The initial management of patients with a known or suspected pericardial effusion is largely determined by clinical status. In the absence of hemodynamic instability or suspected purulent bacterial pericarditis, there is no need for emergent or urgent pericardiocentesis. Diagnostic pericardiocentesis may be performed to establish the etiology of an effusion, although only after thorough noninvasive workup is completed before consideration of an invasive procedure [1]. While the etiology of effusions varies widely in the literature depending upon patient population, a diagnosis based on initial examination alone was highly predictive of effusion etiology in one study [2]. In another large series of patients, between 50% and 60% of moderate to large effusions were due to a previously established medical condition [3]. In addition, the clinical context in which diagnostic pericardiocentesis is performed affects its predictive value, with greater diagnostic yield for large effusions than for acute pericarditis [4,5,6]. Primarily due to the routine use of echocardiographic guidance, the major (1.2%) and minor (3.5%) complications of pericardiocentesis have significantly decreased over the past several decades, with successful single needle passage rates approaching 90% and relief of tamponade in over 97% [7]. As a result, the 2004 European Society of Cardiology (ESC) recommends pericardiocentesis as the method of choice for pericardial fluid removal/sampling [8]. Surgical intervention is recommended for recurring large effusions for which repeated pericardiocentesis has not been effective, loculated or posterior effusions of hemodynamic consequence, purulent pericarditis, traumatic hemopericardium, constrictive pericarditis, and effusions due to aortic dissection [8]. Whenever possible, elective pericardiocentesis should be performed by an experienced operator using echocardiographic guidance. While generally safe, it should be performed in a location with adequate physiologic monitoring to assess any hemodynamic sequelae from complications and to aid in the diagnosis of effusive-constrictive pericarditis.

In contrast to diagnostic pericardiocentesis, the management of hemodynamically compromised patients requires emergent removal of pericardial fluid to restore adequate ventricular filling (preload) and hasten clinical stabilization. Aggressive fluid resuscitation and inotropic agents have been the mainstay of medical management for cardiac tamponade. These measures are largely ineffective and should be used only as a bridge to pericardial drainage [9,10]. The exact method and timing of pericardiocentesis is ultimately dictated by the patient’s overall degree of instability. While echocardiographic and fluoroscopic guidance is preferred, unguided (or blind) pericardiocentesis may be required in patients with severe hypotension not responsive to temporizing measures. In this setting, there are no absolute contraindications to the procedure, and it should be performed without delay at the patient’s bedside.

Urgent pericardiocentesis is indicated for patients with an established effusion who are initially hypotensive but respond quickly to hemodynamic support. Unlike acute tamponade, subacute tamponade is more likely to present with protean
symptoms such as dyspnea and fatigue. Patients with preexisting hypertension may not demonstrate severe hypotension due to a persistent sympathetic response. Echocardiographic assessment of effusion size, hemodynamic impact, and optimal percutaneous approach are of paramount importance [11]. The procedure should be performed within several hours of presentation while careful monitoring and support continue. As in elective circumstances, pericardiocentesis in these patients should be undertaken with appropriate visual guidance, the method of which depends on the physician’s expertise and resources.

Three additional points must be stressed regarding patients undergoing expedited pericardiocentesis. First, coagulation parameters—prothrombin time, partial thromboplastin time, and platelet count (> 50,000 per μL)—should be checked and, when possible, quickly normalized prior to the procedure. If clinically feasible, the procedure should be postponed until the international normalized ratio is less than 1.4. An anti-Xa level is recommended for patients receiving low-molecular-weight heparin. For emergent pericardiocentesis performed on anticoagulant therapy, prolonged and continuous drainage is recommended. Second, many critical care specialists and cardiologists advocate performance of all pericardiocentesis procedures in the catheterization laboratory with concomitant right heart pressure monitoring to document efficacy of the procedure and to exclude a constrictive element of pericardial disease, although excessive delays must be avoided (see Chapter 34). Finally, efforts to ensure a cooperative and stationary patient during the procedure greatly facilitate the performance, safety, and success of pericardiocentesis.

