The pericardium is a relatively inelastic fibro-serous sac composed of two layers about 1-2mm thick, each made of a thin serous mesothelial monolayer lining the fibrous outer portion of the parietal pericardium. During embryogenesis, the heart tube invaginates the inner serosa, which folds onto itself, and thus lines the myocardium and the inner fibrous pericardium in a contiguous layer. This creates the pericardial space in an analogous fashion to the pleural space, and is typically filled with about 15-35mL of fluid.1-3 The pericardium is innervated by the phrenic nerve, is supplied by the internal mammary arteries, and is held in place in the thorax by attachments to the sternum, vertebral bodies and the diaphragm. Pericardial disease and pericarditis — a non-specific inflammation of the pericardium — are commonly encountered clinical conditions and may present in isolation or as part of a systemic disorder. As such, patients should be evaluated individually, based on their clinical presentation, medical history and immunological state with regard to the degree of evaluation and investigation that should be performed to determine the aetiology.
Although it is difficult to accurately determine the true incidence, pericarditis is thought to account for about 0.1% of all hospital admissions and about 5% of non-ischaemic chest pains seen in the ED.4,5 Specific aetiologies (see table 1) in acute pericarditis can be challenging to determine. This is illustrated by a study in 2007 where patients with acute pericarditis were examined and a specific diagnosis was found in only 17% of cases.1 Autoimmune, neoplastic, tuberculous, and purulent were the most common specific aetiologies identified.1 In otherwise healthy and immunocompetent patients, over 90% of cases encountered are either due to a viral infection or idiopathic.1,6-8 As such, investigation for a specific cause is not necessary unless there is clinical suspicion of tuberculous, purulent, neoplastic, systemic involvement, or of a haemodynamically significant effusion (which suggests an aetiology other than viral).4 It is recommended that patients with a fever over 38°C, large pericardial effusion over 20mm depth, features of tamponade, sub-acute course or who are refractory to NSAIDs be admitted for further investigation and treatment, while those without any of these features may be treated as outpatients.
The diagnosis of acute pericarditis is based on the presence of two out of four criteria:
Recurrent pericarditis is diagnosed in patients with a past history of pericarditis, typical pleuritic chest pain and any of the following features: fever (temperature over 38°C), pericardial rub, ECG changes, new or worsening pericardial effusion, or elevation of inflammatory markers. Recurrence usually occurs within 18 months following an initial attack, however, it may manifest at any time.4,8 Fevers over 38°C are unusual for a viral origin and should prompt concern for either a systemic disorder or purulent bacterial pericarditis.
Chest pain is the most sensitive sign of acute pericarditis with an infectious aetiology. However, it may be minimal in uraemic and rheumatological conditions.7,9 The pain is usually sharp in nature, sudden in onset, may radiate to the trapezius ridge, and is characteristically positional in nature. The pain is classically said to be relieved by leaning forward and exacerbated by lying supine.10 The diaphragmatic irritation of pericarditis is referred to the left supraclavicular area and the left trapezius ridge via the phrenic nerve. Trapezius ridge pain, which is specific for pericardial inflammation, should be differentiated from shoulder tip pain in those presenting with associated chest symptoms. There are a few characteristics from the history that can help differentiate the pain of pericarditis from the pain experienced classically in an MI. Typical ischaemic pain is a “crushing”, “squeezing”, “heavy” pain, like someone is sitting on their chest, not changed with respiration or position, radiating to the jaw, neck, and/or arms, lasting minutes to hours. Pericardial pain is typically described as “sharp”, “stabbing”, increasing with inspiration, relieved by leaning forward, exacerbated by leaning back, radiating to the trapezius ridge and back, and lasts hours to days. Table 2 gives a summary of chest pain differentiation.
The pericardial friction rub is the most specific sign for pericarditis and can be noted to be transient, mono-, bi- or triphasic in relation to its occurrence within the cardiac cycle. It is a high-pitched, grating or scratchy sound (described as similar to the sound of walking on dry snow) best auscultated at the lower left sternal edge and the borders of the cardiac silhouette.10 In one study of patients with a pericardial rub, 56% were triphasic. The three phases of the triphasic rub correspond to the atrial systole, the ventricular systole, and the early ventricular diastole.11 This may be confused with a pleural rub, so it must be carefully timed to note its co-incidence with the cardiac cycle or the respiratory pattern. Bedside diagnosis of pericarditis can be made on history and the presence of a rub, but the sensitivity of this finding seems to depend on the frequency of auscultation, as rubs come and go.10 The presence of a pericardial effusion, however small, may reduce a friction rub considerably.
