Endocarditis is infection of one or more heart valves, although infection of mural thrombi is usually included. The disease used to be separated into two types, acute and subacute. The acute type was most often caused by S. aureus, usually affected a normal heart valve, and had a relatively short severe course. Other bacteria less frequently associated with acute endocarditis were Streptococcus pyogenes, Streptococcus pneumoniae, and Neisseria gonorrheae. The subacute type was most often caused by streptococci of the viridans group with the enterococci next in frequency, usually involved previously abnormal heart valves, and had a relatively protracted course of varying severity. However, there is considerable overlap of clinical symptoms and severity of disease between the acute and subacute groups, and infections that would fit into the acute group may occur on previously abnormal heart valves. Therefore, many experts now include both types of endocarditis within one term, infective endocarditis.

Predisposing conditions. Certain conditions predispose to infection of normal heart valves. Some examples are IV injection of drugs by drug abusers, indwelling vascular catheters or other devices, major surgical operations, and factors that decrease immunologic resistance. Conditions that create abnormal heart valves are congenital heart disease, rheumatic fever, atherosclerotic valve lesions, prosthetic heart valves, mitral valve prolapse, and nonbacterial valve thrombi. In addition to the same conditions that predispose to infection of normal valves, dental operations and urinary tract manipulation frequently precede development of infection on already abnormal valves.

Organisms associated with endocarditis. Almost every pathogenic (and many relatively non-pathogenic) bacterium has been reported to cause infective endocarditis. Streptococci are the most frequent organisms (about 60%, range 50%-83%), with viridans group streptococci accounting for about 35% of all cases (range, 30%-52%); group D streptococci about 20%, and enterococci about 12% (range, 5%-20%). Staphylococcus aureus is isolated in about 20%-25% (range, 9%-33%) of cases. Coagulaseegative staphylococci are seen more frequently in association with intravascular catheters. When the patient has decreased resistance to infection (refer to the box), fungi and bacteria that ordinarily are considered uncommonly pathogenic or nonpathogenic may become involved.

Clinical findings. Infective endocarditis falls about halfway between bacteremia and septicemia. Bacteria grow in localized areas on damaged heart valves and seed the peripheral blood; this may be infrequent, intermittent, or relatively continuous, with gradations between the two extremes. Classic signs and symptoms include fever, heart murmurs, petechial hemorrhages in the conjunctivae, small “splinter hemorrhages” in the fingernail beds, and splenomegaly. Hematuria is very frequent, and red blood cell (RBC) casts are a common and very suggestive finding. There often (but not always) is a normocytic and normochromic anemia. Leukocytosis is often present, but it too may be absent. Signs and symptoms are variable among individual patients, and the diagnostic problem is often that of a fever of unknown origin.

Conditions Associated with Increased Rate of Infection or Infection by Unusual or Resistant Organisms
Human immunodeficiency virus infection (e.g., AIDS)
Diabetes mellitus (poorly controlled)
Trauma (severe)
Burns (extensive or severe)
Very young or very old age
Adrenocorticosteroid therapy
Immunosuppressant therapy
Chemotherapy of tumors
Antibiotic therapy (inappropriate or broad-spectrum)
Leukemia and lymphomas
Widespread nonhematologic malignancies
Indwelling urinary catheters
Indwelling vascular catheters

Diagnosis of infective endocarditis and identification of the organism involved are made through blood cultures. The clinical picture, evidence of abnormal heart valves, and the organism that is isolated (e.g., S. viridans) are major considerations in differentiating endocarditis from bacteremia and septicemia. However, in spite of the fact that viridans and group D streptococci together cause about 50%-60% of infective endocarditis, a few investigators state that when these organisms are isolated from blood culture, they are actually contaminants (or are not responsible for clinically significant disease) in 75% or more of the isolates. Blood culture methods used for diagnosis of infective endocarditis are the same as those used for septicemia.

Blood cultures. Many physicians draw one specimen from each of two different sites (“one culture set”) to increase total sample volume and to help decide whether certain organisms are more likely contaminants (e.g., S. epidermidis in only 1 of 2 specimens). One must avoid contamination by skin bacteria by aseptic technique; that is, cleansing the skin with alcohol, then with iodine or an iodophore like Betadine, which has the most efficient bactericidal effect of the common antiseptics available. It is then removed by alcohol. Since alcohol inactivates iodophores, the iodophore must remain on the skin a minimum of 1 minute, then it is removed by alcohol. Some obtain the blood culture before removing the iodophore or iodine tincture preparation. Alcohol requires at least 2 minutes contact time to be effective. Several reports emphasize the need for adequate quantities of blood per bottle (at least 5 ml and preferably 10 ml) to maximize bacterial recovery rate. The optimum specimen quantity depends on the amount of diluting medium and the presence or absence of certain additives, such as sodium polyanetholsulfonate (Liquoid). Repeated blood cultures are even more necessary in possible endocarditis than in diagnosis of septicemia because of the often intermittent nature of the blood involvement; avoidance of culture contamination becomes even more important. About 15% of patients (literature range, 2.5%-64%) do not have positive blood cultures. Uremia is especially apt to be associated with negative cultures.

False negative blood cultures. Some possible reasons for false negative blood cultures include recent antibiotic therapy, insufficient blood obtained for the amount of culture media, use of culture media unsuitable for anaerobes or for bacteria with special growth requirements, slowly growing organisms not detected during usual examination periods, various technical laboratory problems (specimens not obtained at optimal times), and combinations of these factors. One of the most important problems is insufficient blood specimen, especially when the number of organisms is small. Most investigators consider a 1:10 ratio of blood to culture medium to be optimal, and some insist on a minimum of 10 ml of blood (in adults) instead of the usual 5 ml. As noted previously, the optimal amount of specimen depends on the amount of diluting medium and the type of culture medium and system. There have been many blood culture systems advocated: use of many different media, vented and unvented containers, different anticoagulants and additives, hypertonic glucose media, filter or radioisotope detection equipment, and so forth. Interestingly, no system has consistently detected all bacteria all of the time. When two or more culture systems are compared, each system almost invariably detects a certain percentage of organisms that the others miss although some systems provide overall better results than others.

Nutritionally deficient streptococci. Occasional patients are infected by streptococci that grow in blood culture media but not on media used to subculture and identify the organisms. This could lead to a false impression of a negative culture. These streptococci are nutritionally deficient (not all in the same nutrient) and will grow on subculture media that are supplemented with the nutrients (e.g., pyridoxine) that they need. They usually grow as satellites around colonies of staphylococci, much as H. influenzae does.