for a man is 65 years old and for a woman 73 years old. There are about 103,000 MIs in the UK every year and the case fatality (defined as within 30 days of hospital admission with MI) stands at 10% for men and 15% for women at all ages in 2010.2
The use of aspirin for patients in primary prevention of CHD is not recommended. This article overviews secondary care exploitation of the coagulation and fibrinolytic systems in acute coronary syndromes (ACS) to familiarise GPs with secondary care practice in a fast moving field. It also discusses those drugs which GPs may encounter in primary care and reviews current NICE guidance.
Changes in practice
Before the early 1980s patients with a clinical MI might have been cared for by their GP at home or admitted to hospital. Now all patients are admitted unless they decline to go. The criteria for action according to length of chest pain is shorter than past clinical practice; a patient with angina who develops chest pain should use their GTN spray and repeat this after five minutes if the pain has not gone. If it still hasn’t gone after a further five minutes they should call 999.3
In practice this means that GPs see much fewer patients with acute chest pain than previously but still need to provide skilled care for those patients presenting acutely. The importance of quick therapy relates to figures showing that patients who call for help and arrive at hospital within 60 minutes of chest pain with a ST-elevation MI (STEMI) have a 6.1% death rate at 18 months, but those waiting over six hours before hospital arrival have a death rate of 14.9% at 18 months.4
The overarching term for unstable chest pain is ACS and can be subdivided into unstable angina (UA), non-ST elevation MI, NSTEMI or ST elevation MI (STEMI). There are also patients with known coronary artery atheroma causing chronic angina and those with undiagnosed atheroma.
The treatment of ACS has changed considerably over the decades from conservative home care, to admitting the patient for analgesia and CCU monitoring (largely to treat possible ventricular fibrillation), to active intervention to restore coronary artery patency. NICE STEMI guidance states that nearly half of potentially salvageable myocardium is lost within one hour of the coronary artery being occluded.5
There was widespread use of fibrinolytic agents in the late 1980s and 1990s and streptokinase was used as a "clot buster". Only 5% of STEMI patients received fibrinolysis in 2011/12.5 The term PCI in the NICE STEMI guidance has been further expanded to PPCI, primary percutaneous coronary intervention to incorporate coronary angioplasty, thrombus extraction catheters and stent insertion. Still 30% of patients with STEMI will receive no reperfusion therapy for a wide variety of reasons.5
TIMI grade flow
TIMI derives from the Thrombolysis in Myocardial Infarction study and turns up in medical discharge letters sometimes. It refers to the flow through the coronary artery with TIMI 0 being complete occlusion with no perfusion and TIMI 3 being complete perfusion.
TIMI risk score
This can be used in patients with UA and STEMI to risk score the patient for possibility of death or further ischaemia.6
GRACE risk score
This guidance is used to predict patient mortality at six months and helps decide therapy and management.7 Developed in 2003 it looked at over 11,000 patients with ACS who had not died in the first 24 hours of ACS and used eight risk factors to predict inpatient death and death at six months. It has been studied since and over longer periods of time. The risk factors for adverse outcome are: advancing age, severity of heart failure, pulse rate, systolic blood pressure, renal function, ST changes on ECG, raised troponins and cardiac arrest at admission. It is calculated when the patient reaches hospital care and is used to weigh the risks of adverse outcomes from UA or NSTEMI against the risks of therapy ie. bleeding risk mainly. It therefore stratifies patients into low risk (<1.5%) up to highest risk (>9%) of death at six months post ACS and uses this to suggest therapy.7
The European Society of Cardiology defines a MI as myocardial death due to prolonged myocardial ischaemia detectable biochemically by a rise in cardiac troponin and at least one of:
• Ischaemic symptoms ie. chest pain with possible radiation commonly to left arm, jaw, into back, vomiting, sweating (the absence of chest pain is more usual in the elderly and in diabetics)
• ECG changes: new left bundle branch block (LBBB) or ST changes
• ECG changes: development of pathological Q waves
• Imaging evidence of myocardial death or loss of regional cardiac motion
• Angiography or autopsy finding of intracoronary artery thrombus.8
MI occurs usually due to atheroma and thrombus formation but can occur due to reduction in oxygen supply to the heart muscle for other reasons such as tachyarrhythmias, bradyarrythmias, aortic valve disease, aortic dissection, hypertension, hypoxia, anaemia, shock, coronary artery spasm, vasculitis, drug injury, myocarditis, severe illnesses like renal failure or sepsis and injury related to coronary interventions.8
GP action in ACS
The GP should take a careful history and examine the patient. Once the diagnosis of ACS is made the patient should take aspirin 300mg soluble stat. The GP should make a note of the dose and time administered when sending the hospital referral with full past medical and drug history. Aspirin is "anti-platelet aggregation" acting by inhibiting COX enzyme and so preventing production of certain prostaglandins and thromboxane production, all of which encourage platelet aggregation.