The presence of a pericardial effusion, along with a classic history, is highly specific for acute pericarditis. A transthoracic echocardiogram is recommended in the initial workup of all patients with suspected pericarditis to document the presence and extent of a pericardial effusion.12 It is a specific and sensitive investigation to detect haemodynamically significant effusions and the presence of tamponade physiology. The apical four-chamber view with the patient in left lateral position to scan for fluid above the right atrium is one of the most sensitive and specific tests to demonstrate a pericardial effusion (see figures 2-4). Effusions can be classified as small, medium or large on echocardiography.
Small effusions cause minimal separation between the pericardial layers and are less than 1cm at the widest point. Moderate effusions are represented by an echolucent space narrower than 2cm. Large effusions are circumferential and usually at least 2cm wide. It is important to point out, however, that the size of a pericardial effusion is not necessarily indicative of the presence or absence of tamponade physiology. Even small effusions surrounding and compressing the right atrium may cause haemodynamic instability, especially after cardiac surgery. If tamponade is present, echocardiography is able to show diastolic collapse of the right atrium and/or the right ventricle or, in severe tamponade, the left ventricle. The right-sided cardiac chambers are prone to compression, due to their being relatively low pressure structures. Chest X-ray can also be used to detect large effusions of greater than 250mL with the presence of cardiomegaly compared with previous films (figure 5).6
In the setting of acute pericarditis, inflammation of the myocardium (myopericarditis, myocarditis or perimyocarditis) may result in the typical ECG changes of widespread ST elevation (classically ‘saddle-shaped’) and PR segment depression. In fact, in a pure pericarditis, ST and PR segment deviation should not be present. For example, in uraemic patients with fibrinous pericarditis, the typical ST changes are often not present. The ST segment changes are thought to reflect the disruption to ventricular repolarisation, and the PR segment deviation reflects changes in atrial repolarisation. ECG changes are thought to evolve in four stages:
The temporal relationship between these ECG changes is not well understood. One study analysed 50 consecutive ECGs of acute pericarditis by evaluating the prevalence of each ECG abnormality and grouping them according to how quickly the patient presented after symptom onset.13 It found that PR changes were the only ECG findings detected in the patients presenting earliest following the onset of chest pain. The study also found that PR changes are the quickest to normalise, and that ST segment changes persist longer, suggesting that the first ECG changes in pericarditis are those involving the PR segment.13 Differentiating acute pericarditis and MI on the basis of an ECG alone can sometimes be challenging, and thus a thorough history and examination are, as always, of utmost importance. Some of the features of a MI on ECG that can help differentiate these two aetiologies are summarised in table 3.6,14,15 A ratio of ST segment elevation to peak T-wave amplitude >0.25 in leads I, V4, V5, or V6 are all virtually diagnostic of acute pericarditis.14 Significant elevation of cardiac biomarkers in pure pericarditis is unusual. However, small troponin leaks are not uncommon, probably owing to some inflammatory involvement of the epicardium (consistent with a diagnosis of myopericarditis when present).
