Case Presentation: A 7-year-old male with medical history of an unrepaired moderate sized restrictive ventricular septal defect and dilated aortic root was admitted for 2-week history of fevers up to 107°F. He had dental surgery approximately 1 month prior and did not receive endocarditis prophylaxis. He was evaluated by his primary physician at the onset of symptoms and laboratory investigations were sent. Remarkable results were WBC 15 k/uL with left shift, CRP 2.2mg/L, and a single blood culture that returned positive and was eventually identified as Streptococcus mitis. The next day, the patient presented to the emergency department where 2 sets of blood cultures were sent and he received a dose of Ceftriaxone. He was well appearing and his only other symptom was headache. His holosystolic regurgitant murmur was stable. Infectious diseases suggested discharge home with planned reevaluation within 24 hours. The patient was discharged home. The next day, both cultures obtained in the emergency department were positive for gram positive cocci on the gram stain and were later identified as Streptoccocus mitis. The patient was admitted. Patient was clinically stable and had no other symptoms other than headache during the admission. He remained afebrile throughout his hospital course. Multiple blood cultures taken during the admission were negative. He had no stigmata of endocarditis and his transesophageal and transthoracic echocardiograms were negative for vegetations or endocardial evidence indicative of endocarditis. However, he was treated empirically for endocarditis with 2 weeks of ceftriaxone and 2 weeks of oral rifampin and amoxicillin. His ventricular septal defect was repaired approximately 2 months after hospitalization, and he has subsequently been well.
Discussion: Infective endocardits is a complex and rare condition in pediatrics that is diagnosed using the modified Duke criteria. Compared to adults, children generally have a more indolent course and striking vascular or immunologic abnormalities are unusual. Endocarditis is now diagnosed early, within a few days of onset of infection, which further precludes the evidence of classic clinical features. When compared to pathologically confirmed case of endocarditis as the gold standard, the sensitivity of the Duke criteria was 80% in the original propositional study. Therefore, the sensitivity of the Duke criteria may be flawed in certain cases. Our patient had 1 major and 2 minor criteria for endocarditis, leading to a diagnosis of “possible endocarditis”. The presence of another minor criteria would have led to the diagnosis of “definite endocarditis”. Our patient was treated and classified as “definite endocarditis” nonetheless given his history of predisposing cardiac condition, sustained fevers, and infection with a typical organism from the Streptococcus viridans group. In the original proposal of the Duke criteria, the authors acknowledged that the criteria does not provide absolute proof that a patient has endocarditis and was not intended for primary management decisions. Hospitalists are in a unique position to recognize and address these areas of diagnostic uncertainly so that quality care and curative treatments can be provided to patients.
Conclusions: The Duke criteria is critical to the diagnosis of infective endocarditis but is not absolute in that it can not replace clinical judgement if highly suggestive features are present.
