Peter D. Wearden1 and Constantine Mavroudis2 1Nemours Children’s Hospital, Orlando, FL, USA 2Peyton Manning Children’s Hospital, Indianapolis, IN, USA Endocarditis in patients with congenital heart disease often can be traced to intermittent bacteremia due to dental caries, chronic infection, and intraoperative contamination. Turbulent flow associated with residual congenital lesions such as small ventricular septal defects, patent arterial ducts, and coarctation of the aorta causes endovascular damage that can attract and harbor cilia‐forming bacteria. In addition, implanted devices such as valve prostheses, vascular grafts, and transcatheter devices can form niduses for bacterial growth and proliferation. Depending on the patient’s clinical acuity, these individuals can be very difficult to manage owing to hemodynamic instability, active infection, and the complexity of procedural options. The principles of medical and surgical management center around appropriate antibiotic therapy, hemodynamic assessment, and operative planning. The technical operative options that will be necessary to treat the infection and restore hemodynamic stability to effect wellbeing and survival are emphasized. Infective endocarditis is the infection of the endothelial lining of the heart. In addition, the term is very commonly used to include infection of the patent arterial duct and the proximal great arteries and veins, although histologically there is no endocardium present in these structures. The condition commonly involves the cardiac valves, leading to valvar dysfunction. The process starts with damage to the endothelium. Fibrin and platelet deposition ensues, along with bacterial growth of the deposits to form clumps of exudative and proliferative material, termed vegetations. Further deposition of fibrin and platelets allows vegetation growth, isolating the microorganism from leukocytes and antibiotics in a biofilm. Only within the past decade have clinicians recognized the bacterial sessile mode of growth known as “biofilm” in the infectious process [1]. Biofilm is a surface‐attached microcolony of microbes encased in a self‐produced matrix of extracellular polymeric substances acting as a protective slime layer [1]. Microbial cells constitute only about 15% of the biofilm volume; the rest is extracellular polymeric substances [2]. Vegetations are responsible for the remote effects from micro and macro emboli. Diagnosis of infective endocarditis focuses mainly on the epidemiologic risk, culture results, and imaging. Most cases are successfully treated with prolonged antimicrobial management; however, surgical intervention may be necessary for complex or unresponsive cases. Infective endocarditis is less common in children than adults, accounting for ∼1 in 1300 to ∼1 in 3000 pediatric hospital admissions [3–6]. Previously, complications of rheumatic heart disease accounted for a significant proportion of endocarditis episodes in children, but recent studies estimate the contribution of rheumatic heart disease to infective endocarditis at less than 10% in the developed world [3, 7]. Surgical and catheter‐based interventional advances in the treatment of congenital heart disease have increased the risk for infective endocarditis. Surgical repair creates physiology within the heart that may predispose to endocarditis. Surgical techniques frequently require implantation of prosthetic materials, including conduits, shunts, and patches [8]. Nonsurgical repair with intracardiac devices, including coils, valves, and occluder devices, also continues to increase in frequency and can pose a potential risk. Finally, with improved survival secondary to advances in medical, interventional, and surgical management, the lifetime cumulative risk of infective endocarditis increases despite the potential decrease in risk related to surgical and nonsurgical interventions. In 2007, the American Heart Association revised its guidelines for the prevention of infective endocarditis [9], including risk stratification based on the underlying congenital heart defect. The new guidelines focus on providing prophylaxis based on the risk of adverse outcome with infective endocarditis rather than lifetime risk of infective endocarditis acquisition. Conditions considered to be high risk for adverse outcomes related to infective endocarditis include unrepaired or palliated cyanotic congenital heart defects, repaired congenital heart defects with residual defects adjacent to prosthetic material, and completely repaired congenital heart defects with implanted device or prosthetic material for the first six months after repair. Heart transplant recipients with cardiac valvulopathy and patients with prosthetic valves are also categorized as high risk. Many common congenital heart defects (e.g., patent arterial duct, ventricular septal defect) have increased risk of infective endocarditis, yet prophylaxis is no longer recommended due to the lack of efficacy data. Patients who are more than six months after their repair with no residual defects are considered to have the same risk as the general population. Review of this policy change found no significant change in infective endocarditis