The clinical presentation of hemodynamically significant pericardial effusions varies widely among patients. A comprehensive understanding requires knowledge of normal pericardial anatomy and physiology.

The pericardium is a membranous structure with two layers: the visceral and parietal pericardium. The visceral pericardium is a monolayer of mesothelial cells adherent to the epicardial surface by a loose collection of small blood vessels, lymphatics, and connective tissue. The parietal pericardium is a relatively inelastic 2 mm dense outer network of collagen and elastin with an inner surface of mesothelial cells. It is invested around the great vessels and defines the shape of the pericardium, with attachments to the sternum, diaphragm, and anterior mediastinum while anchoring the heart in the thorax [12]. Posteriorly, the visceral epicardium is absent, with the parietal epicardium attached directly to the heart at the level of the vena cavae [13]. The potential space between the visceral and parietal mesothelial cell layers normally contains 15 to 50 mL of serous fluid, which is chemically similar to plasma ultrafiltrate, in the atrioventricular (AV) and interventricular grooves [14]. The pericardium is relatively avascular, but is well innervated and may produce significant pain with vagal responses during procedural manipulation or inflammation [15].

Because of the inelastic physical properties of the pericardium, the major determinant of when and how pericardial effusions come to clinical attention is directly related to the speed of accumulation. Effusions that collect rapidly (over minutes to hours) may cause hemodynamic compromise with volumes of 250 mL or less. These effusions are usually located posteriorly and are often difficult to detect without echocardiography or other imaging modalities such as multislice computed tomography or cardiac magnetic resonance imaging. In contrast, effusions developing slowly (over days to weeks) allow for dilation of the fibrous parietal membrane. Volumes of 2,000 mL or greater may accumulate without significant hemodynamic compromise. As a result, chronic effusions may present with symptoms such as cough, dyspnea, dysphagia, or early satiety owing to compression of adjacent thoracic structures. Conversely, intravascular hypovolemia, impaired ventricular systolic function, and ventricular hypertrophy with decreased elasticity of the myocardium (diastolic dysfunction) may exacerbate hemodynamic compromise without significant effusions present.

Since the first blind (or closed) pericardiocentesis performed in 1840 [16], numerous approaches to the pericardial space have been described. Marfan [17] performed the subcostal approach in 1911, which then became the standard approach for unguided pericardiocentesis as it is extrapleural and avoids the coronary and internal mammary arteries.

The advent of clinically applicable ultrasonography has opened a new chapter in diagnostic and therapeutic approaches to pericardial disease, allowing clinicians to quantitate and localize pericardial effusions quickly and noninvasively [18,19]. Callahan et al. [20,21] at the Mayo Clinic established the efficacy and safety of two-dimensional echocardiography to guide pericardiocentesis. While direct quantification of total fluid accumulation with echo is not yet possible, circumferential effusions of more than 10 mm are considered large (500 mL), and the ESC recommends pericardiocentesis of effusions of more than 20 mm, regardless of the presence of hemodynamic compromise (class IIa indication) [8]. Typically, at least 250 mL of fluid is required for safe pericardiocentesis. The routine use of echocardiography has resulted in two major trends in clinical practice: First, two-dimensional echocardiography is commonly used to guide pericardiocentesis, with success rates comparable to those of traditionally fluoroscopic-guided procedures [22,23,24]. Second, approaches other than the traditional subxiphoid method have been investigated owing to the ability to clearly define the anatomy (location and volume) of each patient’s effusion [20,21]. In one series of postsurgical patients, the subxiphoid approach was the most direct route in only 12% of effusions [25]. With the use of echo guidance, apical pericardiocentesis and parasternal pericardiocentesis are increasingly performed with success rates comparable to those of the subxiphoid approach. In the apical approach, the needle is directed parallel to the long axis of the heart toward the aortic valve. Parasternal pericardiocentesis is performed with needle insertion 1 cm lateral to the sternal edge to avoid internal mammary laceration. All approaches employ a Seldinger technique of over-the-wire catheter insertion. As the subxiphoid approach remains the standard of practice and is the preferred approach for unguided emergent pericardiocentesis, it will be described later.