Article Text

Download PDFPDF

The head says yes but the heart says no: what is non-cardiac chest pain and how is it managed?
  1. J B Chambers1,
  2. E M Marks1,2,
  3. M S Hunter2
  1. 1Cardiothoracic Centre, Guy's and St Thomas Hospital, London, UK
  2. 2Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, UK
  1. Correspondence to Professor John Chambers, Cardiology Administration, Cardiothoracic Centre, Guy's and St. Thomas’ Hospital, F06, East Wing, London SE1 7EH, UK; John.Chambers{at}

Statistics from

Learning objectives

  • To recognise the causes and natural history of non-cardiac chest pain.

  • To recognise features in the history and examination which differentiate cardiac pain from non-cardiac chest pain.

  • To explore and treat the psychological factors initiating and maintaining non-cardiac chest pain.

CURRICULUM TOPIC: Chronic ischaemic heart disease


Chest pain is common and usually non-cardiac in origin. The lifetime population prevalence of non-cardiac chest pain (NCCP) is approximately 20–33%1–7 compared with approximately 6–7% for angina.3 ,8 The incidence of NCCP depends on the clinical setting. It is 70–80% for patients presenting to a general practitioner or a rapid access chest pain clinic9–11 and around 50% attending emergency departments.12 ,13 Normal coronary anatomy is found in 40% having diagnostic coronary angiography.14

Despite this, the focus of clinical care is on excluding coronary disease rather than on the positive management of NCCP. As a result, patients with NCCP are often left with chronic symptoms, high levels of psychological distress,15 ,16 high unemployment and heavy use of healthcare resources.17–19 The causation may be complex with an interaction of organic and psychological processes. However, treatment can be effectively delivered at low cost.20 ,21

This article describes the causes, natural history and management of NCCP with an emphasis on the psychological processes which inform our approach to care.

What is NCCP?

Chest pain may be obviously benign, for example after straining a muscle or being hit in the chest by a football. It only becomes of medical concern when the person seeks advice for one or more reasons: severe or recurrent pain; a family history of coronary disease; health anxiety as a result of personality or induced by societal concerns, for example, healthcare advertisements. The general practitioner may be confident to reassure the person without tests. However, someone referred to a cardiac outpatient or rapid access chest pain clinic, even with a normal somatic sensation, or who has cardiac tests becomes a patient with a diagnosis. NCCP is then defined by the chest pain characteristics or by the absence of ischaemia on functional non-invasive testing or the finding of normal or near-normal anatomy at coronary angiography.

NCCP is therefore defined negatively and is not a unitary diagnosis. It is also to a degree an arbitrary diagnosis based on the willingness of a physician to investigate. This in turn depends on a variety of factors including the presence of coronary risk factors, fear of litigation, clinical inexperience and hospital protocols. The penalties for missing cardiac pain are high and this may help explain the tendency of physicians to overestimate risk.22 The patient characteristics therefore partly reflect the concerns of the physician, the patient and, to a degree, society. However some features are common (box 1). The average age is in the late 40s and men and women are equally represented.3 ,6 ,23 About a quarter are said to report chest pain ‘typical’ of a cardiac origin and about a third have cardiac risk factors.23 Anxiety and depression are found with similar frequency as in patients with proven coronary disease but tend not to resolve20 ,24 and are associated with more disordered illness beliefs.16

Box 1

Vignette of a characteristic patient with non-cardiac chest pain referred to a rapid access chest pain clinic

Man or woman in the late 40s

Mild heavy feeling across the chest all the time

Additional ‘jabs’ of pain at a variable site, most commonly left submammary. May radiate to left shoulder or arm

Pain can come at any time including rest or minimal exertion. May last hours after exercise

Palpitation with the pain or at other times

Irritable bowel (abdominal pain and bloating, flatulence, variable stool consistency, urgency)

Headaches, fatigue, ‘muzzy’ head, difficulty concentrating

Cannot take a satisfying breath

Sighing thoracic respiratory pattern, frequent throat clearing

Breath-holding time reduced to <20 s.

