Eur J Cardiothorac Surg 2004;26:1156-1160
© 2004 Elsevier Science NL
‘Tamponade’ following cardiac surgery: terminology and echocardiography may both mislead
Susanna Price,
Jeremy Prout,
Siân I. Jaggar,
Derek G. Gibson,
John R. Pepper*
Department of Cardiothoracic Surgery, Royal Brompton and Harefield NHS Trust, Sydney Street, London SW3 6NP, UK
Received 18 July 2004;
received in revised form 17 August 2004;
accepted 18 August 2004.
* Corresponding author. Tel./fax: +44 207 351 8530. (E-mail: j.pepper{at}rbh.nthames.nhs.uk).
 |
Abstract
|
|---|
Objective: Echocardiography is widely considered the gold standard for the diagnosis of tamponade. While a relatively common complication of cardiac surgery in adults, determining whether haemodynamics are compromised by a pericardial collection early post-operatively can be difficult. The aim of the current study was to determine the nature and magnitude of the diagnostic challenge posed by cardiac tamponade following cardiac surgery. We therefore examined the accuracy of echocardiography in the diagnosis of tamponade in this patient group. Methods: From January 2000 to January 2002, 2297 adult patients underwent cardiac surgery in a tertiary referral cardiothoracic centre. A retrospective analysis of prospectively collected data, from all patients diagnosed with post-operative bleeding and/or tamponade was performed. Data included demographics, surgery, anticoagulation/anti-platelet medication, clinical/echocardiographic features of tamponade and surgical findings at re-exploration. Results: The diagnosis of ‘tamponade’ was confirmed at re-exploration in 148 patients. When it occurred early (<72h) following cardiac surgery trans-thoracic echocardiography failed to visualise the majority of collections (60%), necessitating trans-esophageal echocardiography. Effusions were small (160±17ml) and localised (92%), showing no echocardiographic features of classical tamponade (79%). Where patients developed tamponade late (>72h) following cardiac surgery, clinical features were atypical, effusions larger (640±71ml, P<0.0001)) and global (77%). Classical echocardiographic features of tamponade were usually present (70%) and readily visualised using trans-thoracic echocardiography. Conclusions: Haemodynamically significant pericardial collections occurring early following cardiac surgery rarely cause classical clinical or echocardiographic features of tamponade. Recognition of this as a separate diagnostic entity is necessary to ensure appropriate surgical intervention is not delayed.
|
1. Introduction
|
|---|
Pericardial effusion is a common complication of adult cardiac surgery, occurring in 50–64% of cases, and compromises cardiac function in 0.8–6% [1–3]. Accurate diagnosis of this potentially lethal, but easily treatable, complication is critical. Classically described echocardiographic features of tamponade include: right ventricular (RV) diastolic collapse, right atrial systolic collapse, a swinging heart and an enlarged non-pulsatile inferior vena cava. In addition there is a reciprocal variation in ventricular size, trans-mitral and trans-tricuspid flows with respiration leading to left heart ‘paradox’. The classical clinical features of pulsus paradoxus, tachycardia and oliguria are often absent [4].
Many of the echocardiographic features of classical tamponade depend upon equal transfer of intra-pericardial pressures to all chambers of the heart. This in turn depends upon the presence of a global effusion and the absence of significant left ventricular (LV) or right ventricular (RV) disease, or pulmonary hypertension [5]. The pressure–volume relationship of the pericardium is such that rapid accumulation of only small volumes may lead to significant localised increases in intra-pericardial pressure in the post-operative period [6]. This may be further confounded by the presence or absence of intact pericardium and pleura. Clearly therefore, it may not be safe to rely on classical echocardiographic features of tamponade to diagnose a haemodynamically significant pericardial collection in cardiac surgical patients. The aim of the current study was to determine the nature and magnitude of the diagnostic challenge posed by cardiac tamponade following cardiac surgery. We therefore studied the mode of diagnosis of ‘tamponade’ in the period immediately (<72h) and late (>72h) following cardiac surgery in a tertiary cardiac surgical unit.
|
2. Materials and methods
|
|---|
We undertook a retrospective review of prospectively collected data from all cardiac surgical patients at the Royal Brompton Hospital (January 2000–January 2002). Patients were selected from our computerised cardiovascular information registry covering adult intensive care and theatre records. Patients included were those with the diagnosis of tamponade, bleeding or pericardial effusion requiring surgical re-exploration. Exclusions were those who underwent non-cardiac surgery or implantation of ventricular assist device (VAD). During the period of the study, clinical protocols remained unchanged and there was no alteration in the senior surgical or anaesthetic consultant staff. In all patients the pericardium was left open. Patients were divided into two groups depending on when presentation occurred: early (<72h) and late (>72h) following initial surgery.