If possible the GP should arrange an ECG as soon as possible whilst awaiting the 999 ambulance. Because speed is all important in deciding outcomes, patients are best sent to hospital by ambulance rather than making their own way there. Of note a posterior MI may appear as ST depression in the anterior chest leads V1–V4. GPs with an interest in cardiology might place leads on the posterior chest to pick up ST segment elevation. There may be rhythm problems, eg. complete heart block, which is usually associated with inferior MI because the right coronary artery usually supplies the SA node, AV node and the inferior heart. The resting rate of myocardium when the SA and AV nodes are not functioning is about 30–40 beats per minute. It is a generalisation that anterior MIs are the most severe (ST elevation in leads V1–4).5
Defibrillation equipment should be brought near and extra staff should attend in case of arrest. Patients who can tolerate therapeutic oxygen and have a finger oxygen stat below 95% should be given oxygen.
GPs may administer opiate and antiemetic IM, most of us are not skilled in IV access. If the patient is more likely to have angina and is not too hypotensive (systolic blood pressure below 90mmHg) then GTN spray can be administered.
This is the more severe MI, defined as a rise in troponin with ST segment rise or new LBBB and usually with associated features of chest pain. After presentation and diagnosis the patient ideally undergoes reperfusion therapy as soon as possible to restore coronary artery patency and improve myocardial oxygenation. Patients need to present within 12 hours of symptoms or have continuing pain suggesting ischaemia beyond that time. Reperfusion may occasionally be via fibrinolysis or more commonly via PPCI.
PPCI has been shown to be superior to fibrinolysis at 30 day and six month trial review by NICE provided it is given at the recommended times.5 Fibrinolysis is not routinely given pre-PPCI as there is a small increase in intracranial bleeding rates, once given it is not repeated in that episode if ischaemia continues. Fibrinolysis is given prehospital or in A&E if patients are within 12 hours of symptoms. PPCI cannot be given within two hours of presentation to healthcare staff but fibrinolysis can be administered within 30 minutes of presentation. In 2011–2012 92% of eligible patients with STEMI were able to have PPCI within 90 minutes of presentation in the UK.5
The drug given may be alteplase, reteplase or tenecteplase, which are all plasminogen activators acting on the fibrin surface to lyse the fibrin and break down thrombus. Exposed thrombus encourages coagulation so a cocktail of other drugs is also given: aspirin to prevent platelet aggregation; clopidogrel to inhibit platelet ADP receptors and so platelet aggregation and binding to fibrin and low molecular weight heparin, LMW heparin, as enoxaparin or unfractionated heparin (UH) (activates antithrombin and inactivates factor 10) or less commonly fondaparinux (F10 inhibitor) may be given for up to eight days in hospital or until PPCI.
In essence most patients will have a cocktail of aspirin, enoxaparin and clopidogrel. It was felt by NICE that as less than 2000 patients with STEMI per year now have fibrinolysis then using the same drugs was important. Patients receiving fibrinolysis may then go on to PPCI if indicated by continual ischaemia.9
NICE reports that 25% of people are contraindicated to have fibrinolysis due to bleeding risk or comorbidity.5 In those patients given fibrinolysis within 90 minutes 75% gain coronary artery patency and 50% have complete reperfusion. 13% may require blood transfusion due to bleeding, 1% may have a disabling haemorrhagic intracranial bleed within 24 hours. The risk of intracranial bleed is increased if the patient is female, older, has low body weight, high blood pressure or a previous history of intracranial bleed. Contraindications include these and risks of recent GI bleeding, ischaemic stroke in the last six months, or trauma in the last three weeks.9
There are a number of other relative risks and the specialist needs to weigh these against the benefit for that patient.
