Introduction

Chest pain is one of the most common presenting complaints in emergency medicine. It is also associated with life-threatening pathology, so it is arguably the most important complaint faced by emergency clinicians. It is certainly one of the most challenging. Failure to diagnose and manage appropriately patients with acute chest pain is a frequent cause of avoidable mortality and morbidity and is a leading cause of malpractice litigation. It is therefore not surprising that physicians often err on the side of caution, yet this can also have adverse consequences for the patient and society. Patient anxieties following unnecessary investigation are often unrecognized but may severely affect quality of life, while over-investigation and unnecessary hospital admission for chest pain waste millions of healthcare dollars each year.

Epidemiology

The incidence of acute chest pain presenting to the emergency department appears to be increasing. Awareness of the importance of early treatment for myocardial infarction has led to public information campaigns that increase emergency department attendances with chest pain. Meanwhile, general practitioners are increasingly being bypassed in favour of an emergency ambulance response. These changes in health service use have coincided in many developed countries with a decrease in the incidence of coronary heart disease. It therefore seems likely that patients presenting to the emergency department with acute chest pain have decreasing prevalence of acute coronary syndrome (ACS) and increasing prevalence of more benign conditions.

Differential diagnosis

The main differential diagnoses are outlined in Table 5.1.1. The most common causes of acute chest pain are ACS (unstable angina or myocardial infarction), musculoskeletal pain, anxiety, gastro-oesophageal pain and non-specific chest pain. The most serious causes (in terms of threat to life) are ACS, pulmonary embolism and aortic dissection. Since ACS is both common and life threatening, it is inevitably the primary focus of assessment.

ACS is discussed in detail in Chapter 5.2, pulmonary embolus in Chapter 5.5 and aortic dissection in Chapter 5.10.

Musculoskeletal chest pain may be related to a precipitating episode, such as chest wall injury or physical over-exertion. Alternatively, it may be caused by inflammation in chest wall structures. Tietze’s syndrome (costochondritis) is most commonly seen in women and is characterized by tenderness of the costochondral cartilages. Epidemic myalgia (Bornholm disease) is due to inflammation of chest wall muscles and pleura occurring after viral infection, typically Coxsackie B. Herpes zoster produces severe pain along the distribution of a thoracic nerve and may be misdiagnosed as musculoskeletal pain if the patient presents before any rash or vesicles have developed.

Gastro-oesophageal pain occurs when gastric contents reflux into the oesophagus or when the oesophageal muscles spasm. Pneumomediastinum can occur spontaneously after vigorous exercise, vomiting or an asthma attack or may be associated with barotrauma from diving or inhalation during drug abuse. Pericarditis is most commonly caused by viral infection, but may be associated with systemic illness, such as uraemia or autoimmune disease, or follow myocardial infarction or cardiac surgery (Dressler syndrome).

Anxiety-related chest pain is a common and frequently unrecognized cause of acute chest pain. It may also coexist and be an important factor alongside other causes of chest pain. The patient with coronary heart disease and anxiety-related chest pain presents a particularly difficult diagnostic and management challenge. Anxiety may be related to a specific serious cause of chest pain and can be exacerbated by misguided efforts to provide reassurance through diagnostic testing. In extreme cases, this can lead to ‘cardiac neurosis’ in which the patient’s anxieties about cardiac disease cause more severe disruption to their daily activities and quality of life than would be expected from the pathology that worries them.

Pleurisy is typically caused by a viral infection and produces pain that is worse on inspiration. It may be differentiated from pulmonary embolus by the presence of systemic features and the absence of breathlessness or risk factors for thromboembolism, although investigation for pulmonary embolism is often required. Pneumonia and pneumothorax can also cause pleuritic pain but should be evident on chest radiography.

There are a number of serious abdominal complaints that may present as chest pain. These include biliary colic (acute biliary pain), cholecystitis, peptic ulcer disease and pancreatitis. Failure to take a careful history and examine the abdomen may lead to delayed diagnosis.

Finally, a substantial proportion of patients will be labelled entirely appropriately as ‘non-specific chest pain’ after emergency department evaluation. These patients have pain that simply cannot be categorized into a clear diagnostic group. It is more honest to accept this than apply an inaccurate diagnostic label.

