Measuring the accuracy of a point system to diagnose tuberculosis in children with a negative smear or with no smear or culture

Abstract

In Brazil, a scoring system was adopted to diagnose tuberculosis in childhood. This study determined the accuracy in diagnosing tuberculosis in children with either a negative smear or with no smear or culture conducted in a reference center in João Pessoa Paraíba - Brazil. It is a phase III validation study, using a cross-section design. The study population consisted of 167 patients attending the outpatient clinics suspected of having tuberculosis. The reference standard for the diagnosis of tuberculosis was a blind and independent review of the medical records, radiology and tuberculin test by two experts. Of the 167 patients, 60 were considered to have tuberculosis (by the reference standard diagnostics). The results for the scoring system with the cut-off of 30 points were: sensitivity 78.57% (95%-CI: 65.56-88.41%), specificity 69.16% (95%-CI: 59.50-77.73%), positive predictive value (PPV): 57.14% (95%-CI: 45.35-68.37%), negative predictive value (NPV): 86.05% (95%-CI: 76.89-92.58%), likelihood ratio (+): 2,55, pre-test probability: 34.36%, and post-test probability (+): 57.14%. This supports the current recommendation for the use of this scoring system in Brazil and similar sites with the cut-off of 30 points. However, as the discriminatory power of the point scoring system may vary across settings, it would be advisable to replicate this phase III study in different settings.

The clinical and molecular diagnosis of childhood and adolescent pulmonary tuberculosis in referral centers

“…Also, in adolescents with ITB, almost every population had DSS value of 30 or greater, and a little more than half the population had detectable Xpert results. These findings support the validation of DSS obtained in other studies and, at the same time, guide the careful use of Xpert, preceded by DSS, in children and adolescents treated at referral centers 3,17,18 . DSS has been standardized for over a decade in Brazil.…”

Latent tuberculosis infection and tuberculosis in children and adolescents

“…For a score of 30, sensitivity and specificity were 78.6% and 69.2%, and a score of 40, they were 48.2% and 87.9%, respectively. ( 17 )…”

Prevalence of culture positive Tuberculosis and utility of a clinical diagnostic tool for the diagnosis of Tuberculosis among HIV Infected Children attending HIV/AIDS Care and Treatment in Dodoma Municipality, Central Tanzania

“…Findings such as generalized lymphadenopathy, persistent fever and weight loss are all seen in HIV/AIDS (Richardson et al 2008;Mulenga et al 2015) as well as in TB. Another retrospective study conducted in Brazil showed that among 167 children suspected to have Tuberculosis, the sensitivity and specificity of the TB clinical diagnostic tools used was 78.57% and 69.16% respectively (Cartaxo et al 2014b). In this and several other studies it was also found that different TB clinical diagnostic tools have variable performance in different patient population and settings, adding to the complexity in developing a universal tool which can reliably diagnose Tuberculosis in most TB endemic regions (Edwards et al, 2007).…”

“…To increase the diagnostic rate of paediatric Tuberculosis in areas with limited diagnostic capacity where TB is highly prevalent, efforts have been made to develop clinically based diagnostic tools to be used for the diagnosis of Tuberculosis in children, but studies investigating their diagnostic accuracy compared to gold standard tests have reported inconsistent findings in different settings and different patient populations (Cartaxo et al 2014a). A study reported from Salvador Brazil on the utility of the TB clinical score tool endorsed by the Brazilian Ministry of Health for diagnosis of Tuberculosis in children found that the use of this score had a 30% chance of misdiagnosing non TB conditions as Tuberculosis (Coelho Filho et al, 2011).…”