Patients with chest pain can be triaged according to whether the aetiology is traumatic or atraumatic. The evaluation of atraumatic chest pain requires an algorithmic approach that first excludes acute myocardial ischaemia before working through the various aetiologies of chest pain. It is important to check whether the patient still has pain, and if not, to find out when his or her last episode of pain occurred.
The character of chest pain can help differentiate between cardiac, respiratory, musculoskeletal, and other causes. The type, severity, location, and duration of pain; the presence of any radiation; and exacerbating or relieving factors may be helpful in pointing towards a diagnosis. Clinical presentation alone cannot reliably determine acute coronary syndrome (ACS).
Certain characteristics of chest pain can give clues to the origin.
Constricting pain may be due to cardiac ischaemia or oesophageal spasm.
Pain that lasts over 15 minutes and is dull, central, and crushing is a feature of ACS.
Pain that radiates to the jaw or upper extremities suggests a cardiac cause.
Pain that is new in onset, or a change to the usual pattern in a patient with previously stable angina, including recurrent episodes occurring frequently with little or no exertion, or lasting longer than 15 minutes, suggests possible ACS.
Sharp pleuritic pain that catches on inspiration may originate from the pleura or pericardium and suggests pneumonia, pulmonary embolus, or pericarditis.
A sudden substernal tearing pain that radiates towards the back is the classic presentation of aortic dissection.
Precipitating and relieving factors can help distinguish between cardiac and gastrointestinal causes (e.g., GORD, peptic ulcer disease, oesophageal spasm). Cardiac pain is more likely to be brought on by exercise or emotion and is typically relieved with rest or nitrates. However, response to glyceryl trinitrate should not be used to make a diagnosis. Pain brought on by food, lying down, hot drinks, or alcohol, and relieved by antacids suggests a gastrointestinal cause. Heartburn and acid regurgitation are typical symptoms of GORD.
Abdominal pathology, such as acute cholecystitis and pancreatitis, may also cause pain referred to the chest. Associated symptoms may help to narrow the diagnosis. Acute cholecystitis may cause symptoms of fever, nausea, vomiting, persistent severe right upper quadrant pain, and jaundice. Acute pancreatitis pain is typically sudden in onset, constant, may radiate to the back, and worsens with movement. The onset of pain is sudden and is associated with vomiting in 80% of cases.
Dyspnoea is an associated symptom in patients with cardiac ischaemia, pulmonary embolism (PE), pneumothorax, or pneumonia. Nausea, vomiting, and sweating may be seen in patients with ACS.
Some patients with ACS may present with atypical symptoms: pre-syncope/syncope, nausea and vomiting, or dyspnea in the absence of chest pain. This occurs more commonly in women, people with diabetes, and those aged 75 years or more.
Past medical history and specific cardiac risk factors, such as known cardiac disease, elevated cholesterol, hypertension, smoking, and family history, support a cardiac cause. In women, pre-eclampsia, gestational diabetes, polycystic ovary syndrome, early menopause and autoimmune diseases increase the risk of cardiovascular disease. It is helpful to determine if the patient has had any previous investigations or treatment for chest pain.
A detailed drug history, prescribed and illicit, should be taken. Use of non-steroidal anti-inflammatory drugs may result in a gastric cause of pain due to gastritis, gastric ulcer, or duodenitis. Cocaine use makes cardiac ischaemia more likely.
Physical examination can further narrow down the differential.
Patients with chest pain may present in acute shock, such as those with:
ACS with cardiogenic shock
Sepsis from pneumonia or cholecystitis
Findings revealed in the cardiac examination include abnormalities in pulse or heart sounds (e.g., new onset of aortic stenosis, mitral regurgitation, or worsening of existing murmur, pericardial friction rub), hypo- or hypertension, and signs of heart failure.
Cardiac examination is often normal in ACS.
Aortic dissection may cause a pulse deficit, indicated by reduced pulse force or volume, or clinical signs of hypoperfusion (e.g., cold, blue, mottled limb) affecting the extremities. This is particularly common in a proximal dissection affecting the aortic arch, and may be unilateral or bilateral. Interarm systolic blood pressure differences of more than 20 mmHg may be present with aortic dissection.
Crepitations on auscultation in one or both bases suggest pneumonia or heart failure. Reduced breath sounds on one side can be caused by a pneumothorax, or focally due to a collapsed lobe.
Tenderness on palpation over the area of chest pain usually indicates a musculoskeletal cause, such as costochondritis. However, many patients with myocardial infarction also have chest wall pain on presentation.
Tenderness unilaterally, in a dermatomal distribution (with or without the typical rash), may be indicative of herpes zoster. The pain typically presents 2 to 3 days before the development of a rash in the affected dermatome. However, some patients present with pain for up to 1 week before development of a rash. Typically, the rash is vesicular on an erythematous base in a unilateral distribution of one dermatome.
