Duke's Criteria of IE

New Criteria for Diagnosis of Infective Endocarditis: Utilization of Specific Echocardiographic Findings DAVID T. DURACK, M.B., D.Phil., ANDREA S. LUKES, B.A. , DAVID K. BRIGHT, M.D., Pharm. D., and the DUKE

ENDOCARDITIS SERVICE,* Durham, worth Carolina

PURPOSE: This study was designed to develop improved criteria fair the diagnosis of infective endocarditis and to compare these criteria with currently accepted criteria in a large series of cases.

PATIENTS AND METHODS: A totd of 405 Con-

secutive cases of suspected infective endocarditis in 353 patients evaluated ih a tertiary care hos&tal from 1985 to 1992 were analyzed using new diagnostic criteria for endocarditis. We defined two “major criteria” (typical blood culture and positive echocardiogram) and six “minor criteria” (predisposition, fever, vascular phenomena, immunologic phenomena, suggestive echocardiogram, and suggestive inicrobidlogic findings). We also defined three diagnostic categorks: ( 1) “definite” by pathologic or clinical criteria, (2) “possible,” and (3) “rejected.” Each suspected case of endocarditis was classified using both old and new criteria. Sixty-nine $athologically proven cases w&-e reclassified after exclusion of the surgical or autopsy findings, enabling comparison of clinical diagnostic criteria in proven cases.

RESULTS: Fifty-five (80%) of the 69 pathologically confirmed cases were classified as

*The Duke Endocarditis Service is comprised of Mark J. Alberts, M.D., Thomas M. Bashore, M.D., G. Riilph Corey, M.D., James M. Douglas, M.D., Linda Gray, M.D., Frank E. Harrell, Jr., Ph.D., J. Kevin Harrison, M.D., Sheila A. Heinle, M.D., Arthur Morris, M.D., Joseph A. Kisslo, M.D., L.M. Nicely, R.C.V.T., Newland Oldham, M.D., Lisa M. Penning, B.S., Daniel J. Sexton, M.D., Michael Towns, M.D., and Robert A. Waugh, M.D., Durham, North Carolina.

From the Departments of Medicine (DTD, TMB, GRC, SAH, JKH, JAK, LMN, LMP, DJS, RAW), Community and Family Medicine (FEH), Surgery. (JMD, NO), Pathology (AM, MT), Radiology (MJA, LG), and Duke University School of Medicine (ASL, DKB), Duke University Medical Center, Durham, North Carolina 27710. Supportwas also provided by the American Heart Association (ASL). Andrea Lukes and David Bright are each the recipient of a Stead Scholarship.

Request for reprints should be addressed to David T. Durack, M.B., D.Phil., Box 3867, Division of Infectious Disease and International Health, Duke University Medical Center, Durham, North Carolina 27710.

Manuscript submitted April 1, 1993, and accepted in revised form August 6, 1993.

clinically definite endocarditis. The older criteria classified oniy 35 (51%) of the 69 pathologically confirmed cases intb the analogous probable category (p ~0.0001). Twelve (17%) pathologically cdnfirmed cases were rejected by older clinical criteria, but none were rejected by the new criteria. Seventy-one (21%) of the remaining 336 cases that were not proven fiatholpg+ally were probable by older criteria, whereas the new criteria almost doubled the number of definite cases, to 135 (40%, p ~0.01). Of the 150 cases rejected by older criteria, 11 were definite, 87 were possible, and 52 were rejected by the new criteria.

CONCLUSION: Application of the h&posed new criteria increases the number of definite diagnoses. This should be useful for more accurate diagnosis &nd classification of patients with suspected endocardttis and provide better entry criteria for epidemiologic studies and clinical trials.

I n a surprisingly high proportion of cases, infective endocarditis is difficult to diagnose with

certainty. This is due to both the inaccessibility of intracardiac vegetations and the highly variable and sometimes nonspecific nature of the clinical manifestations. Overdiagnosis and underdiagnosis of infective endocarditis are common.

In 1981, von Reyn and colleagues [ll published a paper entitled “Infective endocarditisr An analysis based upon strict case definitions.” They analysed 135 cases treated between 1970 and 1977. Seeking to exclude doubtful cases in order to enhance the quality of their case analysis, the authors proposed four diagnostic categories: “definite,” “probable,” “possible,” and “rejected,” with criteria for each (Table I). The categories proposed by von Reyn et al were rapidly accepted, reflecting both their merit and the lack of clear diagnostic criteria at that time.

The von Reyn definitions continue to be widely used today, either unchanged [2-71 or with various ad hoc modifications [8-111. Wide acceptance not-

200 March 1994 The American Journal of Medicine Volume 96

withstanding, these definitions are not well suited for use in clinical settings, as illustrated by the following three case histories.

Case 1 A 42-year-old man with a history of rheumatic

fever at age 17 and subsequent chronic rheumatic heart disease was admitted with persistent fevers above 38°C shortness of breath, and murmurs of mitral and aortic insufficiency. All blood culture specimens were negative, but transesophageal echocardiography revealed an oscillating mass on the anterior mitral valve and fenestrations of the right coronary cusp of the aortic valve. Rheumatoid factor was positive at a titer 1:160. Culture- negative subacute infective endocarditis was diagnosed. The patient underwent aortic and mitral valve replacement after 3 weeks of antibiotic treatment. Vegetations containing colonies of gram- positive cocci were found on both excised valves. Cultures of the valves and vegetations were negative. This patient had infective endocarditis proven by pathology, but the diagnosis prior to surgery was rejected by the von Reyn criteria (Table I). The major reason in this case is the failure of these definitions to assign any diagnostic value to echocardiography.