Clinical features

Clinical assessment is primarily aimed at identifying patients with a significant risk of serious pathology who require further investigation and possibly inpatient care. The most common serious pathology is ACS, so clinical assessment is often focused upon associated features. Other serious conditions, such as pulmonary embolism and aortic dissection, should not be neglected.

ACS is classically associated with chest pain that is crushing, gripping or squeezing in nature and radiates to the left arm, but presenting features in the emergency department may be much more variable, particularly in patients with no past history of coronary heart disease and a non-diagnostic ECG. Table 5.1.2 shows the likelihood ratios of clinical features that may help to diagnose ACS. It is notable that pain radiating to the right arm or to both arms is a powerful predictor of ACS. Pain described as ‘burning’ or ‘like indigestion’ can be associated with ACS, as can pain occurring on exertion. So the diagnoses of gastro-oesophageal reflux or stable angina should be made with great caution. Pain that is sharp or associated with inspiration or movement is less likely to be cardiac, but these findings alone do not exclude ACS. Risk factors for coronary heart disease should be routinely recorded, although they may have surprisingly little diagnostic value. This is perhaps because patients are aware of these risk factors and take them into account when deciding whether or not to seek help for episodes of chest pain. In this respect, social and cultural factors may have an important influence upon patient’s interpretation of their symptoms and health-seeking behaviour.

Clinical examination is of limited diagnostic value and is mainly aimed at identifying non-cardiac causes of chest pain or complications of ACS, such as arrhythmia, heart failure or cardiogenic shock. Pain that can be reproduced by chest wall palpation is less likely to be cardiac, but this finding does not exclude the possibility of ACS. It is also important to determine specifically that chest wall palpation is reproducing the pain that led to presentation. Simply identifying chest wall tenderness has little value – everyone has a tender chest wall if you press hard enough!

Clinical assessment should not just focus upon ACS, but should aim positively to identify other causes. Pulmonary embolism is diagnostically challenging. Suspicion should be raised by chest pain that is clearly pleuritic in nature, haemoptysis, associated breathlessness, features of deep vein thrombosis or risk factors for venous thromboembolism (immobilization, malignancy, recent trauma or surgery, pregnancy, intravenous drug abuse or previous thromboembolism). Clinical examination may reveal tachycardia, tachypnoea or features of deep vein thrombosis (see Chapter 5.5). Aortic dissection is characterized by severe pain radiating to the back with associated diaphoresis. Neurological symptoms or signs, sometimes transient, are common. Clinical examination may reveal discrepancy between blood pressure in right and left arms (see Chapter 5.10).

Clinical assessment of chest pain should always include examination of the abdomen to identify tenderness, guarding, rebound tenderness or a positive Murphy’s sign.

Unnecessary investigation can be avoided if non-life-threatening pathology can be confidently diagnosed by clinical assessment. Pain that is reproduced by chest wall palpation in a patient at low risk of coronary heart disease and with no significant risk factors for pulmonary embolus can be confidently diagnosed as musculoskeletal. A positive diagnosis is particularly valuable for the patient who is primarily suffering from anxiety-related symptoms. In this case, pain is typically described as tightness around the chest and associated with a feeling of restricted breathing. Other features include palpitations (particularly awareness of the heartbeat), sweating, breathlessness, light-headedness, feelings of panic, or paraesthesia of the lips or fingertips.

Clinical investigations

The ECG is the most useful clinical investigation and should be performed on all patients presenting with acute non-traumatic chest pain. Table 5.1.3 shows the value of ECG features for diagnosing myocardial infarction. It is important to recognize that a normal ECG does not rule out myocardial infarction. ST-segment elevation or depression, new Q waves and new conduction defects are specific for acute myocardial infarction and predict adverse outcome. Patients with these features should be managed on a coronary care unit. Other changes associated with myocardial infarction are less helpful. T-wave changes are often non-specific and may be positional or due to numerous other causes. ECG changes in pulmonary embolism are also non-specific.

A standard 12-lead ECG may be augmented by serial ECG recording or continuous ST-segment monitoring. These may detect evolving ECG changes or dynamic ST-segment changes. However, these techniques may also identify non-specific false-positive changes, such as minor T-wave inversions, especially if they are used inappropriately in patients with a low risk of coronary heart disease. ST-segment monitoring was developed for the high-risk coronary care population; in low risk emergency department patients with chest pain, it has a very low yield of significant findings.