hospitalization rates associated with revised prophylaxis guidelines over 11 years across 29 US children’s hospitals [10]. The incidence of infective endocarditis in patients without congenital heart disease has also increased. Immunosuppression secondary to malignancy and transplantation with or without the use of long‐term indwelling catheters are risk factors for infective endocarditis. The incidence of infective endocarditis in neonates is also rising. Advances in the care of premature and other critically ill neonates, which have substantially improved neonatal survival, rely on the prolonged use of arterial and venous indwelling catheters, which have been associated with an increased risk of infective endocarditis even in the absence of congenital heart disease [10–13]. The primary event leading to the development of infective endocarditis is damage to the epithelial lining of the heart, followed by bacteremia with organisms capable of adhering to the damaged endothelial tissue. Several dynamics create the potential for endothelial damage. Turbulent blood flow from high‐ to low‐pressure areas creates trauma to the endothelium. Injury may also be mediated by intracardiac catheters and intracardiac surgical or transcatheter repair. Damage to the endothelial lining of the heart allows the formation of nonbacterial thrombotic endocarditis (NBTE), which results from the deposition of fibrin and platelets on the damaged endothelium. Prosthetic materials including shunts, patches, valve, conduits, and intracardiac catheters provide additional sites for the deposition of fibrin and platelets. Once vegetations form, transiently circulating bacteria adhere to the fibrin–platelet aggregate to propagate the infection. Transient bacteremia occurs frequently with daily activities including brushing teeth or chewing food, which cause local trauma to mucosal surfaces. Additional episodes of bacteremia can occur with dental or selected surgical procedures. The ability of bacteria to adhere to the sterile vegetation influences the risk of infective endocarditis after transient bacteremia. Gram‐positive bacteria have specific factors that promote their ability to adhere to fibrin and platelets. Staphylococcal species have adhesin proteins and some viridans streptococci have FimA protein that promote adherence [14–16]. When the bacteria attach to the platelet–fibrin aggregate they multiply. Once the bacteria embed in the vegetation, they become sequestered from the circulating host immune defenses, avoiding clearance of the infection by forming a biofilm. Additional layers of platelet and fibrin deposit over the bacteria. Prolonged antibiotic therapy may not eradicate the infection because the biofilm acts as a protective slime layer. The array of organisms identified as causative agents in infective endocarditis continues to grow, especially with the availability of molecular diagnostic techniques. However, Gram‐positive cocci with their identified adherence factors remain the dominant causative organisms of infective endocarditis in children [3, 4, 6]. Streptococci have previously been reported as the most common organisms in pediatric infective endocarditis, with viridans streptococci reported most frequently. However, in the most recent reports, Staphylococcus aureus has emerged as the most common overall cause of infective endocarditis in children [3, 8, 11]. Although it is less common than Staphylococcus aureus, coagulase‐negative staphylococcus is also a significant contributor. Other organisms including Gram‐negative bacteria, Candida species, and fungi are rare, but may pose increased risk to children without underlying structural heart defects, including neonates and immunocompromised hosts. Endocarditis related to prosthetic valves or other implanted foreign material requires special consideration. For children with congenital heart disease, more than 50% with infective endocarditis have had previous corrective or palliative surgery. Both early and late postsurgical infections have been reported. Early infections, within a few months of surgery, may be related to surgical contamination or invasive postsurgical monitoring. Staphylococcus aureus and coagulase‐negative staphylococcus are commonly recovered early after interventions [3, 11]. In addition, less commonly reported organisms include Klebsiella, Pseudomonas aeruginosa, diphtheroids, and Candida. Reports of infective endocarditis from anaerobic bacteria in pediatrics are rare [17]. The organisms recovered more than one year after surgical intervention reflect the organisms recovered prior to or without surgery. In a significant proportion of cases, an organism is not identified despite aggressive evaluation. Culture‐negative endocarditis occurs in 5–20% of cases of infective endocarditis [6, 7, 18]. Combined information from clinical, laboratory, and imaging areas assists in the diagnosis of infective endocarditis. The clinical manifestations of infective endocarditis are variable, but include systemic symptoms relating to persistent bacteremia. The presentation for persistent bacteremia ranges from simple fever with malaise to fulminant sepsis with multiorgan system failure. Fever is the