Case 2 A 44-year-old man with a history of intravenous

drug use was admitted with persistent fevers above 38°C worsening shortness of breath, and pleuritic chest pain for 1 week. Two blood cultures were positive for Staphylococcus aureus. Chest roent- genograph showed opacities suggesting septic pulmonary emboli, and transesophageal echocardiography revealed an oscillating mass on the tricuspid valve. This typical example of right-sided staphylococcal endocarditis would be classified as possible under the von Reyn criteria (Table I). The problem could be corrected by including intravenous drug use as a predisposition for infective endocarditis and, as in Case 1, utilizing the results of echocardiography.

Case 3 A previously healthy 57-year-old white man was

admitted for 3 days of fever above 38”C, nausea, vomiting, confusion, and loss of coordination in the right hand. Two blood cultures were positive for S. aureus. The patient had no known primary focus of infection. Magnetic resonance imaging of the brain revealed multiple recent infarcts and a brain abscess. An echocardiogram revealed an oscillating mass on the mitral valve. The patient underwent

TABLE I The von Reyn Criteria for Diagnosis of Infective Endocarditis*

DIAGNOSIS OF INFECTIVE ENDOCARDITIS / DURACK ET AL

Definite Direct evidence of infective endocarditis based on histology from surgery

or autopsy, or on bacteriology (Gram’s stain or culture) of valvular vegetation or peripheral embolus.

Probable (A.) Persistently postitive blood cultures+ plus one of the following:

(1.) New regurgitant murmur, or (2.) Predisposing heart disease’ and vascular phenomena5

(B.) Negative or intermittently positive blood cultures**plus three of the following:

(1.) Fever (2.) New regurgitant murmur, and (3.) Vascular ohenomena

Possible (A.1 Persistently positive blood cultures plus one of the following

(1.) Predisposing heart disease, or (2.) Vascuiar phenomena

(B.) Negative or intermittently positive blood cultures with all threeof the following:

(1.) Fever (2.) Predisposing heart disease, and (3.) Vascular phenomena

(C.) For viridans streptococcal cases only: at least two positive blood cultures without an extra-cardiac source, and fever

Rejected (A.) Endocarditis unlikely, alternative diagnosis generally apparent (B.) Endocarditis likely, empiric antibiotic therapy warranted (C.) Culture negative endocarditis diagnosed clinically, but excluded by

postmortem

,dapted from 111. t least two blood cultures obtained, with two of two positive, three of three positive, or at least 1% of cultures positive if four or more cultures obtained. efrnite valvular or congenital heart disease, or a cardiac prosthesis (excluding permanent cemakers). etechiae, splinter hemorrhages, conjunctival hemorrhages. Roth spots, Osler’s nodes, Janeway ,ions, aseptic meningitis, glomerulonephritis, and pulmonary, central nervous system, coronary or ripheral emboli Any rate of blood culture positivity that does not meet the definition of persistently positive.

mitral valve replacement after receiving intravenous antibiotic treatment for 31 days. Sections of the valve revealed acute inflammation and gram- positive cocci. By clinical criteria, this patient definitely has endocarditis, but his case would be diagnosed as only possible by the von Reyn criteria, prior to surgery (Table I). This could be corrected by giving extra diagnostic weight to community- acquired S. aureus bacteremia when no alternative focus of infection can be identified [121. As in the aforementioned cases, echocardiography provided major diagnostic information.

These three cases illustrate common problems that arise during clinical application of the von Reyn criteria, especially their failure to utilize echocardiographic findings. Echocardiography is now routinely used in evaluating patients with suspected endocarditis, reflecting the improved capability of this technique to detect vegetations [131. This development alone provides sufficient justification for a major revision of the diagnostic criteria for endocarditis [13,14].

Cases 2 and 3 illustrate difficulties in classifying acute cases of endocarditis. The emphasis placed

March 1994 The American Journal of Medicine Volume 96 201

TABLE II Proposed New Criteria for Diagnosis of Infective Endocarditis

Definite Infective Endocarditis Patholoeic criteria

Microkganisms: demonstrated by culture or histology in a vegetation, or in a vegetation that has embolized, or in an intracardiac abscess, or

Patholoeic lesions: vegetation or intracardiac abscess oresent. confirmed by histolo& showing active endocarditis ’

Clinical criteria, using specific definitions listed in Table Ill 2 major criteria, oi 1 major and 3 minor criteria, or 5 minor criteria

Possible Infective Endocarditis Findings consistent with infective endocarditis that fall short of “Defi-

nite,” but not “rejected.” Rejected

Firm alternate diagnosis for manifestations of endocarditis, or Resolution of manifestations of endocarditis, with antibiobc therapy for

4 days or less, or No pathologic evrdence of infective endocarditis at surgery or autopsy,

after antibiotic therapy for 4 days or less

on predisposing heart disease by the von Reyn criteria is appropriate for subacute disease, in which this is a usual feature. In acute endocarditis, this emphasis makes diagnosis more difficult because those patients with acute endocarditis who lack predisposing heart disease must have a new regurgitant murmur to be classified as probable endocarditis by von Reyn criteria (Table I). Without a new murmur, such cases would be rejected (unless a vascular phenomenon occurred, which would result in classification as possible endocarditis).