Pain reproduced by chest wall palpation or by breath-holding

Stresses in life

Thinks he or she may have a heart attack

Frequently off work

Very difficult to manage are those patients who have coronary disease and NCCP. ‘Atypical’ symptoms may be investigated by diagnostic coronary angiography revealing incidental coronary stenoses sometimes leading to inappropriate invasive intervention. However approximately 10% of patients appropriately revascularised by coronary artery bypass grafting (CABG) may return with NCCP.25 ,26 The chest pain characteristics are usually different from before revascularisation with pain often occurring at rest or on minimal exertion. Breathlessness is often more prominent than chest pain. It is important to consider and treat NCCP early after excluding residual cardiac ischaemia rather than performing more complex tests. Perceived diagnostic uncertainty entrenches the idea that revascularisation has been ineffective and that there is a sinister underlying problem not amenable to medical science.

Alerts that a patient should be taken out of the cardiac management protocol are: residual pain after negative investigations at the RACP; multiple emergency room attendances with troponin-negative chest pain; or a normal or near-normal coronary angiogram. We believe that these patients should be managed as outpatients with a helpline to a multidisciplinary team consisting of a cardiologist, cardiac nurse and clinical psychologist.

What causes NCCP?

Descriptions recognisable to a modern cardiologist were recorded in the mid 19th century in civilians and in soldiers in the US Civil War.27 Similar descriptions appeared in World War I28 ,29 and the condition was known as ‘da Costa's syndrome’, ‘circulatory neurasthenia’, ‘irritable heart’, ‘soldier's heart’ or ‘effort syndrome’. Numerous organic theories of causation were considered,30 but after extensive research and clinical experience in the interwar years Paul Wood in his Goulstonian Lectures in 1941 summarised the consensus that it was emotional in origin.31 In the third edition of his textbook he states: “It should be understood that there is no essential difference between ‘effort syndrome’ and ‘cardiac neurosis’, they are merely clothed differently, the former in battle dress and latter in nylon. In civil life the condition accounts for 10 to 15 per cent of all cases referred to cardiovascular clinics…”.

With the introduction of M-mode echocardiography, many patients with NCCP were thought to have mitral prolapse; ‘mitral prolapse syndrome’32 ,33 incorporated many or all of the features previously ascribed to ‘circulatory neurasthenia’. However with the development of 2D echocardiography, followed by the realisation of the saddle-shape of the mitral annulus, it became apparent that mitral prolapse had previously been overdiagnosed by approximately 20-fold. The concept of ‘mitral prolapse syndrome’ faded.34

In the 1970s, the term ‘syndrome X’ was introduced and was defined by the presence of ‘typical’ pain, normal epicardial arteries and evidence of ischaemia on any test including the exercise ECG.35 We now realise that the stress ECG is not a reliable measure of ischaemia and the characterisation of pain as ‘typical’ is too often subjective and loose (see below). Syndrome X was more common in women while NCCP is seen approximately equally in both sexes probably because apparently abnormal ST segment changes are more likely in women. There may be evidence of microvascular dysfunction,36 but stress echocardiography is normal and evidence for subendocardial ischaemia on cardiac magnetic resonance (CMR) is contradictory.37 ,38 It is impossible to summarise a huge literature within this focused review, but recent papers stress the concept of heightened pain perception39–41 which may be affected by psychological factors41 ,42 and which may respond to psychological treatments.23 ,40 ,41 This helps harmonise apparently contradictory models of abnormal physiology and psychology. In our biopsychosocial approach the aim is to integrate these models.

In contradistinction with syndrome X, myocardial ischaemia is excluded before making the diagnosis of NCCP. The main causes of chronic NCCP are thought to be gastro-oesophageal reflux or a motility disorder or a musculoskeletal or psychological problem. The proportion of each varies depending on the specialty of the research study and this may be as a result of selection bias. Most patients, about 80%, are managed solely by their GP.18 Of those referred, the majority are first seen by a cardiologist, and the rest mainly by a gastroenterologist43 with smaller proportions seen by a rheumatologist or psychologist. It is likely that patients seen by a cardiologist with NCCP resemble those referred to a rheumatologist with fibromyalgia and perhaps those seen first by a gastroenterologist, but comparative data do not exist.