A gastrointestinal origin of chest pain is associated with a normal cardiac and respiratory examination, unless there is existing but stable comorbidity. An abnormal abdominal examination (tenderness, rebound, guarding) makes a gastrointestinal aetiology more likely. Possible disorders include: GORD, peptic ulcer disease, cholecystitis, pancreatitis, and gastritis. In acute cholecystitis, physical examination may reveal a right upper quadrant mass. A positive Murphy's sign (the examiner’s hand rests on the right costal margin and deep inspiration causes pain) has a specificity of 96% for acute cholecystitis.
Features of a connective tissue disease such as Marfan's syndrome (tall stature, arachnodactyly, pectus excavatum, joint hypermobility, high-arched palate and narrow face) or type IV Ehlers-Danlos syndrome (translucent skin, easy bruising, small joint hypermobility) may be present in patients with aortic dissection.
Unilateral leg tenderness or swelling may indicate a deep vein thrombosis in patients with suspected PE.
Basic observations include temperature, blood pressure, pulse, and respiratory rate. In addition, the following should be monitored, using clinical judgement on the frequency of monitoring required, until a firm diagnosis has been made.
Oxygen saturation by pulse oximetry
Repeated resting 12-lead ECGs
Any exacerbations of pain and/or other symptoms
Pain relief; checking that this is effective.
A 12-lead ECG is performed in most patients unless a non-cardiac diagnosis can be made with confidence (e.g., pneumothorax). The ECG should be done as soon as possible after presentation.
ST segment changes such as ST elevation or ST depression, QRS abnormalities, arrhythmias, or tachycardia or bradycardia are characteristic findings in cardiac causes. ST segment elevation, measured at the J point, >1 mm in contiguous leads, indicates ST-elevation myocardial infarction (STEMI). ST-wave depression or T wave inversion in contiguous leads may indicate coronary ischaemia. Patients with ACS may have normal or nearly normal ECGs; where there is concern for ACS, serial ECGs should be performed.
Pericarditis may be present with diffuse ST elevation or PR depression.
ECG findings that may be present in patients with PE include sinus tachycardia (most commonly) and S wave in lead I with Q wave and T wave inversion in lead III (S I, Q III, T III pattern).
Patients with pericardial tamponade may demonstrate electrical alternans (alternate-beat variation in the amplitude or axis of the QRS complex).
A chest x-ray can confirm respiratory disorders such as pneumothorax, pleural effusion, or pneumonia.
The chest x-ray is usually normal in ACS.
Chest x-ray can provide clues to another serious cardiac pathology, such as a widened mediastinum in aortic dissection, or a large globular heart in cardiac tamponade.
In patients with concern for ACS, cardiac biomarkers should be ordered on presentation and at least every 6 to 8 hours after presentation. There is considerable variability in protocols, depending on types of biomarkers available.
Cardiac biomarkers (e.g., troponin I and T, creatine kinase [CK], creatine kinase-MB [CK-MB]) found in skeletal and cardiac muscle are raised in many situations including myocardial infarction, following a fall or seizure, myositis, hypothermia, or hypothyroidism. Troponin is more specific for myocardial injury than CK or CK-MB. Troponin is the preferred biomarker in the evaluation of patients with chest pain. CK should only be used if troponin is unavailable. Biotin (vitamin B7) supplements may interfere with the results of some troponin assays, leading to falsely low results.
High-sensitivity troponins, where available, may allow for earlier diagnosis of ACS. Patients with a non-ischaemic ECG and a negative high-sensitivity troponin T, particularly if they have had pain for over 3 hours, have a very low risk of major adverse cardiac events (MACE). In the UK, the National Institute for Health and Care Excellence guideline on the assessment of chest pain recommends that high-sensitivity troponin tests should not be used in people in whom ACS is not suspected.
One cohort study of patients who presented to the emergency department with chest pain but no ACS found that any elevation of high-sensitivity troponin was associated with an increased risk of cardiovascular events and death over a mean follow-up of three years.  Another cohort study of patients who had troponin testing for any clinical reason found that any elevation of troponin was associated with a clinically significant increase in mortality, regardless of age.
A full blood count should be considered to screen for anaemia and evidence of infection.
A renal profile is useful as a baseline test. An observational study of patients attending an emergency department with chest pain found that patients with acute kidney injury at presentation were more likely to be diagnosed with heart failure and myocardial infarction, compared with patients who did not have acute kidney injury at presentation.
Some of the differential diagnoses for chest pain can be excluded or confirmed after history, physical examination, and basic investigations have been carried out. These include STEMI, pneumothorax, pneumonia, pericarditis, and costochondritis.
Coronary angiography with percutaneous coronary intervention (PCI)
Coronary angiography with PCI is required urgently in patients with STEMI and in patients with NSTEMI who have high-risk features such as ongoing chest pain and dynamic ECG changes.
If PCI is unavailable or delayed more than 2 hours, then systemic fibrinolysis should be given for STEMI, but not for NSTEMI.
The HEART score has been developed for risk stratification of patients in the emergency department. Patients with a low HEART score (0-3) have a low risk of MACE at 6 weeks. Patients with a high HEART score (7-10) are at significant risk for MACE. Systematic reviews and meta-analyses have found that the HEART score has a sensitivity of over 95% for predicting major adverse cardiac events.