Like clinical examination, the chest radiograph is mainly intended to identify non-cardiac causes for chest pain, such as a pneumothorax or fractured rib and complications of myocardial infarction, such as left ventricular failure. Although it is often routinely ordered, it is not usually helpful.

Cardiac biomarkers are key investigations in acute chest pain and are a source of much heated debate. They are also a progressively developing technology, so this chapter will focus upon the principles that should guide their use.

Three key features determine the clinical value of a cardiac biomarker. Sensitivity tells us how good the marker is at identifying patients with disease, and thus how useful it is for ruling out myocardial ischaemia. Specificity tells us how good the marker is at identifying patients without disease, and thus how useful it is for ruling in myocardial ischaemia (i.e. a specific test that is positive suggests that the patient is very likely to have ischaemia). The prognostic value (often expressed as a relative risk) tells us how good the marker is at predicting future adverse events, such as death, myocardial infarction or life-threatening arrhythmia.

Intuitively, clinicians tend to be most concerned about sensitivity. If a marker lacks sensitivity then it may miss cases of myocardial infarction leading to potentially catastrophic discharge home without appropriate treatment. However, sensitivity and specificity are often related and may be influenced by the threshold of the marker used to determine a positive test. The lower the threshold used for a positive test, the higher the sensitivity and the lower the specificity. Many evaluations of new markers deliberately optimize sensitivity by selecting a low threshold and sacrificing specificity. This may be an acceptable trade-off in a high-risk population, but emergency department patients with no past history of coronary heart disease and a non-diagnostic ECG typically have a low prevalence of myocardial infarction (<10%). In these circumstances, a test with low specificity will generate many false-positive results requiring hospital admission and investigation, as well as unnecessary anxiety for the patient.

Furthermore, it should be remembered that even a random process can be made to appear sensitive by setting a low diagnostic threshold so that most of the results generated are deemed positive. For example, rolling a pair of dice to diagnose myocardial infarction will have 97% sensitivity if all results, except double six, are deemed to be positive. This is, of course, an extreme example, but biomarkers with diagnostic value that is little better than rolling dice have been promoted by studies that report high sensitivity while hiding away poor specificity. Sensitivity should always be reported with a corresponding specificity.

The prognostic value of a marker is arguably even more useful than its diagnostic parameters, particularly if the marker can predict high-risk patients who will benefit from treatment. If a prognostically useful marker is positive, then we know the patient has the potential to benefit from intervention; if it is negative then we know that, even if further investigation is required to identify the exact cause of their chest pain, they are unlikely to benefit from hospital admission and treatment.

Prognostic considerations explain changes in the definition of myocardial infarction. The original World Health Organisation (WHO) definition of myocardial infarction was based upon creatine kinase, a cardiac marker with limited sensitivity and specificity and only weak evidence of an association with adverse prognosis. Subsequent definitions of myocardial infarction, from the American Heart Association and European Society of Cardiology, used troponin as the biomarker based on research showing elevated troponin levels were associated with increased risk of adverse outcome. Furthermore, research has shown that informing clinicians that their patient has an elevated troponin level can alter patient management and reduce the risk of major adverse cardiac events over the following year. This is important evidence that measuring troponin provides patient benefit by reducing adverse outcomes.

Creatine kinase is released by damaged myocardium, but is also released by muscles and the liver and is measurable in the blood in the absence of pathology. Its MB isoenzyme (CK-MB) is more cardiac-specific but shares the same problems. Substantial myocardial damage is required to produce an elevated CK-MB, but CK-MB may also be elevated in the absence of myocardial injury. Its role in diagnosis is generally limited to situations where troponin testing is unavailable or likely to be unreliable.

There are two troponin assays, troponin I and troponin T, with little to choose between them in terms of diagnostic or prognostic performance. The diagnosis of myocardial infarction is based upon elevation of the troponin level above the 99th percentile upper reference limit of a normal reference population. This value will vary between assays. The precision of the assay is described by the coefficient of variation (CV) at the 99th percentile. There have been several generations of troponin assays with newer assays having higher sensitivity and better precision. High-sensitivity troponin assays have optimal precision (≤10% CV) at the 99th percentile and can detect myocardial infarction earlier after symptom onset than other assays. However, this appears to be at the cost of an increased number of low troponin rises that are of unclear diagnostic and prognostic significance.