Once the diagnosis of pericarditis is entertained, the decision should be made as to whether the patient can be safely treated as an outpatient, or needs to be admitted for further evaluation and inpatient management. A widely cited paper on the outpatient treatment of acute pericarditis is helpful to keep in mind when making this decision. In this study, consecutive cases of acute pericarditis were taken from 1996 to 2001, excluding those with high-risk features (fever over 38°C, subacute onset, immunosuppression, trauma, oral anticoagulant therapy, myopericarditis, large pericardial effusion, or presence of cardiac tamponade).7 Included patients were treated with aspirin 800mg every 6-8 hours for 7-10 days, tapering the dose over 2-3 weeks, along with a proton-pump inhibitor to reduce the incidence of gastroduodenal toxicity. There were 254 patients enrolled as lowrisk patients, and outpatient treatment was successful in 87% of these.7 It should be noted that no serious outcomes were reported in the outpatient cohort after a mean follow-up of 38 months. Results of this trial suggest that outpatient treatment of acute pericarditis in low-risk patients is safe and efficacious.7
In terms of the treatment protocol itself, NSAIDs have been a mainstay of empirical treatment for many years, and their efficacy has been well established in the literature. High-dose aspirin (2-4g daily in divided doses) is the preferred NSAID in those patients with acute coronary syndrome, ischaemic heart disease or who need to be on aspirin for any other reason. It should particularly be used in patients with a recent MI, as it has been shown in animal models that NSAIDs other than aspirin may impair scar formation.6 Ibuprofen (1.6-3.2g daily in divided doses) is the preferred option otherwise, as it has a lower incidence of adverse effects, and has benefits in terms of promoting coronary blood flow, whereas indomethacin (75-225 mg daily, common regimen is to use 50mg tds) should be avoided in those with ischaemic heart disease as it impairs flow through the coronary arteries. A tapering dose is suggested after the acute phase has subsided and symptoms have resolved. Initially patients should be on a thrice-daily dose of NSAID, which can be reduced to twice-daily dosing after 1-2 weeks, and reduced again to once-daily dosing for another week or two, contingent upon persistent symptom resolution. In cases of severe chest pain associated with acute pericarditis, intravenous ketorolac (a parenteral non-steroidal) has been shown in small trials to be effective in relieving pain.
Colchicine is an alkaloid phenanthrene derivative obtained from various species of colchicum, a flowering plant. It decreases leucocyte chemotaxis and phagocytosis and thereby inhibits release of chemotactic glycoprotein. It also inhibits cell division in leukocytes by interfering with the mitotic spindle. Overall, most patients with a first episode of pericarditis have fast resolution of symptoms when treated with NSAIDS alone. As we will summarise below however, when used as an adjunct to NSAID therapy, colchicine reduces symptom duration and decreases the rate of recurrent pericarditis. Colchicine has been shown to be safe when given long term, is generally well tolerated, and may be used as monotherapy in patients who cannot tolerate NSAIDs. As such, most investigators recommend that colchicine be added to NSAIDs in the management of a first episode of acute pericarditis. Treating acute pericarditis to reduce symptoms and rate of recurrence, however, is currently considered an off-label use for colchicine. On the basis of currently available trial evidence, the European Society of Cardiology (ESC) guidelines have given a class IIa recommendation (“reasonable to administer treatment as benefit outweighs risk, but additional studies needed”) to the use of colchicines 0.5mg a day in the setting of acute pericarditis as a combination therapy with NSAIDs.
The 2005 COPE trial (Colchicine for Acute Pericarditis) was a randomised controlled trial that illustrated that colchicine, along with conventional NSAID use in the acute treatment of pericarditis was not only safe, but afforded faster resolution, and reduced recurrence at 18 months.8 It was designed as a prospective, randomised, open-label study, and evaluated 120 patients presenting with their first episode of acute pericarditis secondary to variable aetiologies. The patients were randomised to either conventional therapy with aspirin alone, or to the experimental treatment with colchicine in addition to aspirin. The dose of colchicine used was 1-2mg for the first day and then 0.5-1mg daily for three months (the maximum evidence-based length of treatment with colchicines is six months in recurrent cases). Aspirin was dosed at 800mg every 6-8 hours for 7-10 days and then tapered over 3-4 weeks. Recurrence rates over 18 months were 10.7% in the colchicine group vs 32.3% in the group treated with aspirin alone. Similarly, symptom persistence at 72 hours was significantly reduced in those treated with colchicine at 11.7% vs 36.7% in the aspirin-alone group. In addition, no serious side effects were noted in the group treated with combined therapy.8 A meta-analysis published in 2010 investigated the safety and efficacy of colchicine in the primary and secondary prevention of pericarditis.16
The investigators evaluated a total of five randomised trials including three double-blind trials and two open-label randomised trials with a combined mean patient followup of 13 months. It demonstrated that colchicine use reduced the risk of pericarditis for both primary and secondary prevention, without any statistically significant increase in adverse reactions. However, it also demonstrated that the colchicine group had more drug withdrawals (8%), with gastrointestinal side effects the most common cause.16 Rare side effects of prolonged colchicine therapy include myo- and hepatotoxicity and bone marrow suppression, evidently more common in patients with chronic renal insufficiency. Although the potential for colchicines to be used as a first-line treatment for acute pericarditis is promising, no placebo-controlled multicentre randomised trials have been performed and, as such, the treatment of patients with colchicines in this setting needs further study. Two small randomised, placebo controlled trials (the CORP and CORE trials, totalling about 200 cases) have evaluated the use of colchicine in patients with a first episode of recurrent pericarditis. Using conventional therapy (aspirin, ibuprofen or prednisone) alone or in combination with 1-2mg of colchicine on the first day followed by 0.5mg once or twice daily for six months, investigators were able to show in both trials a significant reduction in the rate of recurrence at 18 months when using combination therapy. In addition, combination therapy with colchicine also significantly reduced symptom persistence at 72 hours.