Symptoms suggestive of gastro-oesophageal reflux are reported in 50–90% of those patients with NCCP23 referred to a gastroenterologist, but are also common in the normal population. Probably a half with NCCP referred to a gastroenterologist have an oesophageal abnormality as a cause of the NCCP1 ,23 ,44 and 5–10% seen in general practice or the emergency room.13 ,45 Reflux is a more common cause than nutcracker oesophagus or other motility disorders.1 ,23 The pain is described as ‘typical’ in just under a half (range 13–100%) of individual research studies.23 However the link between gastrointestinal function and pain is not always straightforward.1 Some patients experience pain with normal levels of acid reflux while balloon dilatation in the oesophagus causes pain at a lower pressure in patients with NCCP than in normal people.46 This has led to the concept of an ‘irritable oesophagus’ and there is a large literature on abnormal proprioception and central sensitisation which can be induced by many factors including psychological abnormalities.46–48

Between 5% and 30%13 ,23 ,49 are thought to have a musculoskeletal origin for chest pain including cervical spondylosis, or fibromyalgia50 A thoracic respiratory pattern sometimes associated with overt hyperpnoea (‘hyperventilation’) is commonly associated with NCCP. This may occur spontaneously or be precipitated by psychophysiological processes such as panic or stress arousal. It can also occur after events including cardiac surgery, pneumonia or a road traffic accident. We showed51 hypocapnia on treadmill exercise in 50% of patients with normal ECGs compared with 14% with ST segment depression and 3% of controls. Failure of the pCO2 to rise in the first minute of exercise was strongly associated with panic anxiety52 and with pain triggered by exercise or by emotion. Thoracic breathing probably causes pain by muscular tension since there can often be associated tenderness consistent in some cases with a diagnosis of fibromyalgia or costochondritis. We showed a closer association between the respiratory rate than end-tidal pCO251 consistent with a musculoskeletal origin. Whether the discomfort from abnormal breathing is counted as musculoskeletal or psychological may account for some of the variability in the incidence of a presumed musculoskeletal origin. For example, in a general practice population45 a ‘chest wall syndrome’ was thought to be the cause of NCCP in 70% of all cases while a psychological cause was judged to be present in only 14%. Most studies find these frequencies reversed.23

In many patients, physical causes of pain may interact with the psychological processes used to deal with pain or other apparently inexplicable sensations. For example, patients may experience a normal somatic sensation such as minor discomfort from sitting awkwardly and misinterpret this as a threatening symptom. This is exacerbated by unhelpful psychological processes, such as attribution to a cardiac cause, attentional focus and hypervigilance to pain-related symptoms and catastrophisation. Such processes probably result from concern about family history or influences from health advertising or the experience of friends. We found20 an organic cause for NCCP in 36 (47%) of a series referred from a rapid access chest pain clinic but 28 of these also required cognitive behavioural therapy (CBT), 10 at low intensity and 18 at high intensity. This underlines the importance of the biopsychosocial approach linking rather than separating all potential physical and psychological influences on the perception of pain.

The natural history of NCCP

NCCP is seen by the cardiologist as benign because the incidence of myocardial infarction or premature death is close to zero17 ,23 ,44 yet patients remain significantly disabled. About three quarters report residual chest pain,23 and continue seeing a physician. A half remains or become unemployed18 and absenteeism is common with a mean time off work of 22 (range 1–240) days in a year.17 Healthcare use is high.18 ,45 ,53 In the 6 months before coming to our clinic,20 63% of referrals saw a GP ≥three times, 63% saw a cardiologist more than once, 33% saw another hospital doctor and 28% attended the emergency room more than once. In the USA in 1989, it was estimated that bed costs alone were $3500 over the year following a normal coronary angiogram.54

