Emergency department presentation of children with tuberculosis.

OBJECTIVES

The objectives were to describe emergency department (ED) presentations of children with tuberculosis (TB) disease and assess the utility in children of TB screening tools developed for adults.

METHODS

Children at most 18 years old with confirmed or probable TB seen at the Children's TB Clinic from 2005 to 2009 who were initially evaluated in the ED for symptoms compatible with TB in the preceding month were included. TB was classified as microbiologically confirmed disease or probable TB disease, as defined by the World Health Organization.

RESULTS

Sixty children (29 with confirmed TB and 31 with probable TB) were identified after presentation to the ED, representing 35% of all children diagnosed with TB at the two hospitals during this interval. Eighty-eight percent were previously healthy. Fifty-five percent were Hispanic, 30% were black or African American, 12% were Asian, and 3% were white. Forty-four (73%) had intrathoracic disease (37 pulmonary parenchymal or pleural disease, four miliary disease, two endobronchial, one pericarditis). Sixteen (27%) had extrathoracic disease (eight meningitis, five cervical lymphadenopathy, two gastrointestinal, one interstitial keratitis), 11 of whom also had abnormal chest radiographs, including all eight children with TB meningitis. Most (76.7%) were diagnosed at the time of their first ED visit or during their first hospital admission, 12% after their second ED visit, 10% after their third ED visit, and one patient after six ED visits to various facilities. In 33 case (55%), the diagnosis was suspected in the ED because of epidemiologic risk factors (15), radiographic evaluation (11), or symptoms (7). Hemoptysis (12%) and night sweats (10%) were uncommon. Neither cavitary lesions (seen in two children) nor apical lesions (seen in 42%) predominated. The five screening tools validated for adults with pulmonary disease were 77% to 98% sensitive in identifying children with intrathoracic TB and 50% to 100% sensitive for extrathoracic TB.

CONCLUSIONS

The point of entry to health care for many children with TB is the ED. The more protean manifestations of TB in children can decrease the utility of screening tools developed to identify adults with TB. While TB in adults often is a microbiologic diagnosis, childhood TB often is an epidemiologic diagnosis. Therefore, questioning caregivers about TB risk factors in the family may identify a higher percentage of children with possible TB.