most common symptom in patients outside of the neonatal age. Fever and other symptoms may be subtle, delaying the diagnosis [5, 7]. In the absence of fever, infective endocarditis should be considered in patients with signs and symptoms of a systemic illness, especially for neonates who may not mount an appropriate fever response. Additional nonspecific systemic symptoms include malaise, anorexia, weight loss, and arthralgia. Cardiac manifestations of infective endocarditis also vary and are dependent on the site of infection within the heart. Valvar lesions may lead to regurgitation or the perforation of the valve leaflet. Subvalvar lesions can cause rupture of the chordae tendineae or papillary muscles, with subsequent severe valvar dysfunction. Lesions near the valve annulus may invade the myocardium, resulting in cardiac conduction disturbances. Additionally, annular abscess and valve rupture occur with infection at the annulus. Large vegetations may produce mechanical obstruction mimicking valvar stenosis. Clinical manifestations of cardiac dysfunction include congestive heart failure, tachycardia, and arrhythmia. Functional changes in murmur quality occur in the patient with underlying congenital heart disease. Fever without a source in the presence of underlying valvar or congenital heart disease should prompt investigation for infective endocarditis. Embolic phenomena depend greatly on the site of infection as well. Areas of embolization primarily include the lung and brain. However, embolization to kidney, spleen, liver, skin, and coronary arteries has been reported [5, 7, 19]. Pulmonary embolization principally occurs with right‐sided vegetations that enter the pulmonary circulation, resulting in infarction and lung abscess. Early in its presentation, infective endocarditis may masquerade as pneumonia after unrealized pulmonary embolization. Neurologic manifestations complicate a significant number of cases. Involvement likely results from embolic events and can include isolated papilledema, altered sensorium, paresis, seizures, cranial nerve palsies, stroke, and coma [7, 20]. Mycotic aneurysms may develop after septic embolization from a vegetation, leading to rupture and death in rare cases. Infections with Staphylococcus aureus are associated with increased risk of neurologic complications with infective endocarditis [21]. Renal manifestations including glomerulonephritis are also fairly common with infective endocarditis. The etiology for renal involvement is multifactorial, including the deposition of immune complexes within the kidney, hypoperfusion secondary to congestive heart failure and sepsis, or deposition of septic emboli. Clinical evaluation for gross and microscopic hematuria should be considered. Skin manifestations can be from either emboli or deposition of immune complexes and are relatively rare in children. Janeway lesions, which present as flat, painless lesions on the palms and soles, result from hemorrhage around microabscesses from septic emboli. Alternatively, Osler nodes, which present as painful red lesions, develop after immune‐mediated infiltration around the dermal vessels. Roth spots found on the retina and subungual splinter hemorrhages on the digits may also occur. Diagnosis of infective endocarditis relies primarily on recovery of a pathogen with a positive blood culture. When a pathologic diagnosis is not made from positive histology or culture from the vegetation, positive blood culture with a causative pathogen is one of the major conditions in the modified Duke criteria (Table 45.1) for infective endocarditis diagnosis [22, 23]. Increased numbers and volumes of blood culture improve the yield of organisms in the diagnosis of infective endocarditis. Generally, three or more individual cultures are sent in a 24‐hour period aiming to meet the Duke criteria for infective endocarditis. Blood volumes for pediatric patients may not meet volumes recommended for adults (10–20 cc of blood per culture) due to restrictions in specimen collection, but volumes of at least 3 mL per sample are generally acceptable in small children, although smaller volumes may be collected in neonates. To avoid confusion with skin flora contamination, sterile preparation at the time of sample collection is essential. Administration of antibiotics prior to specimen collection decreases the yield of positive blood cultures by up to 40% [24]. Several additional techniques to improve pathogen recovery exist. Antimicrobial absorbent resins improve bacterial recovery in patients exposed to antibiotics before cultures are obtained [25]. Prolonged incubation of blood cultures for up to two weeks is also recommended to evaluate for fastidious organisms. These organisms include the HACEK organisms (Haemophilus, Actinobacillus, Cardiobacterium, Eikenella corrodens, Kingella kingae) and nutritionally deficient Streptococcus. Emerging molecular technologies have the potential to further improve organism identification, potentially limiting the diagnosis of culture‐negative endocarditis, both in patients who received antibiotics before cultures were obtained and also in patients with endocarditis attributed to organisms that are extremely difficult to grow in conventional culture. These techniques have improved recovery of organisms in adult and pediatric patients, but require further standardization and are not considered a major criterion in the Duke classification [26–28]. Availability of molecular diagnostics and other techniques to improve recovery of pathogens should be discussed with specialists in infectious diseases and microbiologists in complex cases. Table 45.1 Duke Criteria and Modified Duke Criteria for infective endocarditis* * See [22] for exceptions to criteria. Source: Adapted from Bonow RO et al. Circulation. 