Symptoms continue if an underlying organic non-cardiac source of pain or a psychiatric disorder remains untreated.55 ,56 It is also associated with stress at home or work, with negative life-events,24 health anxiety57 or a number of psychological processes such as catastrophisation, avoidance behaviour, a belief that the heart is the source of the pain18 ,58 ,59 and that symptoms are uncontrollable.24 Chronicity may also be associated with a longer interval before diagnosis58 ,60 and may be more likely in women.61 ,62 Treatment is also less likely to be successful if chest pain has led to a significant change in lifestyle, for example by having a disability allowance, having started to use a wheelchair or being prescribed domiciliary oxygen.

Managing the patient with NCCP

Extending the history

The history needs to be detailed and cover chest pain characteristics, the presence of extracardiac symptoms and evidence of unhelpful psychological processes. The most useful questions are summarised in Algorithm 1.

Algorithm 1

Step-by-step assessment and management of non-cardiac chest pain

History—the most useful questions:


  • Is there discomfort all the time?

  • If you have 10 pains in a row how many happen on exercise/at rest?

  • How long does the pain last?

  • What happened immediately before the first episode of pain?

  • Do you have difficulty filling your lungs?

  • Do you have to gasp?

  • Do you have other difficult physical symptoms (fatigue, pain, headache, knot in throat, irritable bowel)


  • Do you sleep/eat normally?

  • Do you feel depressed or down?

  • Do you often feel stressed or anxious? Are there any major stressors in your life at the moment?

  • Do you worry about your health? Do you have panic attacks?

  • How does the pain affect your life? If the pain stopped, what would you be doing differently?

  • When you have chest pain, what thoughts go through your mind? What emotions do you feel? What do you usually do? How do other people react?

  • Does anything make the pain better? Does anything make it worse?

Extend the physical examination:

  • Does the chest expand on inspiration indicating abnormal thoracic respiration?

  • Ask the patient to hold the breath for as long as possible (>30 s normal and <20 s always abnormal)

  • Does breath-holding induce the usual pain

  • Ask the patient to take deep breaths for 60 s. Does this induce the usual pain?

  • Is there localised chest wall tenderness or pain over the spine or with movement of the shoulder

Initial management

  • Before cardiac tests, prepare patients for a negative result

  • Give unambiguous reassurance that the pain is not dangerous

  • Give a clear, alternative explanation of the cause and mechanism of pain if feasible. Otherwise state that an oesophageal or musculoskeletal origin is known to be most common

  • Consider a trial of a proton pump inhibitor if gastro-oesophageal symptoms are prominent

  • Refer to a gastroenterologist or specialist in back pain if specific symptoms present

Stepped-care intervention with a clinical psychologist

  • Simple explanation and formulation of the factors affecting pain and reassurance sufficient for a minority

  • Low intensity cognitive behavioural therapy (CBT) for straightforward cases including relaxation and breathing techniques (box 4)

  • High intensity CBT for cases with complex psychosocial issues, high levels of anxiety or depression, or who do not respond to low intensity CBT

Chest pain characteristics

‘Typical’ and ‘atypical’ pain, although universally used, are almost meaningless without careful definition. ‘Typical’ is taken to mean pain precipitated by exercise and relieved by rest. In one report,63 60% had pain occasionally on exercise but only 16% had pain reproducibly precipitated by exercise. The subjectivity of the history is underlined by a study64 in which 45% were initially thought to have ‘atypical’ pain. This figure rose to 80% when the history was retaken after the finding of normal coronary anatomy at coronary angiography.