The definitions in the rejected category of von Reyn seem self-contradictory. The first subgroup is “endocarditis unlikely, alternative diagnosis generally apparent,” but the second is “endocarditis likely, empiric antibiotic therapy warranted” (Table I). As a result, this category yields an artificially high number of rejected cases.

After encountering many similar examples of difficulties in diagnosis of endocarditis at our institution, we set out to develop improved criteria for identifying infective endocarditis. Various combinations of carefully defined clinical findings were chosen on the basis of theoretical considerations and clinical experience. The new criteria were tested on a small training sample of approximately 25 patients referred to the Duke Endocarditis Service, followed by further modification and revision. The final criteria were then re-tested in over 400 episodes.

PROPOSED NEW CRITERIA New criteria for the diagnosis of endocarditis are

proposed in Table II. Here, we briefly discuss these new criteria, showing how they can increase the proportion of definite diagnoses.


Definite Endocarditis DEFINITE BY PATHOLOGIC CRITERIA: The diagnosis

of infective endocarditis is certain when appropriate pathologic specimens from surgery or autopsy reveal positive histology and/or culture. Our criteria for definite infective endocarditis by pathologic criteria are essentially similar to those used in the definite category of von Reyn, except for some minor improvements in wording. This group constitutes about a third of our patients with infective endocarditis, leaving a majority in whom the diagnosis must be determined on clinical grounds.

DEFINITE BY CLINICAL CRITERIA: Although only surgical or autopsy specimens can provide absolute proof of the diagnosis, we believe that definite diagnoses can and should be made by applying defined clinical criteria. Our category of definite endocarditis by clinical criteria is intended to iden- tify patients with a very high probability of having a true-positive diagnosis. Appropriate weighting of the various clinical findings is achieved by specify- ing major and minor criteria, analogous to the Jones criteria for diagnosis of rheumatic fever [l&16]. A similar concept was proposed previously by El-Khatib, Wilson, and Lerner [171 for the diagnosis of acute endocarditis in drug addicts.

POSSIBLE: Patients with findings consistent with infective endocarditis that fall short of definite by clinical criteria, but are not rejected, are placed into the category of possible. This category includes a wide range of indeterminate cases that have very different probabilities of infective endocarditis. Therefore, classification into the category of possible endocarditis does not imply that antimi- crobial therapy should or should not be given. Clinicians must draw their own conclusions about the likelihood of endocarditis and the need for empirical therapy in possible cases.

REJECTED: The rejected category is defined strictly. The diagnosis of infective endocarditis can be as difficult to exclude as it is to establish with certainty. Negative or indeterminate pathology from surgery or autopsy does not necessarily exclude the diagnosis, especially when prior antibiotic treatment has been given. Infective endocarditis sometimes can be cured with brief courses of antibiotics-10 days or less [18-211. Therefore, we have classified patients who received more than 4 days of antibiotic therapy as possible infective endocarditis unless a firm alternative diagnosis is made. Long-term cure or survival with 4 days or less of antibiotic therapy provides strong evidence against the diagnosis.

Specific terminology for the proposed new criteria is defined in Table III, and the rationale is discussed briefly, as follows.


DIAGNOSIS OF INFECTIVE ENDOCARDITIS I DURACK ET AL

MAJOR CRITERIA Blood Cultures

Two features of positive blood cultures have special significance for the diagnosis of infective endocarditis: isolation of a typical organism and persistent bacteremia. Typical organisms are those with positive predictive value for infective endocarditis when cultured from blood. These comprise viridans streptococci including nutritional variant strains, Streptococcus bouis, and the HACEK group of fastidious gram-negative organisms: Huemophi- lus species, Actinobacillus actinomycetemcomi- tans, Cardiobacterium hominis, Eikenella species, and Kingella kingae. Isolation of one of these organisms from at least two blood cultures constitutes a major criterion for diagnosis of infective endocarditis (IE), whether or not the bacteremia is community-acquired.

Although S. aureus and enterococci are typical endocardial pathogens, the majority of patients with bacteremia due to these organisms do not have infective endocarditis. Two important factors increase the probability of endocarditis when either S. aureus or enterococci have been isolated: (1) a community-acquired infection, and (2) the absence of an alternative focus of infection. Maki et al [22] and Bayer et al [12] recently reported series of patients with enterococcal and staphylococcal bacteremia, respectively, that support this conten- tion. When both conditions are met, isolation of S. aureus or enterococci from at least two blood cultures constitutes a major criterion for diagnosis of IE.

Persistent bacteremia constitutes a major criterion for diagnosis of infective endocarditis. Persis- tent bacteremia is defined as follows: either two or more positive blood cultures for the same organism separated by at least 12 hours, or three or more positive blood cultures for the same organism with at least 1 hour between the first and last culture.

Coagulase-negative staphylococci, usuaIly Staphylo- coccus epidermidis, are not listed under typical organisms. Although these organisms are a leading cause of prosthetic valve endocarditis, they are also frequent contaminants in blood cultures. Therefore, S. epidermidis bacteremia must be persistent to be considered a major criterion for infective endocarditis.