Troponin elevation is not specific for myocardial infarction and low levels measured using a high sensitivity assay may not indicate significant pathology. Even substantial troponin elevations may not be due to ACS. Troponin can be elevated in pulmonary embolus, sepsis, renal failure, congestive cardiac failure and a number of other illnesses.

The use of troponin has tended to be limited by its lack of early sensitivity. It has been estimated that standard troponin assays take up to 12 hours after symptom onset to achieve optimal sensitivity, so if it is used too early after symptom onset it may produce a false-negative result. This has led to the widespread practice of delaying troponin measurement to at least 12 hours after symptom onset to achieve optimal sensitivity. This practice is problematic because most patients present a few hours after symptom onset, so enforcing a 12-hour delay will typically require hospital admission or use of observation facilities. If there is limited availability of such facilities, clinicians may feel under pressure to discharge the patient without any testing. Thus, a strategy intended to increase patient safety may paradoxically put patients at risk when applied to the real world. Newer assays allow a 6-hour testing strategy and recently developed high-sensitivity assays appear to have good early sensitivity and may be able to rule out myocardial infarction within a few hours of presentation to hospital. However, as mentioned above, this may be at the cost of limited specificity resulting in the generation of troponin elevations that have little diagnostic or prognostic significance. Thus, the choice of testing strategy involves a trade-off between sensitivity and specificity. It is important to know the test used in your institution and its performance characteristics in order to decide on an appropriate serial testing strategy.

Awareness of the limited early sensitivity of standard troponin assays has led to research into alternative biomarkers that can detect myocardial infarction in the initial hours after symptom onset. These biomarkers often have poor specificity so any improvement in sensitivity achieved by combining troponin with another biomarker involves a trade-off in terms of reduced specificity. The development of high-sensitivity troponin assays is likely to undermine the need to develop and use alternative early biomarkers. Future evaluations of alternative early biomarkers will need to include comparison with high-sensitivity troponin.

The most extensively evaluated early biomarkers to date are myoglobin, heart-type fatty acid binding protein (HFABP) and ischaemia modified albumen (IMA). Table 5.1.4 shows the sensitivity and specificity of these biomarkers at presentation to hospital, compared to troponin. None of the alternative biomarkers have sufficient sensitivity to rule out myocardial infarction at presentation. Adding HFABP or myoglobin to a contemporary troponin assay at presentation increases sensitivity at the expense of specificity, but it is not clear whether either biomarker can improve the early sensitivity of a high-sensitivity troponin assay. IMA appears to have insufficient specificity to play any useful role in diagnosis.

Biomarkers are continually being developed and emergency physicians can expect to see headline-grabbing publications extolling their virtues. However, they should be wary before indiscriminately using new markers in their patients with chest pain. As described earlier, the emergency department population with chest pain are a heterogeneous population with a relatively low prevalence of ACS compared to the high-risk patients that usually comprise research study populations. Indiscriminate use of markers with limited specificity will lead to many false-positive results and consequent patient anxiety, unnecessary investigation and waste of resources.

Provocative cardiac testing, usually using an exercise treadmill, has been used in a number of emergency departments. Patients typically undergo a short period of observation and cardiac marker testing to rule out myocardial infarction before receiving an exercise treadmill test. Concerns about the safety of this procedure have been addressed by data from a number of centres, but it should be recognized that selection of low-risk patients plays a key role in ensuring safety.

Exercise treadmill testing has relatively poor sensitivity and specificity and cannot reliably rule in or rule out coronary artery disease. However, it is prognostically useful and predicts risk of adverse events over the months following attendance. It is therefore used to risk-stratify more than to diagnose. A patient with a negative treadmill test may have coronary artery disease but can be reassured that they are at low risk of adverse outcome. It is not clear whether the additional prognostic information provided by exercise treadmill testing justifies widespread use.

CT coronary angiography is developing an increasing role in the diagnostic evaluation of acute chest pain. Unlike exercise treadmill testing, CT coronary angiography detects coronary obstruction with a reasonable degree of accuracy. Sensitivity and specificity are both around 90% against a reference standard of invasive coronary angiography. However, coronary atheroma is a common finding and detecting atheroma, or even coronary obstruction, does not necessarily mean that this was the cause of the patient’s chest pain. Studies of the prognostic value of CT coronary angiography in patients with suspected ACS have been limited by selection of patients with low rates of adverse outcome and have, as yet, failed to show worthwhile prognostic value. This means that, although CT coronary angiography is a promising diagnostic test for acute chest pain, evidence is lacking that its use improves patient outcomes and its role in routine assessment remains unclear.