2006;114:e84–e231; and Jault F et al. Ann Thorac Surg. 1997;63:1737–1741. Additional laboratory findings support the diagnosis of infective endocarditis and may include elevated erythrocyte sedimentation rate, C‐reactive protein, and positive rheumatoid factor. Decreased serum complement and hematuria from glomerulonephritis are immunologic phenomena that may be present; glomerulonephritis is one of the minor criteria for infective endocarditis. Leukocytosis may be present; however, severe sepsis from infective endocarditis may present with leukopenia. Additionally, anemia from chronic illness can be masked by relative polycythemia secondary to cyanotic congenital heart disease. Imaging is the mainstay of defining the intracardiac disease in infective endocarditis. It is key in guiding therapy and prognosis. Findings of vegetations, intracardiac abscess, new partial dehiscence of a prosthetic valve, or new valvar regurgitation are included as major criteria for the diagnosis of infective endocarditis [23]. Furthermore, progression of disease and valvar dysfunction can also be followed echocardiographically. Transthoracic echocardiogram (TTE) has a poor sensitivity, but is usually the first test performed [29]. Transesophageal echocardiogram (TEE) typically provides more detailed information [30, 31], especially when small vegetations and intramyocardial abscesses are suspected. Reynolds et al. [31] found the sensitivity of TTE in detecting vegetations to be only 55% when compared to TEE. Li et al. [23] found TEE to be positive in 19% of cases where TTE was negative. Limitations of echocardiographic imaging include the difficulty in distinguishing old healed vegetations from active infection. Also, it is difficult to differentiate small vegetations from uninfected clots related to suture material and pledgets. Interference from prosthetic materials, such as valves and patches, makes it difficult to look for vegetations around these structures with echocardiography. Repeat studies are very helpful to evaluate the progression of disease. Other imaging modalities may be utilized to assess infective endocarditis. Magnetic resonance imaging (MRI) can occasionally be helpful in assessing ventricular and valvar function. Its value along with contrast‐enhanced computed tomography (CT) scanning is more for detection of remote emboli, and distinguishing between infarction, hemorrhage, and abscess. Interestingly, a recent study showed that CT scan was more accurate than TEE in detecting the perivalvar extent of endocarditis [32]. Fluoroscopy may play a role in detecting valvar dysfunction in prosthetic valves by demonstrating impaired disc excursion. Angiography is rarely needed, except in detection of aneurysms in the affected area, especially intracranial mycotic aneurysms. Coronary angiography is occasionally performed in children with unexplained ventricular dysfunction. Treatment of infective endocarditis depends on prolonged antimicrobial therapy with or without surgical intervention. The prolonged duration of antibiotics recommended is a function of the relatively slow metabolism of the organisms in the vegetation, as well as the structure of the vegetation with the pathogenic organisms covered by biofilm. Duration of antimicrobial therapy depends on the organism recovered and duration of therapy should begin from the date of the first negative culture. Antibiotics should ideally be tailored to the organism, but empiric therapy after multiple cultures are obtained can be directed at the most likely causative organisms pending results, especially in the case of a critically ill patient. Generally, treatment recommendations for duration range from 4 weeks to 6 weeks or longer depending on the clinical scenario, with longer durations for patients with implanted prosthetic materials. Empiric therapy for presumed viridans Streptococcus relies on penicillin G or ceftriaxone, while penicillinase‐resistant penicillin (e.g., nafcillin, oxacillin) should be used for possible staphylococcal endocarditis. For penicillin‐allergic patients and in areas with a high prevalence of methicillin‐resistant Staphylococcus aureus, vancomycin is recommended. The use of an aminoglycoside for synergy is considered optional in the most recent recommendations from the American Heart Association (AHA) secondary to reports of renal toxicity without increased benefit with aminoglycoside use, although some experts recommend its use in fulminant cases [33–35]. Aminoglycosides are recommended for two weeks along with rifampin for the duration of therapy, in addition to the primary