The diagnosis of a haemodynamically significant pericardial collection was confirmed by clinical suspicion (as documented in the patient records) and echocardiographic/CT findings, together with operative findings that confirmed the presence of an intra-pericardial collection ± clot, whose evacuation immediately improved the patient's haemodynamics (an increase in mean arterial pressure (MAP) by 20%±a reduction in right atrial pressure (RAP) by >5mmHg, where RAP was monitored). Clinical features of cardiac compromise recorded in the patient case notes included: hypotension, oliguria, worsening base deficit/acidosis, increasing RAP, chest pain, dyspnea, increasing tachycardia and arterial pulsus paradoxus. Echocardiograms (Hewlett Packard Sonos 2500 or 5500, USA) were performed as clinically indicated by the senior resident cardiologist. In accordance with our clinical practice, all patients initially underwent a trans-thoracic echocardiogram (TTE). Where TTE was negative, trans-oesophageal (TOE) echocardiography was performed. Echocardiographic findings were taken as those recorded in the patient notes and by review of the video recording made at the time of the diagnosis.
Echocardiographic features of tamponade sought in addition to the presence of a pericardial fluid collection included RA systolic collapse, RV diastolic collapse, a swinging heart, and an enlarged, non-pulsatile inferior vena cava. In addition, features that varied excessively with the respiratory cycle were noted, including changes in ventricular size, trans-mitral and trans-tricuspid flow velocities, and the presence of arterial paradox from aortic Doppler recordings. In all cases, where echocardiographic features were noted to be present or absent, the diagnostic approach (TTE/TOE) was documented. Data were analysed using GraphPad Instat 3 and expressed as mean±SEM. Student's t-test was used for comparison of continuous variables, where a P-value <0.05 was considered significant.
|
3. Results
|
|---|
A total of 2297 patients were operated, of which 148 (6.4%) were re-explored for post-operative bleeding/tamponade or pericardial effusion. Of these 148 patients, 20.2% had undergone urgent/emergency procedures. The surgery performed is shown in Table 1. Documented patient demographics included: age 59.4 (18–86 years), sex (male:female 4:1) and body mass index (BMI) 26.7 (18.–40.4kg/m2). A significant number took anti-platelet and/or anticoagulant therapy on the day of initial surgery. Thirty-two percent of those with a diagnosis of ischemic heart disease took aspirin, 17% heparin (fractionated or unfractionated) and 11% clopidogrel. Of the patients with valvular disease, 20% were anticoagulated with warfarin/heparin on the day of initial surgery. Six patients had repeated episodes; hence the total number of documented episodes was 154. Of these, 124 were early (<72h) and 30 were late (>72h) following initial surgery. All these patients survived to hospital discharge.
In patients presenting early following cardiac surgery many had multiple clinical features suggesting the need for an echocardiogram to exclude tamponade (Table 2). Of those returned to theatre early 74 were re-explored urgently (haemodynamic instability associated with excessive bleeding) and therefore had no echocardiogram performed. In patients undergoing echocardiography, the commonest combination of clinical features was hypotension and a worsening base deficit. In contrast, a rising RAP was relatively uncommon. In a small number of patients the sole presentation was oliguria/anuria. Trans-thoracic echocardiography failed to identify any pericardial collection in 60% of patients presenting in the first 72h following surgery (Table 3). Moreover, the classical echocardiographic features of tamponade were absent in 79%. At re-exploration, the volumes found in patients presenting early were small, <200ml in 85% of cases.
