of acute IE, electrocardiography should be performed every 48 to 72 hours to help rule out the
development of septal abscesses.
Rheumatoid factor is present in 50% of patients and subacute IE. It disappears as
successful treatment and may serve as a “poor man’s” substitute for measuring circulating
immune complexes (72). The nonspecific findings of elevated sedimentation rate, anemia
chronic disease, proteinuria, and hematuria are not helpful in the diagnosis of IE.
Because of the prevalence of false-negative blood cultures, especially HCIE, that are due
to the empirical use of antibiotics, several types of imaging techniques have been applied
the diagnosis of valvular infection. Radionuclide scans, such as Ga-67 and In-111 tagged white
cells and platelets have been used in diagnosing myocardial abscesses. These techniques
have been generally been of little help because of their poor resolution and high rate of false
negatives (174).
Echocardiography has become the imaging modality of choice for the diagnosis and
management of valvular infection. Despite the long-term availability of this technique, there
remains a good deal of confusion regarding the indications for its use of as well as the role of
transthoracic echocardiography (TTE) versus transesophageal echocardiography in valvular
infections. Neither TTE nor TEE should be used in patients with a low clinical probability of IE.
Interestingly, pneumonia appears to be the most common alternative diagnoses in these
situations (175). Up to 50% of vegetations, demonstrated by either type of echocardiography,
represents sterile platelet/fibrin thrombi, or nonbacterial thrombotic endocarditis (NBTE).
There are few if any echocardiographic criteria that definitely differentiate infected from
noninfected thrombi. Fifty percent of vegetations actually represent leaflet thickening. There is a
good deal of interobserver variability in reading either type of echocardiogram. Fifteen percent
of cases of IE have no detectable vegetations on echocardiography at any given time (176–179).
Vegetations must be of 3 mm to 6 mm in diameter to be reliably imaged by a
transthoracic echocardiography (TTE). A transoesophageal echocardiography TEE may define
structures down to 1 mm in diameter. The sensitivity of detecting NVIE ranges up to 95%
compared with 68% for TTE. A TTE is ineffective in 15% of patients because of chronic
obstructive pulmonary disease (COPD). It has only a 35% sensitivity for detecting PVE as
compared with greater than 75% for TEE. TEE is also the superior modality for detecting right-
sided vegetations. The negative predictive value of IE by TEE approaches 100% (181).
A TTE should be ordered initially except in the setting of possible PVE, abnormal body
habitus, known valvular abnormality, orS. aureusbacteremia. If there are no positive findings
on TTE, the likelihood of IE is very low, and a TEE should not be performed unless there are
persistently positive blood cultures without a definable source or the TTE study was
technically unsatisfactory. Table 9 presents the indications for performing echocardiography in
NVIE and PVE (182). All cases of proven IE should have an echocardiographic study in order
to set the baseline for that individual and so more accurately monitor the therapeutic response
and to detect the onset of complications especially aortic regurgitation.
The characteristics of the vegetations are useful in predicting the risk of embolization and
abscess formation. Vegetations greater than 10 mm in diameter and those which exhibit
significant mobility are three times more likely to embolize than those without these features.
Vegetations of the mitral valve, especially those on the anterior leaflet, are more likely to
embolize than those located elsewhere. Myocardial abscess formation is positively correlated
with aortic valve infection and intravenous drug abuse (183–186).
CT and MRI currently have almost no role in managing cases of IE. The relative
“slowness” of current technology is the major limiting factor.
DIAGNOSIS
Presumptive Clinical Diagnosis
Whenever there is a BSI with bacteria capable infected in native of prosthetic valve, the
possibility of IE must be actively ruled out. IE is a “cannot miss” diagnosis. The presence of the
continuous bacteremia, by itself, is adequate for the working diagnosis of IE because no other
infection is capable of producing it. A true diagnostic challenge is the clinical scenario in which
the patient’s clinical signs and symptoms are consistent with IE but the blood cultures are
persistently negative (see ‘Mimics of Endocarditis’).
Infective Endocarditis and Its Mimics in Critical Care 235