The pain may occur ‘at any time’. If related to exercise this may be after rather than during exercise or it may last for hours or even days after exercise. Sometimes the pain is triggered by trivial exercise. The Rose Questionnaire is commonly used for epidemiological studies but does not adequately define exertional pain and therefore performs poorly against objective measures of coronary disease.65–67 We developed a ‘typicality’ score68 by comparing the answers to 50 questions in 65 patients with completely normal coronary anatomy and 65 with proven coronary disease. Three questions differentiated patients with normal and diseased coronary arteries (table 1) allowing a typicality score of 0–1 (atypical) or 2–3 (typical). This scoring system was shown to predict events better than treadmill exercise testing in an outpatient population.69 Patients with normal anatomy were also more likely to report breathlessness, dysphagia and symptoms suggesting irritable bowel syndrome. There were no differences in the quality, site or radiation of the pain between the two groups. Relief by sublingual glyceryl trinitrate was reported in a similar proportion of both groups but occurred within 5 min in only 19 (29%) with normal anatomy compared with 41 (63%) with coronary disease.

Table 1

Points in the history

Pain from thoracic overbreathing is variable and may occur at rest, with minimal exertion or commonly ‘at any time’. There is often a sensation of fullness or heaviness which may be present all the time and a shirt or bra may become uncomfortable. Oesophageal pain may be precipitated by exercise but is much more likely to be precipitated by large meals or acidic or fatty foods and by bending or lying down. It may be improved by antacids. Musculoskeletal pain may also be precipitated by exercise but usually involving the arms,70 back or the affected joint. The pain from carpal tunnel syndrome sometimes radiates to the left side of the chest. There may be localised tenderness50 although, surprisingly, this did not differentiate our patients with and without proven coronary disease.68

Ask about the first episode of pain. Sometimes musculoskeletal pain begins with an obvious trauma such as lifting a heavy weight or after changing a computer screen or chair. Pain from a breathing abnormality may follow a psychological stress or an event that alters the breathing pattern such as a chest infection.

Extracardiac symptoms

Overbreathing is associated with a feeling of air hunger, of being unable to fill the lungs properly or take a satisfying breath. The patient may react with a deep breath, which makes it worse. There may be throat clearing and a sensation of globus or ‘something stuck in the throat’. Breathlessness during eating or talking (eg, on the telephone) is common. There may be ‘pins and needles’ around the mouth or the fingers usually on the left and occasionally hyperacusis and photophobia occasionally requiring the use of dark glasses. A muzzy head or inability to concentrate is common.

Formal gastro-oesophageal questionnaires exist,71 but an informal history is sufficient for clinical practice (table 1). Oesophageal symptoms include heartburn and acid reflux into the mouth (water brash) and regurgitation. Dysphagia is specific for an oesophageal abnormality, but is uncommon. Heartburn can also occur with irritable bowel syndrome.

Many patients have multiple stress related symptoms including irritable bowel (table 1).

Evidence of relevant, unhelpful psychological processes

There needs to be detailed assessment of the psychosocial factors associated with chest pain, including cognitions/beliefs, emotions, behaviours and the patient's social environment. General assessment is important to identify current life stresses and potential links between chest pain and stress, the impact of chest pain on the patient's life and goals for the future. There should be brief screening for anxiety and health anxiety, depression and panic disorder (table 1) while questionnaires more suited to a research study are given in table 2.

Table 2

Psychological questionnaires of use in a research context

Extending the examination

Evidence of abnormal breathing is common (box 2 and Algorithm 1). We found an abnormally short breath-holding time (<20 s) in 42% of our patients referred from a rapid access chest pain clinic.20 If pain is not exactly reproduced by holding the breath we ask the patient to breathe more deeply than usual for 60 s. We found that pain was reproduced exactly by voluntary overbreathing in 33%.20

Box 2

Points in the examination

Evidence of a breathing abnormality

Thoracic respiratory pattern


Throat clearing

Dark glasses with no ophthalmological need

Localised chest wall tenderness

Breath-holding time reduced (<20 s) (does this induce the usual pain?)

Voluntary overbreathing at rest for 60 s (does this induce the usual pain?)

Evidence of a musculoskeletal abnormality

Springing the spine

Movement of neck

Crepitus and reduced mobility at shoulder

The musculoskeletal examination includes looking for abnormal spinal curvature, springing the spine for tenderness and assessing mobility and tenderness of joints adjacent to the chest, mainly the neck and chest.