antimicrobial, in patients with staphylococcal infections who have prosthetic materials. Empiric therapy for culture‐negative endocarditis depends on the clinical course and epidemiologic aspects including pretreatment with antimicrobial agents, and presence of and time since placement of prosthetic materials. Current recommendations include ampicillin‐sulbactam plus an aminoglycoside for six weeks for native valve infective endocarditis. For patients with prosthetic material, vancomycin is recommended, with the addition of a broad‐spectrum cephalosporin if infective endocarditis occurs within two months of valve placement [33]. Early surgical intervention should be considered if the patient is unresponsive to empiric therapy. Fungal endocarditis is primarily caused by Candida species, with Aspergillus species occurring less frequently [36]. Morbidity and mortality related to fungal endocarditis are high [11]. Early surgical intervention is suggested by many experts and medical therapy with an amphotericin B‐containing product is suggested for at least six weeks. Traditionally surgery was reserved as a means of last resort. However, early surgery is now recommended with increasing frequency [37–39]. The goal of surgery is removing the infected material, improving valvar function, repairing anatomic complications, and decreasing mortality [18, 39–41]. Data and recommendations specific for children are not available separately, but AHA guidelines [42] recommend surgical intervention in native valve endocarditis in the presence of heart failure (especially related to valvar dysfunction), worsening valvar function, recurrent emboli, evidence of intramyocardial infection, or infections from resistant organisms (see Table 45.2). The decision to proceed to surgery is frequently subject to much discussion and controversy. In the absence of specific recommendations for children, it is important to take all factors into consideration, including the clinical status of the patient, chances of recovery from a neurologic event, the microorganism involved, prospects of valve repair or replacement, and the AHA guidelines [42]. Also, it is recommended that these factors be assessed as a continuum rather than only at a single point in the clinical course of a patient. Table 45.2 Indications for consideration of surgery in infective endocarditis This is perhaps the most important consideration in surgical therapy, as it is the most common predictor of mortality [43]. Congestive heart failure in infective endocarditis is frequently related to the added burden of sepsis on the underlying valvar lesion. For example, a patient with pre‐existing aortic stenosis without heart failure may decompensate into congestive heart failure due to associated fever, anemia, and sepsis. Under such circumstances, correction of these associated factors may improve congestive heart failure. Heart failure from acute valvar dysfunction, destructive intracardiac lesions, and uncorrected shunting lesions is more likely to need surgical intervention. Early surgery for endocarditis is associated with reduced mortality in patients with congestive heart failure when compared with medical treatment alone [39]. Stable patients with valvar dysfunction and no other complications can frequently be treated with antibiotics alone until the acute infection has cleared. Subsequently, patients can be assessed for intervention if the usual surgical indications exist for valvar dysfunction. However, acute valvar dysfunction from leaflet perforation or valve stenosis due to vegetations is frequently associated with congestive heart failure and is unlikely to improve with antibiotics alone. Renal and other organ dysfunction is commonly present. Surgery in this group of patients can frequently improve cardiac function by treating valvar dysfunction. In mitral valve endocarditis, early surgery makes repair more likely than valve replacement [42]. Intracardiac complications of fistula and abscess formation have a high mortality with medical treatment alone. They are frequently accompanied by congestive heart failure. Periannular or myocardial abscess may be a precursor of a mechanical complication. New‐onset atrioventricular block is indicative of myocardial infection. These are indications to proceed with surgery even without significant congestive heart failure [42]. Size of vegetations alone is difficult to interpret in small children. Whereas a 5 mm vegetation on the tricuspid valve in an adolescent may not be significant, a similar‐sized vegetation on the mitral valve of a newborn may be quite significant. Furthermore, a small embolus in a child may result in ischemia of a large anatomic vascular bed. In adults with infective endocarditis, a vegetation larger than 10 mm is felt to be a risk factor for remote emboli, especially stroke [44]. Similarly, patients with mobile vegetations, or vegetations that increase in size while on adequate therapy, are at risk of remote emboli. Such factors should be taken into consideration to determine whether surgical intervention is indicated. Embolic stroke frequently leads to cerebral abscess formation or intracranial mycotic aneurysm formation. Occurrence of two or more