PPCI was brought in nationally after the government’s National Service Framework coronary heart disease update in 2008.4 PPCI restores coronary artery blood flow (TMI3) to normal in 90% of patients.5 The mortality rate for patients after PPCI as inpatients was estimated at 3.5% and at 18 months was 7%. Overall inpatient stroke rate was 0.8% for PPCI. Overall inpatient bleeding rate was 3.5% in the PPCI group compared to 6.2% in patients undergoing fibrinolysis. Patients having PPCI who were admitted directly to the catheter laboratory, bypassing hospital A&E departments, had shorter times to treatment and reduced mortality rates.4
After a diagnosis of STEMI the patient receives a number of anticoagulants to improve perfusion and prevent further thrombus and then undergoes angioplasty usually with stent insertion and continuing anticoagulant medication to take home.
The place of novel oral anticoagulants (NOACs) in this procedure is still being studied but NICE reports via technology appraisals as evidence becomes available. Treatment is irrespective of age and therapy relies on balancing the risks of bleeding with benefits of coronary reperfusion. PPCI can be used in patients who are in cardiogenic shock and those who have had a cardiac arrest. The patient initially receives aspirin 300mg given in primary care and in secondary care has one of three oral ADP receptor antagonist antiplatelet medications: clopidogrel initially as a loading dose or prasugrel or ticagrelor (see part 2 of this series). The last two medications work more quickly at 30 minutes rather than clopidogrel at 3–4 hours. Prasugrel also has an indication in diabetics undergoing PPCI or in patients who have failed clopidogrel therapy.10 All three should be stopped a few days prior to CABG as they increase the risk of bleeding. The patient is preferably directly sent or is transferred to a 24 hour PPCI centre provided PPCI can be offered within 120 minutes of the time that fibrinolysis could be given.
At hospital the patient is given heparin, usually LMW heparin, or bivalirudin. Heparin is given if either of the two NOACs, prasugrel and ticagrelor are used instead of clopidogrel.5
Bivalirudin is a thrombin inhibitor used in patients undergoing PCI and is used with clopidogrel and aspirin, it is not licensed with prasugrel or ticagrelor. Bivalirudin also has a specific use in patients prone to HIT or HITT (heparin induced thrombocytopoenia).
Cardiologists may administer another antiplatelet group, the GP11B/111A platelet receptor antagonists, GPIs, which are abciximab or eptifibatide or tirofiban. In the UK 52.7% of people undergoing PPCI during 2011 were given a GPI.5
A catheter is fed via the radial or femoral artery to the coronary artery for angiogram. Thrombus may be aspirated but this is not routinely done. The thrombosed coronary artery is then balloon angioplastied with a stent being expanded in the occluded area. This stent is a bare metal or a drug eluting stent. The main complication of stents is restenosis. Bare metal stents have a 20% restenosis rate by re-endothelialisation at six months11 and so drug eluting stents were produced to prevent this but, as they can lack endothelialisation this can cause thrombus. Neither stent has currently been found to have superior outcomes.
Patients with UA or NSTEMI were followed up for a year to compare PPCI versus conservative therapies. Rates of death or MI in the two groups were similar at about 8% but there was a halving of angina in the PCI group with less medication use.12 Therefore patients presenting with NSTEMIs are subdivided from high risk to low risk of death using the GRACE risk score in order to decide the best therapeutic strategies and need for angiography. Patients at lowest risk via GRACE score (<1.5% risk of death at six months) receive aspirin and assessment of ischaemia. Grace scores of 1.5–3% are low risk and they may also be managed conservatively but would have dual antiplatelet therapy, referred to as DAPT, with clopidogrel for 12 months.7 Most patients with NSTEMI have a GRACE risk score over 3% and they receive aspirin with clopidogrel for 12 months, DAPT, and undergo PPCI.
A number of different anticoagulants, many acting on platelets and antithrombin are used to improve perfusion and prevent further thrombus formation at the time of STEMI prior to PCI as with STEMIs.