There are many other risk scores available. Guidelines vary in their recommendations on risk stratification in people presenting with acute chest pain to the accident and emergency department.
Some patients need further investigations to confirm the suspected diagnosis.
Once ACS, ventricular arrhythmias, and haemodynamic instability are excluded, patients with chest pain that is clinically considered to be ischaemic in origin can be stratified by their likelihood of having angina and risk for coronary artery disease (CAD).
High pre-test probability of CAD (>90%): should be referred for coronary angiography. Fractional flow reserve (FFR) CT is being increasingly used to clarify the functional significance of stenoses, particularly when intervention is being considered without prior functional testing. FFRCT is a non-invasive, post-processing image analysis technique that allows assessment of flow limitation across coronary stenoses by using data derived from routine diagnostic coronary CT angiography studies.
Intermediate pre-test probability of CAD (10-90%): should have stress testing with imaging (radionuclide, echocardiography, or magnetic resonance imaging).
Low pre-test probability of CAD (<10%): if the patient has an interpretable ECG (not left bundle branch block, pre-excitation [Wolff-Parkinson-White], or ventricular paced rhythm) and is able to exercise, he or she should have an exercise stress test. Otherwise the patient should have a stress test with imaging.
Guidelines differ in their recommendations on use of risk stratification and further specific investigations so it's important to check local protocols.
Patients without symptoms of angina should be evaluated for alternative causes of chest pain.
Transthoracic echocardiography is a non-invasive way of assessing cardiac function. It is necessary if cardiac tamponade is suspected and is helpful in confirming a diagnosis of pulmonary hypertension, pericarditis, aortic stenosis, and mitral valve prolapse.
For a diagnosis of aortic dissection to be made, computed tomography (CT) angiography is more useful. This allows rapid imaging and detection of life-threatening aortic pathology. Magnetic resonance aortography has a sensitivity and specificity equivalent to CT. This imaging modality also provides information on aortic valve pathology and left ventricular function, but the images take longer to acquire so it is only suitable for haemodynamically stable patients. A trans-oesophageal echocardiogram is an alternative if a dissection-skilled operator is readily available.
PE can be difficult to diagnose. A high index of suspicion is required. Initial tests include chest x-ray, ECG, and arterial blood gases, but results of these tests do not definitively establish or eliminate PE as a diagnosis. Validated scores help to predict the likelihood of PE. These include the Wells score and the Geneva score (and their simplified versions). D-dimer blood testing by enzyme-linked immunosorbent assay (ELISA) has a sensitivity and negative predictive value >95%, regardless of the calculated clinical probability, making it useful to rule out the presence of venous thromboembolism. However, specificity is much lower, with estimates from 23% to 63%. It is particularly useful in patients with a low or intermediate clinical probability assessment.
For patients identified to be at very low risk of PE, the Pulmonary Embolism Rule-Out Criteria may be used. If the patient meets all criteria (age <50 years; initial heart rate <100 bpm; initial oxygen saturation >94% on room air; no unilateral leg swelling; no haemoptysis; no surgery or trauma within the last 4 weeks; no history of venous thromboembolism; no oestrogen use), the risk for PE is considered to be lower than the risk of testing, and so a D-dimer is not indicated. D-dimer should be done for patients who do not meet all of the criteria.
A diagnosis of PE can effectively be ruled out in patients with a low probability of PE on clinical scoring with a D-dimer that is not elevated. The risk of PE within 3 months in these patients is less than 1%.
For patients with a high probability of PE on clinical scoring (i.e., PE likely) or an abnormal D-dimer, imaging is required. Multiple-detector computed tomographic pulmonary angiography (CTPA) scanning of the chest is the imaging study of choice.
If CTPA is contraindicated or not available, other imaging studies may be used, such as a ventilation-perfusion (V/Q) scan or a V/Q single photon emission computed tomography (SPECT) scan.
Point of care ultrasound in the emergency department is highly specific for diagnosing pneumothorax, pleural effusion and pericardial effusion. In a prospective observational study of patients attending an emergency department with chest pain and shortness of breath, point of care ultrasound had equal to, or higher, specificity for all indications for which it was used, except pneumonia.
If a gastric diagnosis is the more likely cause for chest pain, then investigations such as oesophagogastroduodenoscopy, oesophageal pH monitoring, oesophageal manometry, barium swallow, and Helicobacter pylori breath test can be considered. A therapeutic trial of proton-pump inhibitors can relieve symptoms in patients with GORD, once cardiac causes of the chest pain have been ruled out.
Further blood tests such as liver profile and either serum lipase or amylase may be necessary if acute cholecystitis or acute pancreatitis is suspected. These diagnoses may also require further imaging such as abdominal ultrasound and abdominal CT (for acute pancreatitis).
Needle decompression of tension pneumothorax animated demonstration
Venepuncture and phlebotomy animated demonstration
How to perform an ECG animated demonstration
Use of this content is subject to our disclaimer