episodes of major embolism during adequate medical treatment is commonly considered an indication for surgery. However, the presence of large vegetations with a single episode or a new stroke, especially during adequate therapy, is a reasonable indication for surgical intervention. Mortality of infective endocarditis with a cerebral embolus is higher [39, 44]. Some studies report Staphylococcus aureus and Candida endocarditis to be more commonly associated with embolic stroke [44–46]. At least half the strokes in adults are not apparent clinically and are discovered by MRI [45]. Presence of a recent stroke frequently is a concern for use of cardiopulmonary bypass due to the possibility of intracranial hemorrhage. As a result, many surgeons prefer to wait a few days or up to three weeks if possible before proceeding to surgery [47]. However, a low incidence of postoperative stroke has been reported in adults in the absence of a hemorrhagic infarction [45, 48]. Persistent fever or bacteremia, failure of the indices of infection to resolve, or progression of echocardiographic evidence of disease are evidence of failed medical therapy. Organisms that are frequently resistant to antibiotics or failure to clear with antibiotics alone, such as methicillin‐resistant Staphylococcus aureus, fungal endocarditis, or Gram‐negative endocarditis, especially with a prosthetic valve, fall into this category [42, 46]. In cases of culture‐negative endocarditis, a detailed examination of all parameters of the disease needs to be considered, as organisms resistant to commonly used therapy may be the causative agents. Progression of anatomic disease such as large or growing vegetations on echocardiographic study, evidence of emboli, or persistent indices of inflammation all point to failure of medical therapy. Culture‐negative endocarditis has a higher risk of complications and need for surgery [49]. Intracardiac prosthetic materials are not well endothelialized within the first two months postoperatively. Bacteremia early after surgery may be more likely to cause infection of the prosthesis, necessitating its removal. Furthermore, endocarditis on a recently placed prosthetic valve is more likely to lead to annular abscess. Hence, the presence of even mild heart failure and early staphylococcal endocarditis of a prosthetic valve is sufficient to warrant surgery [42]. Indications for surgery in late prosthetic valve endocarditis are similar to the ones discussed above. Timing of surgery in infective endocarditis should not be based on duration of preoperative antibiotics alone, but rather on the clinical status of the patient. In stable patients, without progressive valve destruction or other complications, a complete course of antibiotics followed by re‐evaluation for possible need for surgical therapy is appropriate. Very unstable patients will frequently require a period of resuscitation in the intensive care unit prior to surgery. However, in the presence of severe sepsis, or severe congestive heart failure, prolonged delay prior to surgery for controlling sepsis is unnecessary. Patients with cerebral infarction are not necessarily at a higher risk of postoperative stroke [38, 48]. However, if possible, it is prudent to wait 2–3 weeks after the stroke before proceeding to cardiac surgery [48]
CHAPTER 45
Endocarditis in Patients with Congenital Heart Disease
Definition
Epidemiology
Pathophysiology
Microbiology
Diagnosis
Clinical Manifestations
Laboratory Investigations
Duke Criteria (1994)
Pathologic criteria – If either is positive, diagnosis is definite
Demonstrated by culture or histologic examination of a vegetation, a vegetation that has embolized, or an intracardiac abscess specimen
Vegetation or intracardiac abscess confirmed by histologic examination showing active endocarditis
Major clinical criteria – If both are positive, diagnosis is definite
Typical microorganisms consistent with infective endocarditis from two separate blood cultures, microorganisms consistent with infective endocarditis from persistently positive blood cultures, single positive blood culture for Coxiella burnetii or antiphase I immunoglobulin G antibody titer >1 : 800
Echocardiogram positive for infective endocarditis, abscess, new partial dehiscence of prosthetic valve, new valvular regurgitation. Note: worsening or changing of pre‐existing murmur not sufficient
Minor clinical criteria – If all are positive, diagnosis is definite
Major arterial emboli, septic pulmonary infarcts, mycotic aneurysm, intracranial hemorrhage, conjunctival hemorrhages, and Janeway’s lesions
Glomerulonephritis, Osler’s nodes, Roth’s spots, and rheumatoid factor
Positive blood culture but does not meet a major criterion as noted above, or serologic evidence of active infection with organism consistent with infective endocarditis
Modified Duke Criteria (2000)
Imaging Studies
Antimicrobial Therapy
Indications for Surgery
Congestive Heart Failure
Valvar Dysfunction
Mechanical Intracardiac Complications
Remote Embolic Complications
Failure of Adequate Medical Therapy
Culture‐Negative Endocarditis
Prosthetic Valve/Device Endocarditis
Timing of Surgery
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
