The areas covered by this guideline are:

All recommendations are for primary healthcare professionals and apply to adult patients attending general practice with chronic stable angina. The development group assumes that healthcare professionals will use general medical knowledge and clinical judgment in applying the general practice principles and specific recommendations of this document to the management of individual patients. Recommendations may not be appropriate for use in all circumstances. Decisions to adopt any particular recommendation must be made by the practitioner in the light of available resources and circumstances presented by and preferences of individual patients.

Initial assessment

Comment: This guideline is to guide the management of patients with stable angina. It assumes that patients will have had one or more consultations with a GP as a result of which the GP thinks that it is likely that they have angina due to coronary artery disease and not aortic stenosis or hypertrophic obstructive cardiomyopathy. This assessment will be based on a clinical history and relevant examination and requires that the following should, at some point, be known: precipitants of attacks of angina, history of smoking, occupation, amount of exercise taken, drug history, weight, and blood pressure. There are clearly other issues that GPs will want to consider and record but these are the issues covered by this guideline.

The guideline development group thought it important to state, as a general principle, that it is important to ensure clear communication and understanding at all stages in dealings between healthcare professionals and patients.

Age limits

Comment: The group thought that a blanket chronological age limit for investigation or referral was not appropriate. Functional status was thought to be more appropriate. Therefore, a fit, well 83 year old should not be barred from full investigation and referral just because of his or her age.

Precipitating factors

Recommendation

The group thought it important to identify with patients those factors that brought on episodes of angina and, if appropriate, to discuss their management with the patient.

Investigation of angina

Recommendations

Patients being investigated for angina should have:

There was disagreement among both the group and the reviewers about whether or not thyroid function tests should be performed on every patient or only on patients in whom there is clinical suspicion of thyroid disease.

Research question

What is the role of routine biochemical testing in the investigation of patients with stable angina?

Resting 12 lead electrocardiogram (ECG)

Recommendations

Statement: a normal resting 12 lead ECG does not exclude coronary artery disease (II)—Norell et al reviewed the case notes of 250 patients who had presented to a "chest pain clinic" with recent onset chest pain. Of 109 patients with normal ECGs, 48% had non-cardiac pain, 13% had an undetermined diagnosis, and 39% had cardiac pain (half of these with unstable angina).(12) Forty one of these 109 patients had coronary arteriography, and 37 showed significant coronary artery disease. In a case-control study Mukerji et al reviewed the records of patients who had been investigated for chest pain; 15% of those with coronary artery disease had a normal ECG.(13)

Statement: an ECG that is abnormal in any way supports the clinical diagnosis of coronary artery disease (II)—Mirvis et al studied angiographically determined coronary anatomy in a cohort of patients.(14) They compared patients with normal and those with abnormal 12 lead ECGs and showed that ST-T abnormalities correlated with left ventricular contraction abnormalities, male sex, older age, left anterior descending artery stenosis, higher end systolic diameter, and higher blood pressure.

Gregoire and Theroux followed a series of patients who presented with unstable angina and estimated the ability of the 12 lead ECG to predict the presence of coronary artery stenosis (shown at subsequent angiography).(15) A transient abnormality in the 12 lead ECG during pain had a 35% sensitivity and a 68% specificity. The corresponding figures if abnormalities occurred at any time were 65% and 63%.

France et al looked at the ECG records of a cohort of 122 patients who underwent angiography and found that QRS notching or slurring had 62.2% sensitivity and 93.8% specificity for the detection of ischaemic heart disease.(16) Berger et al studied the ECGs of a cohort of 278 patients with chest pain who attended the emergency department and found that an abnormal ECG was 98% sensitive for coronary artery disease; ST-T abnormalities were 86% sensitive and 72.5% specific.(17)

In a further cohort study, Miranda et al looked at patients with resting ST depression, not due to left ventricular hypertrophy, conduction defects, or drugs, who were assessed by exercise test or arteriography.(18) They found that patients with ST segment depression or a previous myocardial infarction had a higher prevalence of severe coronary artery disease than those without.

Statement: an abnormal ECG identifies a population with a poorer prognosis (II)—Madsen et al followed a cohort of patients with suspected myocardial infarction for one year and found that patients with ST segment depression or elevation, T wave inversion, or interventricular conduction delay had a higher incidence of events at one year.(19)Aronow looked at a cohort of elderly, chronically sick patients and found that the amount of ST segment depression correlated with an increased incidence of new cardiac events: >1 mm depression had 3.1 times greater risk; >0.5 mm and <1 mm had 1.9 times greater risk.(20) The study by Miranda et al was also predictive.(18)

Exercise testing

Within this guideline, exercise testing is regarded as a prognostic rather than a diagnostic test. The aim of referral for exercise testing is to group prognostically patients with clinically certain angina and thereby identify those that would benefit from further (invasive) investigation.

An exercise ECG involves a patient having an ECG recorded while he or she is exercising, usually walking on a treadmill. The test should be constantly supervised by trained staff. The speed and slope of the treadmill are both increased throughout the test according to a standard protocol. At its maximum, for most people the test corresponds to walking briskly up a 20% slope; some may need to run.

Referral for exercise testing will be influenced by whether or not open access referral facilities (for instance, direct referral by a GP) are available locally. If they are not then referral for exercise testing will have to be via a cardiologist (see section on referral to a cardiologist).

The recommendations apply only to an initial exercise test. There is no evidence to guide on when (or if) to repeat an exercise test that had previously indicated "low risk".

Recommendations

Statement: an exercise test is a low risk investigation (II)—The safety of the test has been reported in various populations, some of whom were fit adults taking the test as part of fitness assessment. In a population known to have ischaemic heart disease the risks of serious complications (ventricular fibrillation, myocardial infarction) are of the order of 2-4 per 1000 tests; deaths occurred at a rate of 1-5 per 10 000 tests.(21)

Statement: exercise testing is valuable in the assessment of patients with coronary artery disease to establish a prognosis and can provide information in addition to that from invasive testing (II)—Weiner et al analysed 30 variables as predictors of mortality in a cohort of 4083 patients and found that while left ventricular contraction pattern and the number of diseased coronary arteries were most important, in patients with three vessel disease and preserved left ventricular function four year survival was 100% in those who had achieved stage V on an exercise test and 53% in those who achieved only stage 1/2.(22)

Bonow et al studied 106 men and 11 women with no angiographically demonstrable left main stem coronary artery disease and no symptoms or mild stable angina on treatment.(23) They were investigated with exercise testing and gated blood pool scan. All the deaths in four years (11%) were in a group with four factors: ST segment depression >1 mm, a decreased ejection fraction during exercise, exercise tolerance <120 watts, and three vessel disease. The presence of three vessel disease alone equated with a good prognosis.

Sato et al compared three cohorts of Japanese men who had typical angina, atypical angina, or non-angina chest pain.(24) They found that the exercise test was predictive of cardiac events; even in those patients with atypical angina and non-angina chest pain. Mark et al followed a cohort of 2842 patients (70% males) for up to 10 years after an angiogram and exercise test.(25) The results were analysed with the treadmill score (ST segment deviation, exercise time, and angina index). This score allowed identification of high, moderate, and low risk groups. Five year survival in the groups was 72%, 91%, and 97%, respectively; event free survival was 63%, 86%, and 93%, respectively.

Statement: in interpreting results of an exercise test it is important to take into account not just the changes in ST segment but other changes such as duration of exercise, presence of pain, change in blood pressure, and change in heart rate (II)—Detry et al studied the computer-assisted exercise test, evaluating ST segment changes, maximum heart rate, blood pressure, workload, and onset of angina in 387 men.(26) They found that ST segment changes alone were barely better than clinical history. With all five variables, however, there was 83% correct classification. They also concluded that with a typical history an exercise test is not needed for diagnosis but is for prognosis. Detrano et al carried out a meta-analysis of the diagnostic accuracy of exercise tests.(27) They studied 147 reports and concluded that 1 mm ST segment depression had a mean sensitivity for left main stem and triple vessel disease of 86%.

