A twenty-year retrospective study of pediatric tuberculosis in two tertiary hospitals in Rome

Abstract

Background: Tuberculosis (TB) is among the top 10 causes of child death worldwide. Nevertheless, childhood disease has been neglected by tuberculosis control programs.

Methods: This was a retrospective study of patients < 16 years of age diagnosed with active TB in 2 tertiary hospitals in Rome (Italy), between 1990 and 2009.

Results: Two hundred fourteen cases of active tuberculosis were identified (132 definite, 82 probable). Pulmonary involvement was the most common form (75.5%), followed by lymphadenopathy (15.4%) and central nervous system TB (11%). Fever (51.86%) and cough (40%) were the most common presenting symptoms. A total of 23.4% of children were asymptomatic on admission. Sensitivities of the tuberculin skin test and the quantiferon test were 93.4% and 97%, respectively. Both tests performed in 52 children agreed in 49 cases (94%). Sensitivities for culture, Ziehl-Neelsen staining and polymerase chain reaction were 58%, 25% and 66.3%, respectively. The adult source case was identified in 28% of cases. History of contact with a patient with active TB was associated with pulmonary TB (P = 0.0014), whereas negative history of contact was associated with lymph node (P = 0.0064) and central nervous system TB (P = 0.05).

Conclusions: Our study emphasizes the difficulty in managing children with suspected TB, because the absence of constitutional symptoms cannot exclude TB, and bacteriologic confirmation is the exception. Immunologic diagnosis can be a valuable tool to identify TB-infected children because the quantiferon test showed high sensitivity in all age groups. This is of primary importance because early identification of children with latent tuberculous infection and appropriate chemoprophylaxis represent, to date, the most important tool to reduce the burden of TB.