The patient is often referred for further evaluation of chronic chest pain after the finding of negative cardiac investigation sometimes following multiple coronary angiographies confirming normal anatomy. Ideally patients should be seen much earlier than this. If the chest pain is unequivocally ‘atypical’ (Guy's questionnaire score 0) and the coronary risk is low (eg, using European Cardiac Society's web-based scoring) there should be no need to investigate particularly if there is an obvious alternative cause of pain. Patients with ‘atypical’ pain and intermediate or high coronary risk need non-invasive testing, often exercise stress echocardiography ideally done at the time of the rapid access chest pain clinic visit.

Investigation for oesophageal disease may not be useful but a referral to a gastroenterologist for consideration of upper gastrointestinal endoscopy or pH monitoring should be considered if a trial of a proton pump inhibitor for 1–2 weeks is unsuccessful in suspected gastro-oesophageal reflux.

Overbreathing is diagnosed from the history and examination. Exercise testing with measurement of end-tidal pCO2 is a research rather than clinical technique despite being safe and easy.

Negative tests without explanation may not reassure and pursuing tests may reinforce the idea of an underlying organic abnormality.


Management starts during the diagnostic approach. Preparing patients for a negative test may make it easier for them to accept simple reassuring feedback from a cardiologist or cardiac nurse76(Algorithm 1). Treadmill exercise may not be advocated diagnostically but may still be useful therapeutically to demonstrate the ability to exercise safely. Reproduction of the usual symptoms by chest palpation, breath-holding or overbreathing at rest demonstrates their non-cardiac origin. This idea of ‘assessment as treatment’ has been used as an effective model for managing physical symptoms in general hospital outpatients.77

Avoid the tendency to use any blanket term for all patients with normal coronary arteries, for example, syndrome X, coronary spasm, microvascular angina, costochondritis. Another potentially misleading final label is ‘troponin-negative chest pain’, which fails to explain anything about the underlying mechanism and still implies a cardiac focus.

There needs to be unambiguous reassurance including an explanation of the mechanism of the pain and its likely determinants with evidence and feedback tailored to the patient's particular beliefs, fears and concerns.78 Explain that the chest pain is not dangerous and does not portend death or a heart attack. Cardiac medication tends to be continued23 but should be withdrawn or the reason for continuing it (eg, associated systemic hypertension) should be explained.

Those patients with chest pain of a gastrointestinal or musculoskeletal origin or with psychiatric disorders may respond to specific therapies.79 A proton pump inhibitor has been shown to be effective in a number of trials in patients with either non-specific NCCP or pain with features suggestive of gastro-oesophageal reflux.80–82 Although it is most effective if there is documented acid reflux or episodes with a high association with chest pain,43 it is reasonable to use a high dose proton pump inhibitor for 1–2 weeks as a therapeutic trial82 without oesophageal testing.

For those without an obvious organic cause, clear information with simple reassurance may be enough. However many patients with NCCP remain unconvinced.24 Our experience is that these often have abnormal breathing patterns or thoracic muscular tension as a mechanism of pain, with psychological processes dominant in the maintenance of the pain. This group has much in common with other functional syndromes like irritable bowel, non-specific pelvic pain, fibromyalgia and chronic fatigue, which are currently grouped as Medically Unexplained Syndromes.83 For these, CBT has been moderately successful (table 3).83 Most trials have involved between 4 and 12 sessions of individual CBT,84–87 while others have found improvements with group CBT,88 or brief interventions of 374 or even 1 session.89 Another study,90 found a single session of CBT to be ineffective, probably in part as a result of intervening too early after coronary angiography. Others have shown a positive effect from relaxation training, physical training or breathing exercise91–93 which are part of a whole CBT package for pain management.

