OBJECTIVEEighteen to twenty-four percent of patients with infective endocarditis (IE) proved pathologically were clinically possible IE by the Duke criteria. In order to improve the sensitivity, the new criteria (trial) for the diagnosis of IE was proposed by Pediatric Cardiology Association of China and Editorial Committee of Chinese Journal of Pediatrics. The aim of this study was to evaluate and compare the value of the new criteria (trial) for the diagnosis of IE with the Duke criteria.

METHODSGroup A consisted of 193 patients proved with IE at autopsy or surgery, where the cases had the results of blood culture and echocardiography data, and Group B had 23 patients with clinical diagnosis of IE in whom evidence of IE was not found at surgery. All the above cases were collected from 15 hospitals. They were analyzed and classified by the new criteria and at the same time by the Duke criteria. The sensitivity and specificity of both criteria for the diagnosis of IE were compared.

RESULTS(1) In Group A, same microorganisms were detected twice in blood culture in 50 patients (25.9%), while 36 patients (18.7%) had only one positive blood culture. Endocardial involvement was found by echocardiography in 165 cases (85.5%), including vegetation in 160 (82.9%), perforation of aortic valve in 4 (2.1%), and partial dehiscence of ventricular septal defect (VSD) patch in one (0.5%). Vegetation appeared oscillating masses in 100 cases (62.5%). One hundred and eighty (93.3%) patients had predisposing heart conditions, and 151 (72.8%) with congenital heart diseases. Fever was revealed in 178 cases (92.2%). Vegetation or perforation of aortic valve was detected in all patients without fever. Heart failure was complicated in 91 patients, 7 of whom had no fever. Vascular phenomena including petechiae and major arterial emboli occurred in 21 and 28 cases, respectively. Among immunologic phenomena, glomerulonephritis occurred in 9, elevated rheumatoid factor in 17/25 and elevated CRP in 51/71. In Group B, the same microorganism was detected in blood culture twice in only 3 patients and 2 patients had one positive blood culture. Vegetation in tricuspid valve was found by echocardiography in one patient. (2) Ninety-four cases (48.7%) of Group A were clinically confirmed IE by the Duke criteria. The diagnosis was made on the basis of two major criteria in 42, one major and 3 minor criteria in 52.14 of 99 as possible IE were excluded by the modified Duke criteria. On the other hand, a definite diagnosis of IE was made in 156 patients (80.8%) by the new criteria. Of them, 94 met with definite criteria of the Duke criteria, 62 (32%) met with echocardiographic evidence of endocardial involvement (major criteria) and two minor criteria. No patient of Group B was clinically definite with the Duke criteria, but one patient was clinically definite with the new criteria (trial). (3) The sensitivity and specificity for the diagnosis of IE were 80.8% and 95.7%, respectively, with the new criteria (trial), 48.7% and 100%, respectively, with the Duke criteria.

CONCLUSIONWith the addition of echocardiographic evidence of endocardial involvement (major criteria) and 2 minor criteria as definite diagnostic criteria, the sensitivity of the new criteria (trial) is superior to that of the Duke criteria, but there is no significant difference in specificity for the diagnosis of IE between the two criteria.