There is a concern that anticoagulation in the setting of acute pericarditis may potentially lead to the development of a haemorrhagic effusion and cardiac tamponade. However, a multivariate analysis of 453 patients conducted to evaluate the factors associated with poor outcomes in pericarditis found no evidence that anticoagulation was associated with worse outcomes.1 Additionally, a study of patients with acute pericarditis or myopericarditis found the use of heparin was not associated with a significant increase in the incidence of cardiac tamponade.17 Although there is no evidence to support the concern that concomitant anticoagulant use in acute pericarditis results in an increase in adverse outcomes, the use of certain NSAIDs may increase the efficacy of warfarin and thus close attention must be paid to INR monitoring.
Duration of therapy in acute pericarditis is a controversial topic, and while some clinicians treat with high-dose NSAIDs until clinical symptom resolution occurs and then taper, others use hs-CRP (high-sensitivity C-reactive protein) as a marker for resolution of pericarditis and begin to taper the NSAID only when this normalises. Hs-CRP is a well-known marker of acute inflammation, however it is non-specific, as any inflammatory process will cause a rise in this peptide. A 2011 study evaluated the usefulness of hs-CRP in the investigation and management of acute viral or idiopathic pericarditis.18 The study showed that about three-quarters of patients with acute pericarditis had elevated hs-CRP, and of the remainder without elevations, 84% were thought to be the result of either early assessment or previous anti-inflammatory use. Normalisation occurred for 60% of cases in one week, 85% at two weeks, 95% at three weeks and 100% at four weeks. The researchers found that elevated hs-CRP at one week was an independent risk factor for recurrence.18 This suggests that elevated hs-CRP has a high sensitivity in addition to prognostic implications in the treatment of acute pericarditis and may be used to monitor progress or rule out the disease in those patients who may have equivocal clinical signs.
In a minority of cases of acute pericarditis where symptom resolution is prolonged, treatment with NSAIDs and colchicine may not be sufficient to overcome symptoms. Steroid use may be indicated as an adjunct or alternative treatment in patients who are unable to tolerate conventional treatment, have symptoms despite conventional treatment, or have an underlying autoimmune aetiology. Although corticosteroids are excellent in terms of expediting resolution of symptoms, there is a large body of evidence that steroid therapy in this context is an independent risk factor for recurrence and increased major side effects. This may be because most causes of pericarditis are of an infectious aetiology and steroids reduce the efficacy of the immune system to clear the pathogen.19 Some cases result in a a very difficult-to-treat steroid-dependent state. Strong evidence for the optimal dose of immunosuppressant is not available, therefore when steroid use is unavoidable, we can look to a retrospective study that examined the differences between highdose (1mg/kg/day) and low-dose steroids (0.2-0.5mg/kg/day) in the treatment of acute pericarditis.
The researchers found that only high doses were associated with recurrences, severe side effects and hospitalisations.20 Combined with the experience that low-dose steroids have been used effectively to treat serositis from autoimmune conditions, low-dose corticosteroids seem a reasonable initial strategy in the management of pericarditis requiring glucocorticoid therapy. Steroid therapy should continue for 2-4 weeks until symptom resolution and CRP normalisation. They should then be tapered over several months, while introducing colchicine, a NSAID and some form of gastro-protection. A safe tapering strategy for patients on more than 50mg/day is to start by reducing the dose by 10mg a day every 1-2 weeks. Once the dose is down to 50mg, reduce by 5-10mg a day every 1-2 weeks. Once the dose is down to 25mg a day, reduce by 2.5mg a day every 2-4 weeks. Once patients get down to 15mg a day after being on a highdose steroid, changes must be made very carefully and slowly. The dose may then be reduced by 1-2.5mg a day every 2-6 weeks, depending on how they tolerate the reduction. In those patients who are intolerant or do not respond to steroid therapy, the ESC guidelines recommend adding other immunosuppressive agents such as azathioprine or cyclophosphamide.9
In the event of frequent and highly symptomatic recurrences resistant to medical therapy or the development of constrictive pericarditis, pericardiectomy has been shown to be a relatively safe and effective option. It should only be undertaken for frequent and highly symptomatic recurrences resistant to medical therapy. However, all attempts should be made to try to control patients on a steroidfree regimen for one year before surgery, to reduce the incidence of wound complications post sternotomy or thoracotomy secondary to immunosuppression. Those patients with refractory symptoms despite steroid therapy should be referred to a specialist for ongoing management.
When the pericardium is damaged, there is a risk of developing scarring and fibrosis, which may subsequently lead to impaired diastolic filling of the ventricles. The patient usually presents with symptoms of heart failure such as dyspnoea on exertion and signs of fluid overload such as peripheral oedema. The haemodynamic consequences of losing the normal stretch of the pericardium largely results from the increase in ventricular interdependence between the left and right ventricles. Since there is a reduced and rigid space in the pericardium, as the right ventricle fills in diastole, it pushes the interventricular septum towards the left ventricle, reducing left ventricular end-diastolic volume, and thus cardiac output. Constrictive pericarditis can be very difficult to diagnose without using simultaneous invasive left as well as right heart catheterisation. Ultimately, the most effective treatment is a pericardiectomy. The risk of developing constrictive pericarditis depends on the aetiology of the pericardial damage. A study of acute pericarditis patients to determine the likelihood of developing constrictive pericarditis over a median of 72 months found the incidence of constrictive pericarditis was 1.8% overall.21 The number of cases per 1000 person-years by aetiology were: 0.76 for idiopathic or viral cases; 4.4 for connective tissue disease and pericardial injury syndrome cases; 6.33 for neoplastic; 31.65 for tuberculous; and 52.74 for purulent or bacterial (the most common).21 Some other important risk factors to the development of constrictive pericarditis are previous radiation to the chest (eg, for Hodgkin’s disease or breast cancer) and after cardiac surgery. The clinical course is usually progressive, but in a minority of cases, the constriction is transient. These are usually cases without signs of chronic inflammation and recurrent symptoms. One study used cardiac MRI to compare the thickness in the pericardium in those with persistent and transient constriction. Pericardial thickness of at least 3mm predicted the reversibility of constriction.22 In patients with haemodynamic stability, and little evidence of a chronic process, a trial of medical management for 2-3 months is recommended before surgical pericardiectomy.
Cardiac tamponade results from increased fluid in the pericardial space, and thus quickly begins to impinge on the myocardial chambers, and often results in significant interventricular dependence and haemodynamic compromise. In the setting of acute viral or idiopathic pericarditis, it is relatively rare to see the development of pericardial tamponade, with a rate of 14% reported in one series.23 Patients presenting with tamponade will likely develop this over a sub-acute course. Acute tamponade is usually the result of aortic dissections, or trauma to the chest. Sixty-one per cent of patients with neoplastic, tuberculous or purulent pericarditis were shown to develop tamponade in one series. This finding would suggest that patients presenting with tamponade should be investigated for a specific aetiology.23 Signs and symptoms of tamponade include sinus tachycardia, elevated central venous pressure, pulsus paradoxus, ECG changes (electrical alterans and low QRS amplitude are commonly observed), and arterial hypotension. Kussmaul’s sign (paradoxical rise in jugular venous pressure on inspiration) however, is not usually seen in tamponade unlike in constrictive pericarditis.
Evaluation using echocardiography is paramount in cases without overt hypotension and suspected tamponade as it helps to identify the haemodynamic significance. However, it has to be borne in mind that tamponade is a clinical diagnosis and thus echocardiographic signs consistent with tamponade (figure 6) are not by themselves an indication for pericardiocentesis. Tamponade that is haemodynamically stable can be managed by ensuring adequate hydration, vigilant monitoring and most importantly, therapy to address the underlying aetiology. The use of volume expanders was investigated for the conservative management of mild tamponade and found that fluid boluses of 500mL were most likely to increase the cardiac index (cardiac output per body surface area) when patients had a systolic blood pressure under 100mm/Hg.24 For patients becoming haemodynamically unstable, small bore catheter pericardiocentesis is the definitive option, as it is relatively safe, efficacious and can be used to obtain a sample of fluid for further analysis. This is the recommended treatment option endorsed by the European Society of Cardiology, especially if purulent or malignant pericarditis is suspected.9 Other options such as open surgical drainage or a video-assisted procedure may be needed for more complicated cases and for those who require a biopsy of the pericardium.
Pericardial disease in pregnancy is a precarious medical situation because of the potential harm pharmacological treatments pose to the growing fetus. A pericardial effusion can be detected in about 44% of pregnant women during the third trimester.25 The effusion is usually trivial to small in size, rarely requires treatment and usually subsides by about six weeks post-partum.25 The diagnosis of pericarditis is made the same way as it is in nonpregnant individuals, and is also most commonly attributed to a viral or idiopathic aetiology. In the first two trimesters, it is safe to treat the woman as you would any non-pregnant patient with NSAIDs while avoiding colchicine, as colchicine’s effect on the fetus is not well established. Although very scarce in the literature, one report included six pregnant women being treated for acute pericarditis.26
Four of the patients were treated with high-dose aspirin at 800mg three times a day and gradually tapered before the third trimester. One patient had aspirin throughout the pregnancy, and all patients were treated with steroids at 10-25mg a day either throughout the pregnancy or starting at the third trimester. Only the patient who was treated with aspirin throughout the pregnancy developed HELLP syndrome (haemolysis, elevated liver enzymes, and low platelets). Moreover, it is not recommended to continue NSAIDs past 20 weeks due to the potential to induce premature closure of the ductus arteriosus. It may thus be reasonable to give an NSAID in pregnancy up to the 20th week of gestation, while supplementing it with lowdose steroids throughout the pregnancy and lactation if needed, and gradually taper. These treatment decisions should only be made in consultation with a specialist.
Pericarditis may be the presenting symptom of systemic conditions including vasculitides, connective tissue disorders or granulomatous diseases. It is thought that systemic conditions are responsible for up to 7% of acute pericarditis and 10% of recurrent pericarditis.27 Suspicion for systemic conditions should be aroused when patients present with features of pericarditis and also have malaise, fever, weight loss, polyarthralgias, photosensitivity, rashes, dry eyes or mucosal ulcers. Ultimately, the diagnosis can only be made with certainty via pericardial biopsy or characteristic pericardial fluid composition. The treatment however, is similar regardless.
It is treated with NSAIDs combined with low-dose steroids (0.2-0.5mg/kg/day) if needed. The addition of colchicine may be beneficial to reduce recurrences as in viral or idiopathic aetiologies, but its efficacy has not been studied conclusively in pericarditis secondary to systemic diseases. Pericardial effusions are more common and more prominent in patients with systemic inflammatory diseases, in fact, over 50% of patients with SLE-associated pericarditis will have an accompanying effusion.27 However, three times as many patients with pericardial involvement due to rheumatological conditions are asymptomatic compared to those exhibiting the typical symptoms. Table 1 lists some other possible systemic aetiologies of pericardial inflammation.
Pericarditis is a relatively common clinical entity, most often due to a viral illness. The chest pain must be distinguished clinically from an MI, pericardial pain and other causes of pleuritic chest pain. Patients with low-risk features can be treated as outpatients. Patients should be evaluated using echocardiography to exclude pericardial effusion and tamponade. NSAIDs are used as the first-line therapy, however, growing evidence suggests that the addition of colchicines may be safe and efficacious in both preventing recurrence and reducing acute symptoms. Corticosteroid use is associated with increased rates of recurrence, but may be necessary for symptom relief and in cases where NSAIDs are ineffective or not tolerated. Asymptomatic pericardial effusions and pericarditis are a common manifestation of systemic inflammatory conditions and are treated using a similar strategy to viral aetiologies. Continued research is required to clearly define the safety and efficacy of routine colchicine use in the management of acute pericarditis for both viral and idiopathic causes and systemic inflammatory diseases. Therefore, based on current recommendations and available guidelines, we suggest the following approach for treatment of pericarditis.
Available on request from firstname.lastname@example.org
Dr Stefan Bucholz
Consultant cardiologist, Mackay Base Hospital, Cardiac Services Unit, Mackay; and associate professor, James Cook University School of Medicine, Mackay Campus, Queensland.
Dr Sameer Tatavarti
Resident medical officer, department of cardiology, Royal Brisbane Hospital, Brisbane, Queensland.