Table 3

Trials of cognitive behavioural therapy in non-cardiac chest pain

NCCP is not a unitary diagnosis and patients respond to varying levels of explanation and reassurance, medication directed against organic mechanisms and psychological techniques. The integrated treatment of physical and psychological processes is encompassed by a biopsychosocial model, which we believe is best delivered in a combined clinic run by a cardiologist and psychologist. Furthermore a stepped approach, as also advocated by others,85 ensures that patients receive the appropriate level of care without wasting resources.

The biopsychosocial model

A biopsychosocial approach assumes that biological, social and psychological factors are inter-related and are relevant to all patient presentations.83 It assesses the relative contributions of these factors and offers an individual and holistic approach rather than a unitary treatment for all patients.

A verbal and written formulation is given with stepped care according to the needs of the patient. Management (boxes 3 and 4) could include:

  • An explanation for the symptom

  • Medication as clinically indicated for gastro-oesophageal reflux, musculoskeletal pain, insomnia, constipation or depression.

  • Breathing and relaxation exercises88

  • Cognitive behavioural approaches focusing on cognitive processes, such as attention and attribution, and behavioural experiments to test beliefs

  • Supported return to normal exercise and activities

  • Arranging contacts for help with social problems

  • Referral to specialist mental health services

Box 3

Content of the comprehensive biopsychosocial assessment


Cardiologist: In-depth assessment of: chest pain (onset, frequency, duration, nature, radiation, relieving factors, associated symptoms and typicality of pain); risk factors (family history, diabetes, hypertension, smoking and substance use); other comorbidities (eg, arthritis, gastric reflux and other persistent physical symptoms such as irritable bowel syndrome, chronic fatigue syndrome or chronic pain)

Physical examination including assessment of thoracic breathing, peak flow, peripheral oedema and musculoskeletal tenderness. Arrange further cardiac tests if required.

Psychologist: In-depth assessment of the psychological factors associated with chest pain (cognitions/beliefs, emotions and behaviours) and the patient's social environment. General assessment of current life stressors, potential links between chest pain and stress, the impact of chest pain on the patient's life and goals for the future. Brief screening for panic disorder and health anxiety, based on ICD-10 diagnostic criteria.


Cardiologist: Unambiguous assurance that the chest pain is not cardiac. A clear alternative explanation for chest pain, including the physiological mechanism. Where appropriate, medication prescribed directly or via GP, or referral to another service.

Psychologist: Where appropriate perform an individualised cognitive behavioural therapy formulation with the patient demonstrating the interaction of biological, emotional, cognitive, behavioural and social factors in non-cardiac chest pain (NCCP) and forming a basis for treatment.


Standard leaflet reiterating the findings of the clinic, normalising NCCP, placing it in a biopsychosocial model and encouraging a return to normal activities. This is supported by a letter sent to the GP, with a recommendation to stop unnecessary cardiac medications.

Box 4

Stepped-care intervention using the cognitive-behavioural formulation

Step 1: Medical treatment: Medication or referral to another service as advised by the cardiologist, based on the physiological cause of chest pain. If other psychosocial issues are also present the patient should be ‘stepped up’.

Step 2: Brief biopsychosocial—guided self-help (GSH)

GSH, supported by a cardiac nurse or clinical psychologist is delivered in 30-min sessions, with an emphasis on homework and diary keeping. Up to six sessions may be needed focused on the following:

  1. Mindful abdominal breathing: Patients with thoracic breathing patterns and/or anxiety complete a hyperventilation test and are then taught abdominal breathing skills. Homework involves 10 min of daily practice supported by an audio CD of guided breathing with aspects of mindfulness. Patients are reminded to check and adjust their breathing at regular times through the day.

  2. Progressive Muscle Relaxation: Patients with anxiety and/or musculoskeletal tension affecting chest pain are taught progressive muscle relaxation and imagery. Homework involves daily practice supported by a CD.

  3. Increasing Activity: Patients with a tendency to avoid activity are given information emphasising the benefits of exercise. Patients identify goals for activity based upon their values, which are arranged into a hierarchy of difficulty. A graded return to activity is emphasised and homework focuses on this. Psychoeducation emphasises the need to avoid an ‘overactivity–rest’ trap and to drop any unhelpful safety behaviours, such as checking the pulse rate.