Mark et al looked at cases of positive exercise tests with no pain. The 1698 patients were split into three groups: no ST deviation, painless ST segment deviation, and painful ST segment deviation.(28) The five year survival rate was similar in patients with painless and no ST segment deviation and better than those with pain and ST segment deviation. Richardson et al looked at angiography and exercise tolerance testing in 1138 patients to determine the predictive power of 10 exercise tolerance test parameters.(29) Of the 10, exercise ECG maximum rate pressure product and exercise chest pain gave predictive information.

Pratt et al performed angiography on 200 women with chest pain compatible with angina and >1 mm ST segment depression on exercise testing.(30) Four variables were associated with an increased likelihood of coronary artery disease: absence of mitral valve prolapse; exercise duration <5 minutes; inability to reach target heart rate; and time to ST normalisation >6 minutes. False positive results were associated with the ability to reach stage III of Bruce protocol and <4 minutes to ST normalisation.

Weiner et al followed up 2982 patients.(31) They categorised them into four groups: those with ST segment depression and no angina; those with angina and no ST segment depression; those with ST segment depression and angina; and those with no ST segment depression and no angina. The absence of pain did not negate other positive results; the seven year survival was similar whether ischaemia was silent or not.

Statement: the diagnostic usefulness of an exercise test is low in patients with a low pre-test probability of coronary artery disease before the test (II)—Weiner et al, in the CASS study, looked at 1465 men and 580 women.(32) The prevalence of coronary disease before the test was 7-87%. The rate of false positive results was 12% in men and 53% in women. They concluded that the value of the test in identifying patients with coronary heart disease is limited in a heterogeneous population. Diamond and Forrester looked for the features that influence the likelihood of coronary heart disease before the test and concluded that exercise testing is most useful in the mid-range of uncertainty, when coronary disease is neither very unlikely nor very likely.(33)

Statement: an open access exercise testing service can be used appropriately by GPs (III)—Sulke et al looked at 110 GP referrals for exercise tests and compared patient characteristics and test results before and after guidelines.(34) After the introduction of the guideline 67% of the referrals were judged to be appropriate compared with 46% before; more referrals were for prognosis and fewer for diagnosis or with non-cardiac pain. There were also fewer patients with low probability of coronary heart disease and more with moderate risk.

Statement: when an exercise test is performed to identify whether a patient is in a group that would benefit from prognostic investigation and treatment, it should be performed with the patient taking his or her normal medication (II)—Lim et al studied 84 patients with typical angina or a history of definite myocardial infarction and mild symptoms.(35) An exercise test, radionucloetide ventriculography, and angiography were all performed with the patients taking and not taking their medication. The authors concluded that the exercise test should be performed with patients taking their medication to optimise identification of prognostically important disease and avoid angiography in those whose symptoms are well controlled on treatment.

Statement: doctors requesting an exercise test on a woman should record her oestrogen status in terms of whether or not she is menopausal or taking oestrogen replacement therapy (II)—Morise et al studied 326 men and 234 women and looked at the results of the exercise test and subsequent coronary angiogram.(36) When they assessed the result of the exercise tests along with other parameters, inclusion of women’s oestrogen status, defined by menopause or oral oestrogens, resulted in the diagnostic accuracy of exercise tests becoming similar in men and women.

Statement: if an exercise test is performed on a patient with diabetes this should be clearly stated on the request form as this may influence performance and interpretation of the test (II)—Ranjadayalan et al compared results of exercise tests in 32 patients with diabetes and 36 without and found that there was no correlation between having diabetes and time to ST segment depression or exercise capacity but there was prolongation of the angina perceptual threshold in patients with diabetes.(37)

Research question

What are the indications for repeating an exercise test in a patient with stable angina?

Risk factor modification

Assessing risk factors

Recommendations

Statement: a patient’s absolute risk of subsequent cardiovascular events or death is related to his or her underlying risk factors, both modifiable and non-modifiable (I)—The NHS Centre for Reviews and Dissemination publication on cholesterol and coronary heart disease concludes that while blood cholesterol is an important risk factor for coronary heart disease it should be considered in the context of other modifiable risk factors (smoking, raised blood pressure).(38) Lowering cholesterol concentration (which is effective in reducing mortality and morbidity from coronary heart disease) is one of several methods of reducing the risk of cardiovascular events. They suggest that the cost effectiveness of other strategies (some antihypertensives, aspirin, and b blockers) is greater than for statins. An overall framework within which to consider modifiable risk factors is provided by the New Zealand hypertension and lipid guidelines.(39)

Cholesterol

Recommendations

Statement: most (but not all) patients with stable angina are at high risk of subsequent cardiovascular events or death (I)

Statement: patients at high risk of subsequent cardiovascular events or death benefit from having raised serum cholesterol concentrations lowered (I)

Statement: lowering cholesterol concentration with statins is effective at reducing subsequent mortality and morbidity from coronary heart disease (I)

Statement: in patients with ischaemic heart disease there is no good evidence to support the effectiveness of a range of "diet alone" interventions (low fat diets; garlic, oats, and soy protein) in lowering total mortality or fatal or non-fatal coronary events (I)

Statement: in patients at high risk of ischaemic heart disease (>3% annual death rate for coronary heart disease) fibrates are, on balance, beneficial in reducing CHD event rates for coronary heart disease; not all patients with stable angina will be at such risk (I)

Davey Smith et al showed that the benefit of lowering cholesterol concentration was limited to those patients who were at high risk of cardiovascular disease.(40) While it is, from the paper, difficult to say whether patients with stable angina would always fall into the "high risk" group, the guideline development group thought that it was reasonable to say that all patients with angina should have their serum cholesterol concentration measured.

The NHS Centre for Reviews and Dissemination publication on cholesterol and coronary heart disease concludes that, while blood cholesterol concentration is an important risk factor for coronary heart disease, it should be considered in the context of other modifiable risk factors (smoking, raised blood pressure).(38) Lowering cholesterol concentration (which is effective at reducing coronary heart disease mortality and morbidity) is one of several methods of reducing the risk of cardiovascular events. From 22 published randomised controlled trials of lipid lowering treatment the authors calculated a relative risk for statins versus placebo of 0.79 (95% confidence interval 0.73 to 0.86) for total mortality, 0.74 (0.66 to 0.83) for coronary heart disease mortality, and 0.70 (0.61 to 0.80) for non-fatal myocardial infarction. Most of the data on statins come from four studies.(41) (42) (43) (44)

Blood pressure

Recommendations

Statement: for patients with high blood pressure, antihypertensive medication reduces the risk of cardiovascular and all cause mortality (I)—(Comment: the details of levels of intervention, target levels of control, and appropriate drugs vary between published guidelines.) In patients with angina who are at raised risk of cardiovascular events and more likely to be elderly, any reduction in blood pressure will lower their risk of subsequent cardiovascular events.(45) Although they focused on older people with hypertension, the review of Mulrow et al included trials within which a proportion of the patients had known ischaemic heart disease.(45) As such, the guideline development group has assumed that patients with stable angina and hypertension are likely to benefit at least as much as those in the trials within the review.

Mulrow et al also conducted a meta-analysis to examine the long term effects of antihypertensive drug treatment on morbidity and mortality in patients over 60 years.(45) They identified 15 randomised controlled trials (with 21 908 patients) that lasted at least a year and assessed antihypertensive drug treatment in elderly people (at least 60 years old) with hypertension and provided data on morbidity and mortality. The average prevalence of cardiovascular risk factors, cardiovascular disease, and competing comorbid diseases was lower among trial participants than the general population of elderly people with hypertension. Most were 60 to 80 years old. Most trials were conducted in Western, industrialised countries and evaluated diuretics and b blockers.

Event rates per 1000 participants over about five years indicated that antihypertensive drug treatment was beneficial. Cardiovascular morbidity and mortality was reduced from 177 to 126 events (95% confidence interval of difference 31 to 73). Cardiovascular mortality was reduced from 69 to 50 deaths (9 to 31). Total mortality was reduced from 129 to 111 deaths (4 to 28). The data from the three trials restricted to people with isolated systolic hypertension indicated a significant benefit: cardiovascular morbidity and mortality over about five years was reduced from 157 to 104 events per 1000 participants (12 to 89). The numbers of participants who dropped out of trials because of adverse drug effects were often not reported. The four trials that did report the data showed a wide variation in drop out rates, ranging from no significant differences between treatment and control groups to as many as one out of four dropping out because of side effects of treatment.

They concluded that treating healthy older people with hypertension is highly effective. Benefits of treatment with low dose diuretics or b blockers are clear for people in their 60s and 70s with either diastolic or systolic hypertension. Differential treatment effects based on patient risk factors, pre-existing cardiovascular disease, and competing comorbidities could not be established from the published trial data.

Smoking and smoking cessation

Recommendations

Comment: Most of the papers identified on smoking cessation dealt with coronary heart disease or risk factors for coronary heart disease rather than stable angina specifically.

Statement: smokers are more likely than non-smokers to suffer from angina and stopping smoking probably does not alter anginal symptoms (II)—A group of studies suggests that the association between smoking and angina is not as strong as the association with death from ischaemic heart disease.(46) (47) (48) (49) Two of these are from the Framingham study. Three other studies found that for smokers the relative risk for angina was 2.6(50) or 3.0(51) and for coronary artery stenosis was 2.8.(52) However, stopping smoking may not affect the symptoms of angina.

Statement: stopping smoking probably lowers mortality in patients with ischaemic heart disease (II)—The evidence from cohort studies suggests benefit from stopping smoking. In a group of 4165 patients with angiographically documented coronary heart disease followed over five years, Vlietstra et al showed an adjusted mortality of 22% in smokers and 15% in those who had stopped smoking.(53) Two papers from the CASS study found over 30% reduction in mortality in people over 70 years who gave up smoking and over 54% reduction in younger groups.(54) (55)

Neaton and Wentworth showed that smokers of up to 25 cigarettes a day had a risk ratio of coronary heart disease events of 2.1 and smokers of more than 45 cigarettes a day had a risk ratio of 3.4, both compared with non-smokers.(56) These figures were for over a five year period. Doll et al followed a cohort of 34 000 British doctors over 40 years and found that non-smokers had 30% less ischaemic heart and other heart disease.(57)

Statement: brief advice from a health professional and nicotine replacement therapy can help patients to stop smoking (I)—The tobacco addiction module of the Cochrane Database of Systematic Reviews draws together evidence of the effectiveness of smoking cessation interventions.(58) These reviews and supporting information have been summarised in the NHS Centre for Reviews and Dissemination publication Effectiveness Matters "Smoking cessation: what the health service can do," which reviews the effectiveness of smoking cessation interventions.(59) Both sources conclude that there is good evidence of effectiveness for brief advice from a health professional; nicotine replacement therapy with advice; and advice and support to pregnant women. There is insufficient evidence of effectiveness for antidepressants/anxiolytics; aversive conditioning; acupuncture; hypnosis; mecamylamine; and self help materials, booklets, pamphlets, and manuals. There was some supportive evidence for the effectiveness of clonidine but doubts about its usefulness. The publication examines weight gain after smoking cessation and states that "smokers should be informed that weight gain is common but that the associated health risks are far outweighed by the benefits of stopping smoking. Nicotine replacement therapy can delay weight gain until smokers feel ready to follow a weight control strategy."