Evidence of Endocardial Involvement ECHOCARDIOGRAPHY: Visualization of the primary

endocardial lesion by echocardiography, when com- bined with other diagnostic information, is vital to the diagnosis of endocarditis. Because the reliabil- ity of echocardiography depends upon the quality of available equipment and the skills of the staff, some investigators still question its role [21. How-

TABLE III

Definitions of Terminology Used in the Proposed New Criteria

Maior Criteria Positive blood culture for infective endocarditis

Typical microorganism for infective endocarditis from two separate blood cultures

Viridans streptococci,* Streptococcus bovis, HACEK group, or Community-acquired Staphyloccus aureus or enterococci, in the

absence of a primary focus, or Persistently positive blood culturt, defined as recovery of a microor-

ganism consistent with infective endocarditis from: (i) Blood cultures drawn more than 12 hours apart, or

(ii) All of three or a majority of four or more separate blood cultures, with first and last drawn at least 1 hour apart

Evidence of endocardial involvement Positive echocardiogram for infective endocarditis

(0 Oscillating intracardiac mass, on valve or supportingstruc- tures, or in the path of regurgitant jets, or on implanted mate- rial, In the absence of an alternative anatomic explanation, or

(ii) Abscess, or (iii) New partial dehiscence of prosthetic valve, or

New valvular regurgitation (increase or change in pre-existing murmur not sufficient)

Minor Criteria Predisposition: predisposing heart condition or intravenous drug use Fever: t 38.o”C (100.4”F) Vascular phenomena: major arterial emboli, septic pulmonary infarcts,

mycotic aneurysm, intracranial hemorrhage, conjunctival hemorrhages, Janeway lesions

Immunologic phenomena: glomerulonephritis, Osler’s nodes, Roth spots, rheumatoid factor

Microbiologic evidence: positive blood culture but not meeting major criterion as noted previously+ or serologic evidence of active infection with organism consistent with infective endocarditis

Echocardiogram: consistent with infective endocarditis but not meeting major criterion as noted previously

Eikenella spp., and Kingella kingae. *including nutritional variant strains. iExcluding single positive cultures for coagulase-negative staphylococci and organisms that do not cause endocarditis.

ever, the high detection rate of modern echocardiography for vegetations and other lesions provides overwhelming support for its central role in diagnosis of infective endocarditis today [23-261.

An echocardiogram should be performed in all patients with suspected endocarditis [13]. The transesophageal technique is more sensitive than transthoracic echocardiography for the detection of vegetations [13,14,23] and is safe [27,28]. In practice, transthoracic echocardiography is performed first because it is easier and cheaper to perform. Transesophageal echocardiography is usu- ally unnecessary if transthoracic echocardiography is positive, but it is recommended whenever endocarditis is strongly suspected and the results of the transthoracic echocardiography are negative or equivocal [131, and is especially valuable in patients with prosthetic valves. Transesophageal imaging also is much more sensitive for the detection of two important complications of endocarditis: abscesses and valve perforations [14,28-301.

Three echocardiographic findings provide sufficient evidence of endocardial involvement for each to be considered as a major criterion. The first is an oscillating intracardiac mass located at sites where


vegetations typically occur, such as on valves, chordae, or in the path of turbulent jets of blood passing through incompetent valves or septal de- fects. Alternative explanations, such as ruptured chordae and stable healed vegetations from previous episodes of infective endocarditis, should be excluded. The second finding is an intracardiac abscess. Abscesses are detected much less commonly than vegetations but, if present, provide strong evidence for infective endocarditis. The third major finding is new partial dehiscence of a prosthetic valve. Because periprosthetic leaks occur after valve implantation in 10% to 15% of patients in the absence of infection, we emphasize that the dehiscence must be documented to be new.

NEW REGURGITANT MURMUR: Similarly, detection of a new regurgitant murmur provides evidence of endocardial involvement. Again, this must be documented to be new in order to constitute a major criterion for active endocarditis; change in a pre- existing murmur is not sufficient.

MINOR CRITERIA Predisposition

Presence of one or more predisposing heart conditions increases the prior probability that a patient has infective endocarditis. The risk posed by various cardiac conditions for endocarditis has been ranked previously 1311, As predispositions, we included those conditions presenting high or intermediate risk and exclude those ranked as low risk [311. An important addition in our proposed criteria is the inclusion of intravenous drug use as a predisposing factor.

Fever Although the presence of fever clearly is nonspe-

cific, absence of fever has negative predictive value for the diagnosis of infective endocarditis. In our own series, fever was reported in 87% of definite cases (see Results). Therefore, we have included fever greater than 38°C (100.4”F) as a minor criterion.

Vascular Phenomena When the diagnosis of infective endocarditis is

suspected, major arterial emboli (including septic pulmonary emboli) provide supporting evidence for the presence of an infected intravascular lesion. Peripheral necrotic skin lesions, as commonly seen in S. aureus endocarditis, are probably due to arterial emboli and are included under vascular phenomena. Minor emboli to capillaries and nonspecific peripheral signs that might be due to tiny emboli are excluded.

Mycotic aneurysms, although much less common than emboli, are strongly associated with infective endocarditis. Central nervous system hemorrhages can result from either emboli or mycotic aneurysms in the setting of endocarditis. Conjunc- tival hemorrhages and Janeway lesions are classi- cal peripheral signs that are common in infective endocarditis. Whether the pathogenesis of some of the peripheral lesions of endocarditis is primarily vascular, immunologic, or both is uncertain, so we have assigned them arbitrarily for the purposes of these criteria (Table III).

Petechiae and splinter hemorrhages are included as vascular phenomena in von Reyn’s criteria. Splinter hemorrhages can occur in healthy individuals and/or in other cardiac conditions, whereas petechiae can occur in many other dis- eases. We consider these manifestations of infective endocarditis to be nonspecific and have excluded them from our proposed criteria 132,331.