  4. Cognitive Therapy: Patients with unhelpful beliefs about chest pain are supported in cognitive restructuring. Information and experiences from previous sessions are used to demonstrate how certain beliefs about chest pain are false or unhelpful. They are encouraged to use alternative, helpful thoughts.

Step 3: Complex biopsychosocial—individual cognitive behavioural therapy (CBT):

CBT covers the same information as GSH with additional information and support. Sessions are 50 min, and up to 10 are available per patient. Patient-centred therapy-led sessions are tailored to a patient's individual formulation. Additional aspects of CBT include:

  • Psychoeducation about stress, anxiety and specific psychological disorders

  • Using patient-specific behavioural experiments to explore beliefs

  • Problem-solving and cognitive work around complex psychosocial issues (such as bullying, bereavement, domestic conflict)

  • Individualised cognitive therapy for patient-specific illness beliefs

  • Behavioural activation for depression

An assessment of a pilot clinic delivering this biopsychosocial approach20 showed that the proportion with pain occurring more than once monthly fell from 100% at baseline to 61% at 3 months (p<0.001). Pain interference reduced and depression and anxiety scores improved. Use of healthcare resources also improved with a fall in consultations for chest pain over 6 months from a mean 2.6 to 0.1 (p<0.05). We did not perform a cost-benefit analysis but expect that the saving to the National Health Service across the country would be considerable. This approach is also effective for chronic pain in other sites94 and the various components have individually been tested in small randomised trials.


As we develop focused technologies for treating coronary disease, it is important that we do not forget the far more numerous group with NCCP. These patients feel equally or more disabled and are a major drain on healthcare resources, yet can be treated with inexpensive low-technology methods. Multidisciplinary clinics are effective for chest pain as for other chronic pain syndromes. They reduce the negative effects of ‘mind-body’ dualistic approaches, which commonly leave the patient feeling misunderstood, with persistent worries about their health and repeated help-seeking.

Key messages

  • Non-cardiac chest pain (NCCP) is common and patients with ‘atypical pain’ (table 1) and low cardiac risk do not usually need formal investigation.

  • When tests are needed prepare the patient for a negative result which makes acceptance easier.

  • Give authoritative reassurance that NCCP is common and not dangerous and give a plausible cause for the pain if feasible.

  • Withdraw cardiac medication or explain why they are being continued (eg, for systemic hypertension).

  • Do not give a pseudocardiac diagnosis without evidence (eg, ‘coronary spasm’).

  • If the cause is not obvious, explain that gastro-oesophageal reflux and musculoskeletal abnormalities are the most common explanations.

  • Consider a trial of a high-dose proton pump inhibitor.

  • Offer breathing exercises if there is evidence of an abnormal breathing pattern.

  • Consider psychological causative and maintaining factors (biopsychosocial approach) and offer low-intensity or high-intensity cognitive behavioural therapy.

You can get CPD/CME credits for Education in Heart

Education in Heart articles are accredited by both the UK Royal College of Physicians (London) and the European Board for Accreditation in Cardiology—you need to answer the accompanying multiple choice questions (MCQs). To access the questions, click on BMJ Learning: Take this module on BMJ Learning from the content box at the top right and bottom left of the online article. For more information please go to:

  • RCP credits: Log your activity in your CPD diary online (—pass mark is 80%.

  • EBAC credits: Print out and retain the BMJ Learning certificate once you have completed the MCQs—pass mark is 60%. EBAC/ EACCME Credits can now be converted to AMA PRA Category 1 CME Credits and are recognised by all National Accreditation Authorities in Europe (

Please note: The MCQs are hosted on BMJ Learning—the best available learning website for medical professionals from the BMJ Group. If prompted, subscribers must sign into Heart with their journal's username and password. All users must also complete a one-time registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit.


View Abstract


  • Contributors JBC wrote the draft; EMM and MSH made substantial contributions to the final article.

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles