It was only half a century ago that cardiac percussion was considered by many noted cardiologists of the day to be a useful adjunct in the clinical evaluation of the patient with cardiac disease. Since then, the advent of improved imaging techniques such as echocardiography has resulted in a general consensus that the practice is outdated. It is, however, our contention that the loss of this skill is testimony to a trend away from clinical art and towards clinical science. Though to a degree this trend has brought improvements in practice, revealed by evidence-basing, for some of us at least who are part of this great tradition, medicine will always be both an art and a science. Aside from such romantic debate, we aim to make evident the usefulness of cardiac percussion to current medical practice. The discussion is based on personal use of the technique, with some speculation on potential uses not within our experience. The argument against percussion of the heart is that with roentnographic imaging defining size and borders and with echocardiography being a more definitive gauge of the cardiac indices, there is no value to it. By the same argument there is no reason to auscultate the heart. The point is that such techniques direct us to the need for further characterization by more informative methods. Having said all this, we would agree that cardiac percussion for the purpose of determining the cardiac borders is prone to error. Indeed, as early as 1899 Williams1 noted the errors of ordinary percussion as opposed to roentnographic determination for this purpose, particularly in small hearts. What we propose is the elucidation of abnormal areas of dullness, whose determination allow us to obtain quick information at the bedside. Indeed, in his popular A Primer of Medicine,2 and based on work by Dressler, Papworth alludes to at least fifteen cardiac pathologies that may be diagnosed on the basis of cardiac percussion. By dullness in this instance we mean an extremely flat note. Of course percussion around the praecordium will inevitably reveal a dull note, but an element of resonance is always noted. In the healthy individual such an area of marked dullness is normally found adjacent to the sternal margin, beneath the fourth rib on the left. This dull area is approximately two inches wide and merges with hepatic dullness at the level of the sixth rib caudally. Enlargement of the heart results in more proximate apposition of its chambers to the chest wall, so causing a flatter percussion note above this vicinity and extension of the normal area of marked dullness. As the anterior aspect of the heart is primarily right ventricle, it is right heart dilation that is most readily amenable to detection by this method. It is characterized by extension of marked dullness to the left and right of the normally dull area, such that there is substernal dullness and left-sided flattening extending more than the two inches described. Areas of abnormality should be compared to the normal area of dullness to judge the tone as being definitively dull. There are three useful diagnoses we can make with this technique: Pulmonary hypertension/left atrial enlargement; Right atrial enlargement; and Right ventricular hypertrophy/dilation. Chronic pulmonary hypertension and left atrial enlargement, as occur in long-standing mitral stenosis, may also be detected using the percussion method.3-5 Here, dilation of the right ventricular outflow tract is evidenced by marked dullness in the third left interspace, extending as much as three inches in this direction in severe cases. This is in effect the clinical equivalent of the roentnographic ‘straight left heart-border.’ Enlargement of the right atrium, as in tricuspid incompetence,6,7 is intimated by an area of marked dullness extending to the fifth intercostal space and sixth rib on the right of the sternum. Where such dilation is severe this area of dullness can extend upward to the fourth interspace on the right. Right ventricular dilation causes dullness over the sternum at the level of the fifth and sixth ribs, and extends to the left a few inches from the fourth intercostal space to the sixth. It thus encompasses the normal area of dullness and a substernal component. Note, however, that the difference between right ventricular hypertrophy and dilation cannot be appreciated as both present similarly owing to overlap with the normal area of dullness. The above-mentioned features of extended areas of dullness were posited by Dressler3,4 to be particularly useful in determining cases of mitral stenosis and tricuspid regurgitation. However, perhaps the greatest value of the cardiac percussion method lies in appreciation of the more diffuse pattern of marked dullness that accompanies pericardial effusion. Here, the finding of a flat percussion note over the lower half or two-thirds of the sternum in the absence of evidence for mitral stenosis is suspicious of pericardial effusion. In moderate effusions dullness often extends to both sides of the sternum, especially the left side where the second and third interspaces are involved. In our experience, however, we have discovered that the extremely dilated hearts of chronic mitral stenosis, dilated cardiomyopathy and heart failure can mimic this picture, so the clinical context is important. This method can clinically differentiate significant pericardial effusion4 from the differentials of restrictive cardiomyopathy and constrictive pericarditis, as the latter do not have the same pattern of dullness—indeed, one of the authors has used it to make the diagnosis, subsequently confirmed on echocardiography. In this diagnosis, it can be confounded by gross dilation due to congestive heart failure which, if significant, can mimic the findings of effusion, and the authors have experienced several cases where pericardial effusion has been queried on Dressler's criteria but all that was discovered on echocardiography was global chamber dilation and function consistent with congestive impairment. Such a finding in a patient without previous cardiac symptoms but with cardiac risk factors and no valvular disease ought to suggest ischaemic cardiomyopathy, particularly if there is no corroborative evidence of effusion on electrocardiography, such as small complexes or electrical alternans. Though asymptomatic, such patients may have significant coronary artery disease with an attendant increased risk of event and ought to receive coronary angiography. In addition we suggest that in the emergent pericardial tamponade with haemodynamic instability, the finding of small electrical complexes is not always present. Rapid percussion showing dullness in three areas (to the left of, directly above, and to the right of the sternum) should suggest the diagnosis, to be confirmed by echocardiography. The finding may thus differentiate tamponade from acute massive pulmonary embolism, which presents with isolated signs of pulmonary hypertension. In the case of dual pathology, in the form of dilated cardiomyopathy and acute massive pulmonary thromboembolism, differentiation is possible from life-threatening tamponade by the finding of dullness of the upper third of the sternum in the latter. This said, as the method is quick it should not delay further definitive imaging. It is also the finding of the authors, albeit anecdotally, that the method may be particularly accurate in patients with obstructive airways disease, where barrel-chest deformity would normally lead to resonance in all relevant areas of the praecordium. Findings of dullness are therefore more specific in these patients. In particular, the method may be useful in diagnosing cor pulmonale (with right ventricular hypertrophy and pulmonary hypertension), and one author has diagnosed both this and pericardial effusion in patients with emphysema using area-based dullness. Similarly in obese patients the diagnoses of obstructive sleep apnoea/obesity hypoventilation can be suspected at an earlier stage if cardiac percussion is used in admission assessments, particularly as jugular venous distention can be difficult to visualize in these individuals and percussion signs are generally relatively easier to appreciate in them. One of the authors has inferred a diagnosis of obstructive sleep apnoea based on symptoms of daytime sleepiness, snoring and the cardiac percussion findings of right ventricular hypertrophy. Nocturnal oximetry confirmed the diagnosis. Another less common instance where the technique may be useful is in the diagnosis of calcific mitral stenosis, where first heart-sound and opening snap are diminuted and the murmur may be hard to appreciate. On a speculative note, albeit intuitive, young patients with presenting features of dizziness/syncope/arrhythmia and percussion signs of isolated right ventricular dilation without pulmonary hypertension may fit with the diagnosis of arrhythmogenic right ventricular dysplasia and perhaps should be put forward for further investigation. All this being said, there were some false-positives associated with the technique used in isolation when subsequent echocardiography was undertaken; however, when used in clinical context with other findings we have achieved many notable diagnostic successes, some of which have been mentioned above. In conclusion, therefore, we propose that this abnormal dullness area-based approach to cardiac percussion, first outlined by Dressler himself over half a century ago, is an invaluable tool in a subset of cardiorespiratory diagnoses in our experience, and thus we propose that its use in modern clinical practice should be encouraged and further investigated.