The combination of observation and cardiac marker testing to rule out myocardial infarction, followed by provocative cardiac testing to risk-stratify or CT coronary angiography to diagnose coronary artery obstruction, has been adopted in many hospitals in the form of a chest pain unit or pathway. These have a number of potential benefits for patients and health services and some evidence to suggest that they reduce the probability of admission, reduce the risk of discharge with ACS, improve patient satisfaction and quality of life and reduce health service costs. However, as an organizational intervention, the effect of the chest pain unit will depend upon local circumstances and may be influenced by staff attitudes, professional roles and local leadership. Furthermore, the presence of a chest pain unit may attract additional attendances with chest pain. Whether this represents identification of unmet demand or unnecessary work is a matter of opinion.

A number of clinical risk scores have been developed to risk-stratify patients with suspected ACS. The Goldman algorithm and acute cardiac ischaemia time insensitive predictive instrument (ACI-TIPI) were developed and validated on large cohorts of patients with chest pain in the 1980s and 1990s. The Goldman algorithm uses a series of questions about the patient’s age, clinical history and ECG findings to categorize patients into a low (<7%) or high (>7%) risk of myocardial infarction, based on the WHO definition used at the time. ACI-TIPI can be incorporated into a computerized ECG. The user enters the patient’s age, sex and whether chest or left arm pain is the primary symptom. The computer then uses these data and analysis of the ECG to generate a probability of acute cardiac ischaemia.

The thrombolysis in myocardial infarction (TIMI) score has been developed and validated as a predictor of adverse outcome in patients with diagnosed ACS (see Chapter 5.2). Studies have evaluated the TIMI score in emergency department patients with suspected ACS and shown that higher scores are associated with a higher risk of adverse outcome. This has led to the TIMI score being used to risk stratify patients with chest pain before a diagnosis of ACS has been confirmed. The ECG and cardiac biomarker components are the most powerful predictors among the elements of the TIMI score and a modified version of the TIMI score, with extra weighting for these elements, has been proposed for patients with acute chest pain. It is not clear whether either the original or modified TIMI score provides better prognostic information than that provided by ECG and biomarkers alone.

The Global Registry of Acute Coronary Events (GRACE) freedom-from-event score has recently been developed to predict adverse events in people with non-ST-elevation acute coronary syndrome and can identify up to 30% of the admitted population who are at low risk of death or any adverse in-hospital event. It is more complex to calculate than the TIMI score but bedside calculation can be easily facilitated by mobile technologies. The GRACE freedom-from-event score offers an alternative to the TIMI score for predicting adverse outcome but further validation in the wider chest pain population is required.

Treatment

Treatment of acute chest pain is directed at the specific cause. The treatment of ACS is outlined in Chapter 5.2, pulmonary embolus in Chapter 5.5 and aortic dissection in Chapter 5.10.

Musculoskeletal chest pain, whether due to muscular strain, chest wall injury, Tietze’s syndrome or epidemic myalgia, should be treated with simple analgesia and the patient advised to see their general practitioner if the pain persists beyond a few weeks. It is also worth considering whether anxiety may be exacerbating the patient’s symptoms.

Gastro-oesophageal pain can be treated acutely by antacids, although the diagnostic value of observing relief of pain with antacids (the so-called ‘GI cocktail’) is debatable. ACS can present as burning or indigestion-type pain and, pain being typically fluctuant, may ease coincidentally with administration of an antacid. Gastro-oesophageal pain should be diagnosed with caution and ideally only after ACS has been investigated and ruled out. In these circumstances, a course of treatment with a proton pump inhibitor is appropriate. Follow up will depend upon local practice along with the duration and severity of symptoms.

Anxiety-related symptoms range from simple chest wall muscular tension to panic attacks, hyperventilation syndrome and cardiac neurosis. Treatment should therefore be tailored to the patient’s individual needs. In many cases, anxiety will be an understandable reaction to concerns about heart disease or other serious pathology. The first step is therefore to provide clear and unequivocal reassurance. If diagnostic uncertainty makes this impossible then it may still be possible to provide reassurance by highlighting the excellent prognosis of patients with chest pain whose tests are negative. Information leaflets have been shown to reduce anxiety after emergency department attendance with chest pain. Patients with more severe symptoms may benefit from relaxation techniques, cognitive behavioural therapy or treatment with an antidepressant. These are best arranged through the patient’s general practitioner.