Immunologic Phenomena Several immunologic phenomena are associated

with infective endocarditis. Elevated levels of rheumatoid factor that resolve with treatment have been reported in 30% to 50% of cases of subacute infective endocarditis [34,35] and in up to 25% of acute cases [361. Pre-existing positive rheumatoid factor cannot be used as a minor criterion, Im- mune complex glomerulonephritis, although un- common, can be an important complication of endocarditis [371. Immunologic mechanisms have also been implicated in the development of Osler’s nodes and Roth spots 1321.

Echocardiography Echocardiographic abnormalities consistent with

infective endocarditis but which do not meet our strict definitions for a major criterion (as previously discussed) are commonly found. Examples include nonoscillating targets, new valvular fenestrations, and nodular valvular thickening, which are consistent with infective endocarditis. Al- though these findings are less predictive for infective endocarditis than the major echocardiographic criteria, they have some diagnostic significance in the setting of suspected infective endocarditis and have been assigned minor criterion status. This allows for some flexibility in utilizing differing interpretations of some of the less certain echocardiographic manifestations of endocarditis by individual readers.

It should be noted that some echocardiographic findings, for example, valvular thickening with regurgitation or a nonoscillating target that was known to exist before an episode of suspected



endocarditis, or targets that persist after cure, could be utilized either as a predisposing heart condition or as a minor echocardiographic criterion. However, such a finding should not be used as both.

Microbiology When a bacteremia is neither typical nor persis-

tent, it has low predictive value for endocarditis. Nevertheless, such bacteremias can provide some supporting evidence for a diagnosis of infective endocarditis. A positive blood culture, even if noso- comial or associated with another focus of infection, can be used as a minor criterion. Single isolates of coagulase-negative staphylococci and organisms such as viruses and atypical mycobacte- ria, which for all practical purposes do not cause endocarditis, are excluded from this category.

Serologic evidence of active infection with organisms such as Brucella species, Chlamydia species, Coxiella burnetii, Legionella species, and Rochali- maea species could provide important supporting evidence for the diagnosis in these rare forms of infective endocarditis and, therefore, is included in this minor criterion.

PATIENTS AND METHODS To evaluate these new criteria, we conducted a

review of 421 consecutive cases of endocarditis or suspected endocarditis in 369 patients over a 7-year period. These cases included referrals to the Duke Endocarditis Service since its inception on October 1, 1989, through May 31, 1992, and discharge diagnoses of endocarditis from January 1, 1985, through May 31, 1992, in addition to 16 episodes from outside these dates. The medical records of 405 (96%) of these episodes were available for review and abstraction. Approximately a third of the cases were identified by referral to the Duke Endocarditis Service, whereas the remaining two thirds were identified from discharge diagnoses.

The Pearson x2 test was performed for examining the difference between two proportions. A two-sample Wilcoxon rank-sum test was performed for examining the difference between two medians of continuous variables. To compare the diagnostic sensitivities of the different criteria, the McNemar x2 test was performed. For all analyses, two-sided tests of significance were performed with an alpha of 0.05. Statistical analyses were performed using SAS software (SAS Institute Inc., Carry, N.C.).

RESULTS A total of 405 episodes in 353 patients were

analyzed. Two-hundred four (50%) of the cases in

DIAGNOSIS OF INFECTIVE ENDOCARDITIS / DURACK ET AL -

TABLE IV Demographic and Clinical Features in 204 Definite Episodes

A@ Median Mean &3 Range

sex 1 month to 92 years

Female (%) 101 (50) Male (%) 103 (50)

Race White t%) Black (%) Other i%)

Types of episode Acute (%) Subacute (%I Indeterminant (%)

Type of valve Native (%) Prosthetic (%)

Primary treatment Antibiotics alone (%) Antibiotics t surgery (%)

In-hospital mortality (%) Etiologic organisms

Streptococci 88 (43) Viridans 47 (23) Enterococci 14 (7) S. bovis 10 i5j S. agalactiae 11 (5) Other 6 (3)

Staphylococci 90 (44) S. aureus 76 (37) S. epidermidis S. warneri 7 Iti

HACEK 7 (3) Other species 9 (4) Polymicrobic 3 (1) Culture negative 7 (3)

\CEK = Haemophilusspp.,Actinobaci//usactinomycefemcomitans, Cardiobacferium homini 'ceneila spp.,and /tinge/la kingae.

185 patients were classified as definite endocarditis, with 69 (34%) being pathologically proven at surgery or autopsy. There were 149 (37%) possible and 52 (13%) rejected cases among the remaining 168 patients.

The main demographic and clinical features of the 204 definite episodes are presented in Table IV. No statistically significant differences were found between the two subgroups of definite endocarditis, pathologically and clinically diagnosed. A lower proportion of black patients had pathologically proven endocarditis compared with white patients (36% versus 48%, p = 0.11). This trend was due to a lower proportion of black patients undergoing valve surgery.

Twenty-eight of the 185 (15%) patients with definite endocarditis had recurrent endocarditis or a history of previous endocarditis. Of 11 patients with more than 1 episode during our 7-year study period, 2 patients had 5 episodes, 2 patients had 3 episodes, and 7 patients had 2 episodes. An addi- tional 17 patients gave a prior history of endocarditis.

