Diagnosing pediatric tuberculosis presents a challenge because of the relatively low proportion of children with gastric cultures positive for Mycobacterium tuberculosis and because of the imprecision of chest radiography in detecting mediastinal or hilar adenopathy. Genetic amplification techniques, such as polymerase chain reaction (PCR) assays, that test sputum specimens have been shown to accurately diagnose tuberculosis in adults. However, little is known about the effectiveness of PCR testing in diagnosing pediatric tuberculosis. Neu and associates evaluated the effectiveness of a commercial PCR assay in detecting M. tuberculosis in the gastric aspirate cultures of children. They also compared the sensitivity of PCR testing with that of computed tomography (CT) in assessing the degree of adenopathy. In addition, they examined the applicability of the Centers for Disease Control and Prevention's (CDC) clinical case definition for tuberculosis in children with negative cultures (see the accompanying table).

Children under 15 years of age with signs of tuberculosis, including an abnormal chest radiograph, who were admitted to the emergency department of a large New York hospital were eligible for the study. All patients underwent a tuberculin skin test, radiographic studies, gastric aspirates for culture and serum sampling for the presence of M. tuberculosis DNA using the Amplicor test. In addition, a CT scan was obtained in all patients who had an equivocal radiograph or in whom clarification was needed.

Twenty-seven children were enrolled in the study. Of these, 19 had positive skin tests, and all had some abnormality on screening chest radiographs. Nineteen eventually had CT scans, and mediastinal or hilar adenopathy was clearly detected in six patients with equivocal or absent adenopathy on chest radiographs. Two children had previously undetected infiltrates that resolved with antibiotic therapy. All gastric aspirate specimens were acid-fast smear-negative. M. tuberculosis was cultured from five specimens obtained from two children; however, PCR testing detected M. tuberculosis DNA in only one of these five specimens. Sixteen patients met the CDC criteria for tuberculosis.

The authors conclude that CT scans improve the precision of the diagnosis of adenopathy and should be considered as a guide to therapy. The sensitivity of PCR testing of gastric aspirates in children remains inadequate, and it is not recommended. The CDC clinical case definition appears to be too restrictive and should be augmented to include immunosuppressed children who have negative skin tests or those with minimal or no radiographic changes during therapy. In addition, more weight should be given to a known epidemiologic link with an infectious source case.