Managing anxiety in the emergency department patient is often complicated by difficulties in satisfactorily ruling out serious physical illness. A diagnosis of anxiety may be considered likely, but until cardiac testing is complete (perhaps even involving coronary angiography), the treating physician may be reluctant to discuss treatment of anxiety with the patient. This is inappropriate. If the patient has significant anxiety-related symptoms then this will adversely affect their quality of life and should be addressed regardless of whether they ultimately also need treatment for cardiac disease.

Non-specific chest pain obviously presents a challenge. With no clear diagnosis it is difficult to advise an appropriate treatment. However, patients can be advised that, although no clear diagnosis can be made, about half such patients presenting to the emergency department have no further episodes of pain over the following month. Those who do suffer further episodes of pain are unlikely to be troubled. Treatment is therefore unlikely to be required.

Finally, an acute episode of chest pain provides an opportunity to identify and manage cardiac risk factors at a time when the patient is likely to be most receptive to lifestyle advice. Smokers should be advised to use the episode as a stimulus to stop smoking and referral to a smoking cessation service arranged. General dietary and exercise advice may also be helpful. Blood pressure, blood glucose and lipid profile may be requested as part of clinical assessment, although any abnormalities identified should preferably be referred to the patient’s general practitioner, who will be best placed to provide overall cardiovascular risk assessment, intervention and long-term follow up.

Prognosis

Prognosis will also depend upon the underlying pathology and the prognoses of various causes of chest pain are discussed in the relevant chapters of this book. Patients with no obvious diagnosis after clinical assessment, ECG and troponin testing have an excellent prognosis. There is some evidence that patients who attend the emergency department with chest pain have a higher risk of adverse cardiac events than the general population, even if cardiac disease is ‘ruled out’ at initial presentation, but this risk is not high enough to warrant active intervention beyond ensuring that any cardiac risk factors identified have been addressed.

Likely developments over the next 5–10 years

Chest pain is responsible for a substantial and growing number of emergency medical admissions in many countries. This is placing a major burden upon healthcare systems. The value of hospital admission for low risk ACS and pulmonary embolus is being questioned and it is likely that there will be increasing efforts to manage acute chest pain without admission to hospital. These efforts may be successful for younger patients with no co-morbidities and a single potentially serious cause for their chest pain, but may be difficult to implement among the growing population of older patients with co-morbidities or multiple potentially serious causes for their pain.

High-sensitivity troponin assays are increasingly replacing other contemporary assays. This will lead to more positive results and concerns that some of these may represent false-positive results with little prognostic value. Emergency physicians may therefore need to be more circumspect in their use of troponin and avoid measuring troponin in patients with a very low risk of ACS in whom an elevated high-sensitivity troponin measurement just above the diagnostic threshold is likely to cause confusion rather than provide a clear diagnosis.

Point-of-care troponin testing has been available for some time but widespread adoption has been limited by a number of factors, including lack of usability and difficulties converting reduced turnaround times into reduced length of hospital stay. The development of hand-held point-of-care machines using finger-stick capillary blood samples could improve the usability of point-of-care testing and extend use to the pre-hospital setting.

CT coronary angiography is likely to be more widely used in the assessment of acute chest pain. In the absence of randomized trial evidence of effectiveness, it will not be clear whether patients are benefitting from increased use. Both high sensitivity troponin assays and CT coronary angiography have the potential to increase the diagnostic yield of positive tests in the investigation of acute chest pain leading to over-burdening of cardiology services.

Chest pain management is likely to be influenced by changes in health service policy, which are, in turn, likely to depend upon local social, political and economic factors. These changes will be variable and may be unpredictable. On one hand, public awareness of the medical significance of chest pain and policies aimed at increasing rapid access to care may lead to increased numbers of patients presenting with chest pain. On the other hand, reorganization of services and attempts to control costs may result in an opposite effect. Specifically, development of primary angioplasty services may lead to centralization of chest pain services and patients bypassing facilities that do not provide primary angioplasty.