All patients who were re-explored for late tamponade underwent echocardiography, except one, where the diagnosis was made with CT scanning (Table 3). In patients presenting late, symptoms and signs were atypical and the diagnosis of pericardial effusion was usually made on routine post-operative echocardiography. The commonest associated features were chest pain and dyspnoea (Table 2). Hypotension or an elevated RAP was almost never observed in this patient group. In contrast to patients presenting early following cardiac surgery, the majority of those presenting late were found to have classical echocardiographic features of tamponade (Table 3). At re-exploration, the volume of pericardial collection found was significantly greater than in those patients presenting early following cardiac surgery (late 640ml±71, early 160ml±17, P<0.0001).
|
4. Discussion
|
|---|
Compromise of cardiac function by pericardial fluid, frequently referred to as cardiac tamponade, is a well-documented complication of cardiac surgery in adults. Previously described risk factors include advancing age, aortic root surgery, complex surgery or an episode of poor anticoagulant control in the perioperative period [1,3,7]. The diagnosis of cardiac tamponade depends upon a constellation of clinical signs, echocardiographic features and invasive pressure measurements. Each feature may be absent, but the diagnosis may still be entertained. The original description of the effects of cardiac compression due to haemorrhage into the pericardium was in the early 19th Century [8]. The classical explanation of cardiac tamponade denotes it as a single pathophysiological entity: the accumulation of fluid in the pericardial space results in an increase in the intra-pericardial pressure, which, when exceeding the ventricular end-diastolic pressures impairs ventricular filling, and so reduces stroke volume and cardiac output. The development of these features depends upon the presence of a global effusion, with equal transmission of the elevated pericardial pressure to all chambers of the heart, in the absence of other conditions (LV dysfunction, LV or RV hypertrophy and pulmonary hypertension). These classical features of cardiac tamponade have usually been described in patients with large effusions, often with an underlying medical cause. However, it is recognised that patients following cardiac surgery provide a much more specific diagnostic challenge, since effusions may be localised, underlying cardiac pathology present, and positive pressure ventilation used—all factors likely to alter the classical clinical and echocardiographic findings [9–12]. In these circumstances, waiting for the clinical and echocardiographic features of classical tamponade to develop may delay essential treatment and even threaten survival. For this reason we suggest using a more general term, such as ‘cardiac compromise by pericardial collection’ or ‘haemodynamically significant pericardial collection’ to refer to these cases, and tamponade only in those with the classical picture, usually occurring more than 72h after primary surgery.
The rapid accumulation of small volumes in the pericardial space results in significant elevation in intra-pericardial pressure that might be expected to have a profound effect on the filling and thus the function of any of the cardiac chambers, especially the low-pressure atrial and RV systems [13]. Indeed, isolated left and right atrial tamponade following cardiac surgery are well-recognised entities [12]. In our series, when this occurred early after cardiac surgery presentation was with the classical features of a low cardiac output state, whose true significance only became apparent when fluid/haematoma was demonstrated in the pericardial space. The volumes were small, and the collection localised, but their pathogenic significance became apparent with re-operation. Since the volumes were small TOE was required to demonstrate any collection in the majority of cases. The echocardiographic features of classical ‘tamponade’ were nearly always absent. To have awaited the appearance of echocardiographic features of tamponade would have led to inappropriate pharmacological and other treatment (inotropes, intra-aortic balloon pump, haemofiltration). Once recognised the correct treatment was straightforward—namely early re-exploration.
In contrast, the effusions occurring late following cardiac surgery were larger and displayed many of the classical echocardiographic features of tamponade (68% of cases). The fluid volume was much larger and the rate of accumulation presumably slower. Thus, where fluid accumulation/haematoma development occurs more slowly, the pericardium accommodates larger volumes with only a small rise in intra-pericardial pressure distributed globally. Over time, as intra-pericardial pressure rises, the echocardiographic features of tamponade develop. Of note, clinical features in these patients were atypical and did not necessarily suggest that an echocardiogram was indicated. Moreover although only one of the patients was hypotensive (even when compared with pre-operative blood pressure recordings) at the time tamponade was diagnosed, evacuation of the collection resulted in an immediate and significant rise in mean arterial pressure, together with complete resolution of all the echocardiographic abnormalities in all patients. Thus, late following cardiac surgery the threshold should be low for performing echocardiography in a patient presenting with any symptoms (however atypical) even if haemodynamically stable. The decision to drain any collection should then depend on the size of the collection as well as features of tamponade.
The study design (retrospective review of prospectively collected data) has inherent limitations. From this study, and other reports in the literature [2–4,9,10,12,15], it is clear that a prospectively designed study is required. This should include systematic physiological echocardiography (TTE and TOE), together with invasive pericardial pressure monitoring, to accurately determine echocardiographic features of haemodynamically significant pericardial collections in the perioperative period.