Statement: transdermal nicotine is safe to use in patients with ischaemic heart disease (I)—The working group for the study of transdermal nicotine in patients with coronary artery disease examined the safety of transdermal nicotine for smoking cessation in patients with coronary artery disease.(60) In a five week, double blind, randomised, placebo controlled, multicentre study, 156 patients with coronary artery disease who smoked at least one pack of cigarettes daily were randomised to receive either transdermal nicotine (14 mg/day) or transdermal placebo. After one week, patients who had smoked more than seven cigarettes were able to have their dose of blinded study medication increased to 21 mg/day of transdermal nicotine or corresponding placebo. All participants attended weekly group counselling sessions. Eight of 79 patients randomised to receive placebo and three of 77 patients randomised to receive transdermal nicotine withdrew during the trial because of adverse effects (P=0.13), most of which were cardiovascular. Transdermal nicotine did not affect frequency of angina attacks, overall cardiac symptom status, nocturnal events, arrhythmias, or episodes of ischaemic ST segment depression. Smoking cessation was achieved by 36% and 22% of patients receiving transdermal nicotine and placebo, respectively (P<0.05).

Joseph et al randomly allocated 584 patients with a diagnosis of cardiovascular disease to a 10 week course of transdermal nicotine or placebo as an aid to smoking cessation.(61) Participants were monitored for 14 weeks for the primary end points (death, myocardial infarction, cardiac arrest, admission to hospital for arrhythmia, congestive heart failure, or worsening angina symptoms), secondary end points (admission to hospital for other reasons and cardiac related outpatient appointments), side effects, and smoking behaviour. There were no significant differences between groups in primary or secondary end points or adverse effects. The transdermal nicotine group were more likely to have abstained from smoking by the 14 week point (21% v 9% in the placebo group; P=0.001), however this difference was not sustained after the trial (abstinence at 24 weeks; 14% v 11%; P=0.67)

Comment: The BNF recommends caution in the use of nicotine products in patients with cardiovascular disease.

Exercise

Recommendations

There is no consistent evidence that exercise influences the progress of stable angina. The evidence that was identified for the role of exercise in the prevention of coronary heart disease is conflicting. Some trials have suggested some improvement in myocardial perfusion,(62) (63) coronary blood flow, peak exercise, and arteriographic changes.(64) Five trials showed no benefit.(65) (66) (67) (68) (69) These trials were all of different design and used different exercise regimes and durations; none had been designed solely to study patients with stable angina.

The evidence from cohort and case-control studies is also conflicting. One study showed that people defined as "highly active" lived on average 2.1 years longer.(70) Another showed that both sedentary work and sedentary occupation was associated with more ischaemic heart disease after correction for the risk factors.(71) However, a cohort study of over 9000 middle aged men found that after correction for risk factors for coronary heart disease there was no association between physical activity and myocardial infarction.(72)

Lewin et al randomised 77 patients with angina to either an eight week outpatient rehabilitation programme or a waiting list control group.(73) The programme included supervised exercise sessions, stress management, and education over two full mornings a week for the eight week period. At the end of the eight week period, the treatment group showed significantly greater improvements than the waiting list group in terms of episodes and severity of angina, use of glyceryl trinitrate, time on treadmill, and time to 1 mm ST depression but not in terms of duration of angina, resting heart rate, or maximum ST depression.

Cupples and McKnight randomised 688 patients who had had angina for at least six months to receive education or usual care.(74) The intervention group were given education regarding cardiovascular risk factors at four monthly intervals over a two year period. After two years, more patients in the intervention (44%) than in the control group (24%) reported taking daily physical exercise. The intervention group also reported eating a healthier diet than the control group and less restriction by angina in everyday activity. There was no difference between groups in smoking habit, blood pressure, cholesterol concentration, or body mass index (BMI).

Weight reduction and dietary management

Recommendations

While the review by Brand et al shows benefit from weight reduction in overweight patients with hypertension our search did not identify any direct evidence of benefit of weight reduction in normotensive overweight patients with stable angina.(75)

Statement: weight reducing diets in overweight people with hypertension can result in modest weight loss and may decrease dosage requirements of people taking antihypertensive medications (I)—The systematic review by Brand et al evaluated whether weight loss diets are more effective than regular diets or other antihypertensive treatments in controlling blood pressure and preventing morbidity and mortality in hypertensive adults. Six trials involving 361 participants assessed a weight reducing diet versus a normal diet. The data suggested that weight loss in the range of 4% to 8% of body weight was associated with a decrease in blood pressure in the range of 3 mm Hg systolic and diastolic. Three trials involving 363 participants assessed a weight reducing diet versus treatment with antihypertensive medications. These suggested that a stepped care approach with antihypertensive medications produced greater decreases in blood pressure (in the range of 6/5 mm Hg systolic/diastolic) than did a weight loss diet. Trials that allowed adjustment of participants’ antihypertensive regimens suggested that patients required less intensive antihypertensive drug treatment if they followed a weight reducing diet. Data were insufficient to determine the relative efficacy of weight reduction versus changes in sodium or potassium intake or exercise.

Weight reducing diets in overweight people with hypertension can result in a modest weight loss in the range of 3-9% of body weight and are probably associated with modest blood pressure decreases of roughly 3 mm Hg systolic and diastolic. Weight reducing diets may decrease dosage requirements of people taking antihypertensive medications.

Statement: patients who have survived a myocardial infarction and start a "Mediterranean diet" or increase their intake of fatty fish have a lower rate of subsequent cardiovascular events (I)—Two trials have suggested that specific dietary strategies may benefit patients at high risk of subsequent cardiovascular events. The participants in the trials had all survived a myocardial infarction; it is unclear how directly these results translate to a population with stable angina, only some of whom will have survived a myocardial infarction.

de Lorgeril et al compared Mediterranean diet (<35% energy as fat; <10% energy from saturated fat; <4% energy as linoleic acid; >0.6% energy from a linolenic acid) with usual care in 605 people who had survived a first acute myocardial infarction.(76) A dietician assisted patients in adapting their usual diet to the Mediterranean diet, taking into account patient preferences. Participants were followed for a mean of 27 months. Of the primary end points (cardiac death and non-fatal acute myocardial infarction) there were 33 events in the control group and eight in the Mediterranean diet group. Taking secondary end points (mainly unstable angina) into account, there were 59 events in the control group and 14 in the study group; a relative risk for the diet group of 0.24 (0.13 to 0.44).

Burr et al randomised 2033 men who had recovered from a myocardial infarction to receive or not receive advice on each of three dietary factors: a reduction in fat intake and an increase in the ratio of polyunsaturated to saturated fat; an increase in fatty fish intake (at least two portions a week); and an increase in cereal fibre intake.(77) The advice on fat only produced a small reduction (3-4%) in serum cholesterol concentration and was not associated with any difference in mortality. The men advised to eat fatty fish had a 29% reduction in two year all cause mortality compared with those not so advised (this was almost entirely due to a lower rate of deaths related to ischaemic heart disease). Men given advice on fibre did not have a significantly different mortality than other men.

Research question

What is the cost effectiveness of a Mediterranean style diet and a diet with an increased intake of fatty fish in patients with stable angina?

Occupation

Having angina may affect a patient’s occupation. For some occupations this may necessitate a switch to lighter duties, for some the patient may have to stop work; whether or not this is a permanent change will be affected by their response to treatment. Vocational drivers’ continued fitness to work is governed by clear regulations. Special rules also apply to certain other occupations—for example, merchant seamen, airline pilots. For details the relevant authorities should be consulted.

Driving

The Road Traffic Acts 1972 require immediate notification by an applicant or licence holder to the Driver and Vehicle Licensing Agency at Swansea on diagnosis of any disability that is likely to affect safe driving either at the time of driving or in the future, except in the case of disabilities, such as fractures, which will be completely cured within three months. Vocational drivers’ continued fitness to work is governed by clear regulations. The medical practitioner’s role is to advise the patient on the basis of the severity of the condition.(78) If the driver’s fitness is so severely affected as to present a considerable hazard to other people, and if the driver fails to notify the agency, there may be grounds for the doctor to consider whether he or she should notify the agency directly.