Table V lists the various combinations of major and minor clinical criteria that classified episodes



TABLE V

Combinations of Major and Minor Diagnostic Criteria Used in the Clinical Diagnosis of 190 Definite Episodes

Pathoo;gic* Clinical Total 0 (%I (%)

2 Major criteria 1 Major and 3 Minor criteria 5 Minor criteria 1 (2) 4 (3) 5 (3)

Total 55 135 190

ourteen pathologically proven cases were possible by clinical criteria and are excluded.

as definite endocarditis. Excluding the 14 episodes that were classified as definite endocarditis by pathologic evidence only, there were 190 episodes classified as definite endocarditis by clinical criteria. The majority (110,58%) were associated with 2 major criteria; however, a significant proportion (75, 39%) were associated with 1 major and 3 minor criteria. Five cases (3%) were diagnosed by the presence of 5 minor criteria.

The percentages of major and minor criteria used in 204 clinically definite episodes are given in Table VI. Most definite episodes (195, 96%) had positive blood cultures, with 165 (81%) meeting the major criteria and 30 (15%) meeting the minor microbiologic criteria. Most definite episodes (177, 87%) also had echocardiographic evidence of endocarditis, with 119 (58%) meeting major criteria and

58 (28%) meeting minor criteria. The overall distri- butions of criteria for pathologically and nonpatho- logically confirmed episodes were similar.

Of the 204 definite episodes, the overwhelming majority (201 of 204, 99%) had at least 1 echocardiogram performed. The only three episodes lack- ing an echocardiogram occurred in severely ill patients. Two of these patients died within 1 day after admission; the other patient presented with severe intracranial hemorrhage and died 21 days after admission. Of the 201 episodes with at least 1 echocardiogram, all had a transthoracic echocardiography; in addition, 61 had a transesophageal echocardiography performed. As shown in Table VI, 119 (58%) definite episodes had at least 1 major echocardiographic criterion. Of these 119 episodes with a major echocardiographic criterion, 76 (64%) were identified using transthoracic echocardiography, 28 (24%) were identified by transesophageal echocardiography, and 15 (13%) were positive by both techniques.

Table VI shows the frequency with which all of the major and minor criteria were distributed in the definite, possible, and rejected categories defined by the new criteria.

In order to evaluate the proposed new criteria, we compared the new criteria with the previously accepted criteria [l] (Table I), using all 405 sus-

TABLE VI

Frequency of Diagnostic Criteria in the Subgroups Defined by the New Criteria

Major criteria Positive blood cultures

Typical microorganism Persistently positive blood culture Both

Evidence of endocardial involvement Positive echocardiogram

Oscillating target Abscess New dehiscence of prosthetic valve

New valvular regurgitation Minor criteria

Predisposition Heart disease Intravenous drug use Both

Fever ~38.0% Vascular phenomena

Arterial emboli Mycotic aneurysm Central nervous system hemorrhage Janeway lesions Conjunctival hemorrhage

Immunologic phenomena Glomerulonephritis Osler’s node Roth spot Rheumatoid factor

Microbiologic evidence Suggestive echocardiography

Possible (%) (n = 149)

63 (42) 31 (21)

E IEi

i: I;;

$i:i

121 (81)

‘i i I:;/ 15 (10)

100 (67)

i; Ii; l(1)

i

i 1;1 0 1 (1) l(l)

R~efe;$b’

:126; 2 (4) 0

4 (8) 3 (6) 0 1 (2) 2 (4)

ii i I::; 5 (10) 2 (4)

26 (50)

i

i

i

;

i


DIAGNOSIS OF INFECTIVE ENDOCARDITIS / DURACK ET AL -

TABLE VII TABLE VIII

Comparison of Clinical Diagnoses by von Reyn Criteria and the New Criteria in 69 Cases of Pathologically Proven Infective Endocarditis

Comparison of Clinical Diagnoses in 336 Cases Evaluated for Diagnosis of Infective Endocarditis, Excluding Pathologically Proven Cases

Probable von Reyn Criteria

Possible Rejected Total (%.)

New criteria Definite 32 19 4 Possible Rejected ; a ;

E I;:; 0 (0)

Total (%I 35 (51) 22 (32) 12 (17) 69 (100)

Probable van Reyn Criteria

Possible Rejected Total (%I

New criteria Definite 65 Possible 6

I Rejected 52 (15) Total (%I 7; (21) 11: (34, 1 !?I (45) 336 (100)

petted episodes. To compare both sets of criteria without involving any unproven diagnoses, we separated the 69 pathologically proven cases from the remaining 336 cases. For these 69 pathologically proven cases, we then excluded the pathologic findings and reclassified each based solely on clinical and echocardiographic criteria (Table VII). The sensitivity of the new criteria was 80% (55 of 691, significantly higher than the sensitivity of 51% (35 of 69) yielded by the older von Reyn criteria (p < 0.0001).

For the remaining 336 cases that were not pathologically proven, there were important differences in the number of clinical diagnoses yielded by the 2 sets of criteria (Table VIII). The new criteria classified 135 cases (40%) as definite. Of these 135 definite cases, only 65 (48%) were classified in the analogous probable category of von Reyn, and 11 (8%) were “rejected” (p <O.Ol). The new criteria define rejected more strictly than the previous criteria, resulting in a smaller number of rejected cases. Of the 149 possible cases classified by the new criteria, the majority (96%) were either possible or rejected by the von Reyn criteria.

COMMENTS The challenge in the diagnosis of infective endo-

carditis is to interpret, weigh, and combine diagnostic findings appropriately. We have proposed new criteria that are more versatile and practical than earlier systems, based on carefully defined major and minor criteria. Their application should increase the number of definite diagnoses that can be made on clinical grounds.

