ABSTRACT
INTRODUCTION: This study evaluates the performance of individual and combinations tests used for pediatric tuberculosis diagnosis at a reference center. MATERIALS AND METHODS: Diagnostic test outcomes from children with presumed pulmonary tuberculosis evaluated from January 2005 - July 2010 were compared to a standard diagnosis made by an expert panel of physicians. RESULTS: Presence of at least one sign/symptom, history of contact, or abnormal chest X-ray (aCXR) individually showed the highest sensitivity (85.7%). While the combination of history of contact, at least one sign/symptom, positive tuberculin skin test, and aCXR had low sensitivity of 20%, but the specificity and a positive predictive value were 100%, respectively. The combination of tests used in the International Union Against Tuberculosis and Lung Disease and the Brazilian Ministry of Health systems showed sensitivity of 28.6% and 71.4% and specificity of 95.8% and 97.0%, respectively. CONCLUSIONS: In the absence of a gold standard, the combination of clinical history, tuberculin skin test, and aCXR, as well as the Brazilian scoring system serve as simple, low-cost approach that can be used for pediatric TB diagnosis by first-contact care providers.
Subject(s)
Tuberculosis, Pulmonary , Tuberculosis , Adolescent , Child , Diagnostic Tests, Routine , Humans , Predictive Value of Tests , Tuberculin Test , Tuberculosis/diagnosis , Tuberculosis, Pulmonary/diagnosisABSTRACT
Muitos casos de tuberculose na infância são subnotificados pela dificuldade do diagnóstico na criança. O diagnóstico da tuberculose nessa faixa etária se baseia na presença de contato com adulto bacilífero, associado à prova tuberculínica reatora, sintomas sugestivos de tuberculose e alterações radiológicas. Os escores diagnósticos são estabelecidos através da combinação desses achados. Este artigo é uma breve revisão de diversos escores propostos na literatura para o diagnóstico de tuberculose pulmonar em crianças. Existem vários escores propostos mundialmente, com concordância entre eles de moderada a fraca. Segundo a última revisão sistemática sobre os diversos escores disponíveis no mundo, o sistema que tem mais estudos de validação, com consistentes sensibilidades e especificidades, é o do Ministério da Saúde do Brasil. É imperativa a existência de técnicas ou sistemas diagnósticos mais eficazes, rápidos e de baixo custo para o diagnóstico de tuberculose na infância e com validação no local de sua aplicação. Porém, até lá, sugerimos que os escores diagnósticos sejam utilizados de acordo com as normativas locais, como triagem diagnóstica para uma avaliação criteriosa com o especialista, evitando-se diagnósticos tardios ou equivocados. Em locais onde há escassez de especialistas, profissionais bem treinados poderão utilizar os escores disponíveis para diagnosticar adequadamente a maioria dos pacientes.
Childhood tuberculosis is underreported, because of the difficulty in diagnosing tuberculosis in children, in whom the diagnosis is based on a history of contact with active tuberculosis in an adult, positive tuberculin skin test results, symptoms suggestive of tuberculosis, and radiological alterations. Diagnostic scores are established by evaluating those findings inconjunction. In this review, we discuss a number of such scores that have been proposed.Various scores for the diagnosis of pulmonary tuberculosis in children have been proposed, with only moderate to weak concordance among them. The most recent systematic review of the scores available worldwide showed that the Brazilian National Ministry of Health has sponsored more validation studies, with consistent sensitivity and specificity, than has anyother such system. Techniques or diagnostic systems for childhood tuberculosis that are more efficient, rapid and affordable should be developed and validated for use at the target locale. In the interim, we suggest that diagnostic scores be employed in accordance with local regulations, such as screening through careful evaluation by a specialist, avoiding delayed or inconclusive diagnoses. Where there is a shortage of such specialists, well-trained professionals, using the available scores, will be able to diagnose tuberculosis adequately in the majority of patients.