Most patients presenting with NSTEMI are given 300mg of aspirin as soon as possible and then clopidogrel at a high loading dose 300–600mg stat and they then continue low dose aspirin and clopidogrel for 12 months. The risk of major bleeding in the CURE study on aspirin was 2.7% and when clopidogrel was combined was 3.7% with no increased mortality.13
The NOAC alternatives to clopidogrel are prasugrel and ticagrelor which with aspirin can be recommended for the patient. Ticagrelor is recommended particularly in patients with UA who are over 60 years old and have previous atheroma, it reduces the absolute risk of time to MI, CVA or death from vascular disease by 1.9% compared to aspirin with clopidogrel. It has an increased risk of breathlessness, raised uric acid, raised creatinine and short ventricular pauses compared to clopidogrel but no increased risk of major bleeding.
Prasugrel is indicated if there has been failure with clopidogrel use or the patient is a diabetic and has been shown to reduce the rate of MI compared to clopidogrel and aspirin but also increased the risk of major bleed from 1.8% on aspirin and clopidogrel to 2.4% on prasugrel and aspirin with a small increase in fatal bleeds.10 A PPI, not omeprazole, should be given with DAPT if the patient has a history of GI bleeding or PU or multiple risk factors for bleeding such as warfarin, H. Pylori infection, over 65 years old, SSRIs or on prednisolone.14
The patients at higher risk Grace score (>3%) next receive antithrombin therapy and angioplasty either within 24 hours or within four days (96 hours). If angioplasty is planned within 24 hours then the antithrombin therapy is UH, also used if the patient has significant renal impairment. LMW heparin, enoxaparin, may also be used. If angiography is planned after 24 hours then fondaparinux is recommended (anti factor 10 so preventing prothrombin becoming thrombin.) An alternative antithrombin to heparin is bivalirudin but the patient does not then receive fondaparinux or a GPI antiplatelet (eptifibatide or tirofiban). UH is also given in the PPCI lab to the fondaparinux treated group. Fondaparinux reduced bleeding risk and was as effective as enoxaparin making it the preferred antithrombin for most patients having PPCI.7 Patients at over 3% Grace risk may also receive further platelet inhibition via a GPI, eptifibatide or tirofiban or abciximab.
Some patients will require CABG and it is up to the cardiac surgeon to discuss stopping medications in advance if indicated with the patient but bleeding due to clopidogrel, ticagrelor and prasugrel is reduced if they are stopped a few days pre-surgery.7
The European Society of Cardiology (ESC) suggest that patients over 75 years old are more likely to present with atypical symptoms and treatment decisions should be made in the context of their wishes, their estimated life expectancy, their comorbidities and quality of life. Choices and doses of antithrombotic drugs should be tailored to consider the risks of bleeding and renal impairment and the elderly should be considered for early angiography and possible PCI if fit enough.14
Aspirin is known to reduce death from MI and cardiovascular events in patients who have had a MI. A 1970s study suggested that placebo patients had a death rate of about 8.5% over an average follow up of 22 months, reduced to 5.8% by quite high dose aspirin.15 Aspirin with a proton pump inhibitor (PPI) was better than using clopidogrel alone in the prevention of bleeding complications.16 Aspirin should be continued long-term. Patients intolerant of aspirin should take clopidogrel monotherapy long-term. For patients who have other evidence of atheroma eg. intermittent claudication, clopidogrel can be considered as a long-term option.
In effect patients post MI usually receive DAPT as aspirin and clopidogrel for 12 months unless contraindicated because the majority have had PPCI and stent insertion. Adding in clopidogrel reduces mortality rate but there is a rise in bleeding rate. The CURE study had a rate of major bleeding in NSTEMI patients of 2.7% on aspirin and 3.7% on aspirin and clopidogrel combined.17 This is why a bleeding assessment needs to be undertaken. Clearly our once fit patient ages and changes medically and the GP needs to review the bleeding risk for that patient if circumstances change. Prasugrel or ticagrelor are also ADP platelet inhibitors and may be used with aspirin. Guidelines now suggest that stopping the clopidogrel or similar drug at 12 months should be a matter for clinical judgement and best evidence combined, particularly over concerns about late stenosis of stents, not an automatic stoppage.