Echocardiography is widely held to be the gold standard in the diagnosis of tamponade [14]. Late following cardiac surgery, where patients present with atypical symptoms or an unexpected low-output state, the condition is often under-diagnosed [15] and TTE is pivotal to the diagnosis. In contrast, our data suggest that early following cardiac surgery echocardiography is often misleading. In addition, clinical and echocardiographic features are so different from those of classical tamponade, that the condition should be considered a separate diagnostic entity. In such cases the surgeon must be aware that demonstration of only a small pericardial collection (even if only on TOE) or the absence of any classical features of tamponade, does not exclude the diagnosis of a haemodynamically significant collection which requires prompt relief.
|
Acknowledgments
|
|---|
SP is funded by the British Heart Foundation as the Jill Dando Adult Congenital Heart Disease Fellow.
|
References
|
|---|
- Malouf JF, Alam S, Gharzeddine W, Stefadouros MA. The role of anticoagulation in the development of pericardial effusion and late tamponade after cardiac surgery. Eur Heart J 1993;14:1451-1457.[Abstract/Free Full Text]
- Pepi M, Muratori M, Barbier P, Doria E, Arena V, Berti M, Celeste F, Guazzi M, Tamborini G. Pericardial effusion after cardiac surgery: incidence, site, size, and haemodynamic consequences. Br Heart J 1994;72:327-331.[Abstract/Free Full Text]
- Alkhulaifi AM, Speechly-Dick ME, Swanton RH, Pattison CW, Pugsley WB. The incidence of significant pericardial effusion and tamponade following major aortic root surgery. J Cardiovasc Surg (Torino) 1996;37:385-389.[Medline]
- Tsang TS, Oh JK, Seward JB. Diagnosis and management of cardiac tamponade in the era of echocardiography. Clin Cardiol 1999;22:446-452.[Medline]
- Hoit BD, Shaw D. The paradoxical pulse in tamponade: mechanisms and echocardiographic correlates. Echocardiography 1994;11:466-487.
- Spodik DH. Acute cardiac tamponade: pathologic physiology, diagnosis and management. Prog Cardiovasc Dis 1967;32:64-66.
- Sellman M, Intonti MA, Ivert T. Reoperations for bleeding after coronary artery bypass procedures during 25 years. Eur J Cardiothorac Surg 1997;11:521-527.[Abstract]
- Morgagni JB. De Sedibus et causes Morborum, Lipsiae sumptibus Leopoldi Vossii, 1829. (Beck CS. Wounds of the Heart. The Technic of Suture). Arch Surg 1926;13:205-227.[Abstract/Free Full Text]
- Bommer WJ, Follette D, Pollock M, Arena F, Bognar M, Berkoff H. Tamponade in patients undergoing cardiac surgery: a clinical-echocardiographic diagnosis. Am Heart J 1995;130:1216-1223.[CrossRef][Medline]
- Russo AM, O'Connor WH, Waxman HL. Atypical presentations and echocardiographic findings in patients with cardiac tamponade occurring early and late after cardiac surgery. Chest 1993;30:1216-1223.
- Faehnrich JA, Noone Jr RB, White WD, Leone BJ, Hilton AK, Sreeram GM, Mark JB. Effects of positive-pressure ventilation, pericardial effusion, and cardiac tamponade on respiratory variation in transmitral flow velocities. J Cardiothorac Vasc Anesth 2003;17:45-50.[CrossRef][Medline]
- Ionescu A, Wilde P, Karsch KR. Localized pericardial tamponade: difficult echocardiographic diagnosis of a rare complication after cardiac surgery. J Am Soc Echocardiogr 2001;14:1220-1223.[CrossRef][Medline]
- Yamada E, Zhang Y, Davies R, Coddington W, Kerber RE. Phased-array intracardiac echocardiographic imaging of acute cardiovascular emergencies: experimental studies in dogs. J Am Soc Echocardiogr 2002;10:1309-1314.
- Tsang TS, Oh JK, Seward JB, Tajik AJ. Diagnostic value of echocardiography in cardiac tamponade. Herz 2000;25:734-740.[CrossRef][Medline]
- Meurin P, Weber H, Renaud N, Larrazet F, Tabet JY, Demolis P, Ben Driss A. Evolution of the postoperative pericardial effusion after day 15: the problem of the late tamponade. Chest 2004;125:2182-2187.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
S. J. Durham and J. P. Gold
Late Complications of Cardiac Surgery
Card. Surg. Adult,
January 1, 2008;
3(2008):
535 - 548.
[Full Text]
|
|
|
Top
Abstract
1. Introduction