The analysis of pathologically confirmed cases is the only objective method available to formally compare clinical diagnostic criteria for endocarditis without using circular reasoning. The analysis of our 69 pathologically confirmed cases, after exclusion of the pathologic evidence, indicates that the sensitivity of the new criteria is significantly improved. These 69 patients could represent a biased subset because they required surgical inter- vention or died, increasing the likelihood that they

had acute, severe, or complicated forms of endocarditis. However, the demographics and many of the features of the pathologically proven and clinically diagnosed cases, including the percentage of acute episodes, distribution of etiologic organisms, and in-hospital mortality, were similar (Table VI). We also recognize the possibility of referral bias in the selection of patients because our institution is a tertiary referral center [38]. This, however, should not discredit the comparison of criteria because both were applied to the same population.

We recognize that our category of definite infective endocarditis by clinical criteria does not provide absolute proof that a patient has endocarditis. The new criteria were developed with the specific intent of improving sensitivity while maintaining specificity. We have demonstrated increased sensitivity in clinical application.

The probable category of the von Reyn criteria identified patients who have a high probability of having endocarditis. Not surprisingly, the majority of von Reyn probable cases were classified as definite by the new criteria (75 of 78, 96%). Inter- pretation of the possible category of von Reyn is more difficult. It contains some patients who clearly have endocarditis but also includes many patients who clearly do not have endocarditis. Approxi- mately one half of the von Reyn possible cases were classified as definite by the new criteria (59 of 115, 51%).

Our rejected category is precisely defined, resulting in fewer rejected cases. Only one third (501139, 36%) of the von Reyn rejected cases were classified as rejected by our criteria. Thus, the new criteria seem more specific, although the precise specificity cannot be calculated because the number of patients who definitely do not have endocarditis cannot be determined. We believe that this subset represents patients with an extremely low chance of a false-negative diagnosis.

A definite diagnosis of endocarditis is desirable to guide management, especially valve-replacement surgery. However, in the early stages of the evaluation of a suspected case of endocarditis, it is



often appropriate to give empiric treatment before there is sufficient information to confirm or reject the diagnosis. Good diagnostic criteria may help to clarify the early assessment of such patients, but their main value often comes later, when further information has accumulated. Accordingly, it should be emphasized that these criteria are not intended primarily to govern management decisions. Obviously, we believe that all cases classified as definite infective endocarditis should receive appropriate antibiotic treatment. Decisions regard- ing treatment for patients with possible endocarditis must be based on the individual circumstances of each case.

6. Van Der Meer JTM, Thompson J, Valkenburg HA, Michel MF. Epidemiology of bacterial endocarditis in the Netherlands. 1. Patient characteristics. Arch Intern Med 1992; 152: 1863-8. 7. Van Der Meer JTM, Thompson J, Valkenburg HA, Michel MF. Epidemiology of bacterial endocarditis in the Netherlands. II. Antecedent procedures and use of prophylaxis. Arch Intern Med 1992; 152: 1869-73. 8. Gallagher PG, Watanakunakorn C. Listeria monocytogenes endocarditis: a review of the literature 1950-1986. Stand J Infect Dis 1988; 20: 359-68.

9. Lederman MM, Sprague L, Wallis RS, Ellner JJ. Duration of fever during treatment of infective endocarditis. Medicine 1992; 71: 52-7. 10. lmperiale TF, Horwitz RI. Does prophylaxis prevent postdental infective

endocarditis? A controlled evaluation of protective efficacy. Am J Med 1990; 88: 131-6.

We suggest that all patients with definite infective endocarditis by the new criteria, whether pathologically or clinically diagnosed, should qualify for inclusion in epidemiologic or treatment trials of infective endocarditis. A recent set of guidelines approved inclusion of the possible category of von Reyn in treatment trials 121, but we believe that the diagnosis in such patients is not sufficiently firm to justify inclusion in either treatment trials or epidemiologic studies.

11. Nahass RG, Weinstein MP, Bartels J, Gocke DJ. Infective endocarditis in intravenous drug users: a comparison of human immunodeficiency virus type l-negative and -positive patients. J Infect Dis 1990; 162: 967-70. 12. Bayer AS, Lam K, Gintzon L, Norman DC, Chiu C-Y, Ward JI. Staphylococ- cus aureus bacteremia. Clinical, serologic, and echocardiographic findings in patients with and without endocarditis. Arch Intern Med 1987; 147: 457-62. 13. Khanderia BK. Suspected bacterial endocarditis: to TEE or not to TEE (editorial comment). J Am Coll Cardiol 1993; 21: 222-4. 14. Daniel WG, Schroder E, Nonnast-Daniel B, Lichtlen PR. Conventional and transoesophageal echocardiography in the diagnosis of infective endocarditis. Eur Heart J 1987; 8: 287-92. 15. Dajani AS, Ayoub E, Bierman FZ, et a/. Guidelines for the diagnosis of rheumatic fever. Jones criteria, 1992 update. JAMA 1992; 268: 2069-73. 16. American Heart Association ad hoc Committee. Jones criteria (revised) for guidance in the diagnosis of rheumatic fever. Circulation 1984; 69: 203A- 8A.

The validity of these proposed new criteria have not yet been tested in special populations such as infants and young children with suspected infective endocarditis. However, an independent valida- tion on a separate group of adult patients has been conducted (Bayer AS et al published in this issue). Future studies will report on the outcome of patients in the possible and rejected categories.

In conclusion, the diagnosis of endocarditis is often complex and uncertain. These problems can be minimized by utilization of carefully defined and clinically tested criteria, as proposed in this study. After wider application of the new criteria, refinements may be needed. Meanwhile, we believe that they offer substantial improvement over previous diagnostic criteria for infective endocarditis.”

17. El-Khatib MR, Wilson FM, Lerner AM. Characteristicsof bacterial endocarditis in heroin addicts in Detroit. Am J Med Sci 1976; 271: 197-201. 18. Hunter T. The treatment of some bacterial infections of the heart and pericardium. Bull NY Acad Med 1952; 28: 213-28. 19. Durack DT. Review of early experience in treatment of bacterial endocarditis, 1940-1955. In: Bisno AL, editor. Treatment of infective endocarditis. New York: Grune & Stratton, 1981: 1-14. 20. Chambers HF, Miller RT, Newman MD. Right-sided Staphylococcus aureus endocarditis in intravenous drug abusers: two-week combination therapy. Ann Intern Med 1988; 109: 619-24. 21. Bricaire R, Berkson L, Vilde JL, Frottier J, Verliac F, Bastin R. Cure of streptococcal endocarditis with six days of antibiotic therapy. JAMA 1986; 255: 1291.

REFERENCES

22. Maki DG, Agger WA. Enterococcal bacteremia: clinical features, the risk of endocarditis, and management. Medicine 1988; 67: 248-69. 23. Mugge A, Daniel WG, Frank G, Lichtlen PR. Echocardiography in infective endocarditis: Reassessment of prognostic implications of vegetation size determined by the transthoracic and transesophageal approach. J Am Coll Cardiol 1989; 14: 631-8. 24. Daniel WG, Mugge A, Martin RP, et a/. Improvement rn the diagnosis of abscesses associated with endocarditis by transesophageal echocardiography. N Engl J Med 1991; 324: 795-800. 25. Steckelberg JM, Murphy JG, Ballard D, et a/. Emboli in infective endocarditis: the prognosticvalueof echocardiography. Ann Intern Med 1991; 114: 635-40.

1. von Reyn CF, Levy BS, Arbeit RD, Friedland G, Crumpacker CS. Infective 26. Shapiro SM, Bayer AS. Transesophageal and Doppler echocardiography in endocarditis: an analysis based on strict case definitions. Ann Intern Med the diagnosis and management of infective endocarditis. Chest 1991; 100: 1981; 94: 505-17. 1125-30. 2. Wilson WR, Gilbert DN, Bisno AL, et a/. Evaluation of new anti-infective agents for the treatment of infective endocarditis. Clin Infect Dis 1992; 15 (Suppl 1): 89-95. 3. Bruyn GAW, Thompson J, Van Der Meer JWM. Pneumococcal endocarditis in adult patients. A report of five cases and review of the literature. Q J Med 1990; 74: 33-40. 4. Baddour LM. Twelve-year review of recurrent native-valve infective endocarditis: a disease of the modern antibiotic era. Rev Infect Dis 1988; 10: 1163-70. 5. Sussman JI, Baron EJ, Tenenbaum MJ, et al. Viridans streptococcal endocarditis: clinical, microbiological, and echocardiographic correlations. J Infect Dis 1986; 154: 597-603.

27. Nikutta P, Mantey-Stiers F, Becht I, et a/. Risk of bacteremia induced by transesophageal echocardiography: analysis of 100 consecutive procedures. J Am Sot Echocardiogr 1922; 5: 168-72. 28. Daniel WG, Erbel R, Kasper W, et a/. Safety of transesophageal echocardiography, a multicenter survey of 10,419 examinations. Circulation 1993; 83: 817-21. 29. Birmingham GD, Rahko PS, Ballantyne R. Improved detection of infective endocarditis with transesophageal echocardiogram. Am Heart J 1992; 123: 774-81. 30. Karalis DG, Bansal RC, Hauck AJ, et al. Transesophageal echocardiographic recognition of subaortic complications is aortic valve endocarditis: clinical and surgical implications. Circulation 1992; 86: 353-62.



31. Dajani AS, Bisno AL, Chung KJ, et a/. Prevention of bacterial endocarditis. Recommendations by the American Heart Association. JAMA 1990; 264: 2919-22. 32. Kerr A, Jr. Subacute bacterial endocarditis. Springfield, Illinors: Charles C. Thomas, 1955: l-343. 33. Kilpatrick ZM, Greenberg PA, Sanford JP. Splinter hemorrhages-their clinical significance. Arch Intern Med 1965; 115: 73b-5. 34. Messner RP, Laxdal T, Quie PG, Williams RC Jr. Rheumatoid factors in subacute bacterial endocarditis-bacterium, duration of disease or genetic predisposition? Ann Intern Med 1968; 68: 746-54.

35. Williams RC, Kunkel HG. Rheumatoid factor, complement, and congluti- nin aberrations in patients with subacute bacterial endocarditis. J Clin Invest 1962; 41: 666-75. 36. Sheagren JN, Tuazon CU, Griffin C, Padmore N. Rheumatoid factor in acute bacterial endocarditis. Arthritis Rheum 1976; 19: 887-90. 37. Gutman RA, Striker GE, Gilliland BC, Cutler RE. The immune complex glomerulonephritisof bacterial endocarditis. Medicine 1972; 51: l-25. 38. Steckelberg JM, Melton LJ, llstrup DM, Rouse MS, Wilson WR. Influence of referral bias on the apparent clinical spectrum of infective endocarditis. Am J Med 1990; 88: 582-8.


Date post:	03-Jan-2016
Category:	Documents
Upload:	andikhg
View:	48 times
Download:	4 times

Duke's Criteria of IE

Documents