On occasion (STEMI with conservative or fibrinolysis therapy or post CABG) the length of clopidogrel therapy may be shortened to four weeks or with consideration of prolonging clopidogrel as DAPT for 12 months.16
Patients requiring warfarin for other reasons (eg. AF) run a higher risk of bleeding complications with antiplatelet therapy combined. A review of guidelines of triple therapy with aspirin, clopidogrel and warfarin suggests that there is not enough evidence to make firm recommendations about risk but in general the bleeding risk was increased by triple therapy.18 If a patient is taking a NOAC on admission instead of warfarin then patients presenting with ACS should, if possible, be switched to warfarin. Patients undergoing PPCI and stenting would then continue with warfarin and clopidogrel for 12 months after which their antiplatelet use should be reviewed. Patients undergoing conservative therapy, balloon angioplasty or CABG should take warfarin and aspirin for 12 months and then the use of aspirin should be reviewed. If the patient cannot tolerate aspirin then clopidogrel could be used.16 There is no guideline for combining warfarin with prasugrel or ticagrelor.
Patients whose history of STEMI and NSTEMI is over 12 months ago should receive aspirin long-term unless they are aspirin intolerant when they should receive clopidogrel. In addition if they also have other features of atheroma then clopidogrel long-term is preferable to aspirin, this is a change from past practice.16
Patients with stable angina
Patients with stable angina but continuing symptoms despite therapy should be considered for either CABG or PCI as both these procedures can relieve anginal symptoms. Their antiplatelet therapy would then be determined by procedure and stent insertion as above. All patients with angina should be taking long-term aspirin as it reduces non fatal MI and vascular events but not vascular deaths, with a small increase in bleeding risk.3 NICE guidance points out that it cannot discern those patients likely to benefit most so suggests that comorbidity and bleeding risk be assessed before offering long-term aspirin.
Patients with normal coronary arteries at angiography but anginal pain are classified as Syndrome X and there is no evidence for the use of aspirin in this group.
In secondary care bleeding risk can be assessed using CRUSADE, a bleeding risk score incorporating haematocrit, eGFR, sex, pulse rate, heart failure, diabetes, other vascular disease and systolic blood pressure.14
How do GPs estimate bleeding risk for their patients? In general anticoagulants and antiplatelets (the term oral anticoagulant is being more usually reserved for warfarin nowadays) are more and more likely to be suggested by secondary care. GPs may be familiar with using HASBLED to consider bleeding risk in patients with AF and considering warfarin use. The principles of bleeding risk are partly embedded within HASBLED and the physiology of bleeding applies to any scenario so risk factors for bleeding are:
• Kidney disease including CKD
• Liver disease including alcoholism
• Previous bleeding problem or tendency including recent trauma or surgery
• Age over 65 years
• Memory problems
• Difficulty adhering to therapy
• Tendency to falls
• Active peptic ulcer disease
• Co-medications especially steroids, NSAIDs, other anticoagulants.
Patient decision aids19 using Cate models (smiley faces) are useful for GPs and patients alike when discussing medication risks but there is a need for more patient information about the secondary care aspects of therapy benefit and harms. The drug which doesn’t work is also the one that the patient isn’t taking. A recent article suggests that only 66% of patients are adherent to their secondary prevention CHD medications.20 In an ageing population the role of the GP in supporting patients in therapeutic decision making is central and difficult.
Aspirin is still a mainstay of therapy, clopidogrel is more long-term in all MI use and other platelet ADP inhibitors are finding their place. Low clopidogrel responders exist due to variance in liver activation (it is a pro-drug) and there are also high responders who may be at risk of increased bleeding complications. Peptic ulcer disease is more common with aspirin because prostaglandin E2 formation is inhibited via COX and it is ulcer protective. 10-15% of patients will die of their MI within 30 days. PPCI post STEMI improves mortality by 2% compared to fibrinolysis20 and reduces bleeding complications. PPCI in NSTEMI reduces the incidence of angina. There is no mortality benefit of PPCI in angina therapy at present. Identifying
the patients who would benefit from therapy by PPCI and from various anticoagulants and those who would have complications is very difficult.
Bleeding complications are subdivided into major, minor and trivial. In the CURE study13 major bleeding was defined as bleeding that was disabling, intraocular or needing over two units of blood. Life threatening was bleeding that was fatal or reduced haemoglobin by over 5g/dl, or caused hypotension requiring intervention surgically or with inotropes, or intracranial haemorrhage, or needing over 4u of blood transfusion. Minor bleeding was bleeding that caused the medication to be stopped for a time. Most GPs suspect their patients would define all these bleeding complications in practice as "major". The British Society for Haematology22 recommend stabilising the patient, checking liver and kidney function, PTT, aPTT, thrombin time, fibrinogen concentration and therapy may be IV fluids, red cell transfusion, endoscopy, platelet transfusion, coagulation factor transfusion, protamine sulphate or vitamin K depending on the contributing anticoagulant.
