INTRODUCTION AND EPIDEMIOLOGY
The clinical presentation of endocarditis is often nonspecific and variable, with potential to affect nearly every organ system in an indolent or fulminant course. The vast majority of endocarditis is infective; diagnosis relies on a set of explicit criteria, which include findings from blood culture, echocardiography, and close clinical observation. Unrecognized infective endocarditis has frequent complications and high mortality.
In developed countries, the incidence of infective endocarditis ranges from 2 to 11.6 cases per 100,000 patient-years1,2,3,4,5,6,7 and is higher in urban versus rural settings, likely reflecting the impact of injection drug use. The disease is uncommon among children, where it is associated with structural congenital heart disease, rheumatic heart disease, or nosocomial, catheter-related bacteremia. The disorder affects men more commonly than women, and the hospital mortality rate is up to 18%, varying according to the microorganism involved and presence of complications.1,2
Most cases occur either in those with a predisposing identifiable cardiac structural abnormality (congenital or acquired), prosthetic valve, or a recognized risk factor for disease (including injection drug use, intravascular devices, poor dental hygiene, chronic hemodialysis, or infection with the human immunodeficiency virus). The mitral valve is the most commonly affected site, followed in decreasing frequency by the aortic, tricuspid, and pulmonic valves.
For native valve–related infective endocarditis in the developed world, mitral valve prolapse is a common predisposing cardiac lesion. Other underlying structural defects include congenital defects (most commonly bicuspid aortic valve), degenerative cardiac lesions (particularly calcific aortic stenosis), and rheumatic heart disease. In developing countries, rheumatic heart disease creating valvulopathy remains the leading underlying risk factor. For native valve–related lesions, left-sided disease predominates, and mortality ranges from 16% to 27%. Short-term mortality increases in those with left-sided native valve endocarditis when accompanied by other severe comorbid illnesses, abnormal mental status, congestive heart failure, or a bacterial etiology other than Streptococcus viridans and Staphylococcus aureus, and when treated with medical therapy absent valve surgery.8
The estimated risk in injection drug users is 2% to 5% per year, with a mean age of diagnosis being 30 years old. When endocarditis occurs in injection drug users, it has a predilection for the tricuspid valve. Other features include increased susceptibility to recurrence (approximately 40%) and increased mortality in those with concurrent human immunodeficiency virus and evidence of immunosuppression (defined as a CD4+ T-cell count of <200/mm3). Large vegetation size and fungal organism are predictive of poor outcome in injection drug use–associated right-sided endocarditis.9
Indwelling vascular devices create greater risks that microorganisms will attach to valves during bacteremia. Healthcare-associated endocarditis occurs when (1) a diagnosis is made >72 hours after admission in patients with no evidence of endocarditis on admission or in whom the disorder develops within 6 months after hospital discharge; or (2) cardiovascular manipulations have ...