Further articles in this series will be published in GM Journal.
Conflict of interest: none declared
1. Deaths registered in England and Wales 2012. http://www.ons.gov.uk.
2. Coronary heart disease statistics. A compendium of health statistics 2012 edition British Heart Foundation Health Promotion Research Group Department of Public Health, University of Oxford. http://www.bhf.org.uk/
3. Management of stable angina cg July 2011. www.nice.org.uk
4. Treatment of Heart attack National Guidance Final Report of the National Infarct Angioplasty Project Dept of Health. 2008. http://www.bcis.org.uk/resources/documents/NIAP%20Final%20Report.pdf
5. Myocardial infarction with ST-segment elevation. cg 167. July 2013. http://www.nice.org.uk
6. Antman E.M, Cohen M, Bernink PJLM et al. The TIMI Risk Score for Unstable Angina/Non–ST Elevation MIA Method for Prognostication and Therapeutic Decision Making. JAMA. 2000; 284 (7): 835–42
7. Unstable angina and NSTEMI. The early management of unstable angina and non-ST-segment-elevation myocardial infarction. March 2010. cg 94.www.nice.org.uk
8. Third Universal Definition of MI by Thygesen K et al of the Global MI Taskforce. European Heart Journal 2012; 33: 2551–67
9. ESC guidelines for the management of acute myocardial infarction in patients presenting with ST segment elevation European Heart Journal 2012; 33: 2569–619
10. Prasugrel for the treatment of acute coronary syndromes with percutaneous coronary intervention Oct 2009.TA 182.www.nice.org.uk
11. Newsome LT, Kutcher MA, Royster RL. Coronary artery stents: Part I. Evolution of percutaneous coronary intervention. Anesth Analg 2008; 107(2): 552–69
12. Fox KA, Poole-Wilson PA, Clayton TC. et al. Interventional versus conservative treatment for patients with unstable angina or non-ST-elevation myocardial infarction: the British Heart Foundation RITA 3 randomised trial. The Lancet 2002; 360 (9335): 743–51
13. The CURE investigators. Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation. N Eng J Med 2001; 345: 494–502
14. Hamm C, Bassand J-P, Agewall S et al. The Task Force. ESC guidelines for the management of acute coronary syndromes in patients presenting without ST-segment elevation. European Heart Journal 2011; 32: 2999-3034. http://assets.escardio.org
15. Aspirin in coronary heart disease. The Coronary Drug Project Research Group. Circulation 1980; 62(6 Pt 2): 59–62
16. Post myocardial infarction. Secondary prevention in primary and secondary care for patients following a myocardial infarction. June 2013. Partial update of NICE cg48. www.nice.org.uk. draft consultation at Aug 2013.
17. Peters RJ, Mehta SR, Fox KA et al. Effects of aspirin dose when used alone or in combination with clopidogrel in patients with acute coronary syndromes: observations from the Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) study. Circulation 2003; 108: 1682–87
18. Hermosillo AJ, Spinler S. Asprin,Clopidogrel and Warfarin: Is the combination appropriate and effective or inappropriate and too dangerous? The Annals of Pharmacotherapy May 13 2008; 42(6):790-805
19 Acute coronary syndrome (ACS) patient decision aid: aspirin plus clopidogrel versus aspirin alone. 2007. accessed at http://www.npc.nhs.uk
20. Wald DS,Lewis R. Medication adherence in cardiovascular disease: how to address one of the challenges of preventive medicine. Prim Care Cardiovasc J. 2013; 6: 60–62
21. Wijeysundera HC, Ko DT. Does Percutaneous Coronary Intervention Reduce Mortality in Patients With Stable Chronic Angina. Are We Talking About Apples and Oranges? Circulation. Cardiovascular Quality and Outcomes 2009; 2: 123–26
22 Makri M, Van Veen J J, Tait C R. et al Guideline on the management of bleeding in patients on antithrombotic agents. British Journal of Haematology 2012. 160 (1): 35-46 accessed on line at British Committee for Standards in Haematology http://www.bcshguidelines.com/