Treatment-decision score for children with HIV and suspected tuberculosis

Organisation: Position: Deadline Date: Location:

A new treatment-decision score for the diagnosis of tuberculosis (TB) in children with HIV performed well, and when used in an algorithm may facilitate prompt treatment decision making for children at a high risk for mortality, according to data.

Clinical assessments, chest radiography, Quantiferon Gold In-Tube (QFT), abdominal ultrasonography, and sample collections for microbiology, including Xpert MTB/RIF (Xpert) were performed on 438 children with HIV who were suspected to have TB in Burkina Faso, Cambodia, Cameroon, and Vietnam. Using logistic regression 4 TB diagnostic models were developed: (1) all predictors included, (2) QFT excluded, (3) ultrasonography excluded, and (4) QFT and ultrasonography excluded. These models were internally validated using resampling and scores built on the basis of the model with the best area under the receiver operating characteristic curve and parsimony.

Of the 438 children enrolled, 57.3% had TB, including 12.6% with culture- or Xpert-confirmed diagnosis. The 4 models included the following characteristics, which were noted to improve model prediction: Xpert positivity, fever lasting >2 weeks, unremitting cough, haemoptysis and weight loss in the prior 4 weeks, contact with a patient with smear-positive TB, tachycardia, military TB, alveolar opacities, and lymph nodes on chest radiograph, together with abdominal lymph nodes on the ultrasound and QFT results. For models 1, 2, 3, and 4 the areas under the receiver operating characteristic curves were 0.866, 0.861, 0.850, and 0.846, respectively.

The score developed using model 2, demonstrated a sensitivity >90% in the case-control population. Results also demonstrated that model 2 had a diagnostic sensitivity of 88.6% and a diagnostic specificity of 61.2%; it also had a positive predictive value of 77.4%, and negative predictive value of 78.1%. Further, when applied to the overall cohort, the score accurately identified 85.7% of children with TB if missing data was assumed negative; this increased to 90.8% when missing data was assumed positive.

Over-estimation of the models’ diagnostic performance is a possibility due to incorporation bias resulting from the lack of a reference standard for childhood TB, independent of candidate predictors. A second limitation to the study is missing data for the predictors in almost one-quarter of the enrolled children, mostly younger children with severe clinical status. Finally, the study eligibility criteria differed from World Health Organisation (WHO) criteria for investigation of TB; therefore, this score is not directly applicable to children presenting with the WHO criteria.

The study is strengthened by using data from 4 different countries, which increases generalisability. Further, internal validation demonstrated that the models would provide good predictions. Investigators concluded that the high sensitivity and algorithmic approach of the score should, “enable rapid treatment decision in children with presumptive tuberculosis.” Researchers cautioned that further external validation is required to validate the scoring system and overall approach before its clinical usefulness can be fully confirmed.

Abstract
Background: Diagnosis of tuberculosis should be improved in children infected with HIV to reduce mortality. We developed prediction scores to guide antituberculosis treatment decision in HIV-infected children with suspected tuberculosis.
Methods: HIV-infected children with suspected tuberculosis enrolled in Burkina Faso, Cambodia, Cameroon, and Vietnam (ANRS 12229 PAANTHER 01 Study), underwent clinical assessment, chest radiography, Quantiferon Gold In-Tube (QFT), abdominal ultrasonography, and sample collection for microbiology, including Xpert MTB/RIF (Xpert). We developed 4 tuberculosis diagnostic models using logistic regression: (1) all predictors included, (2) QFT excluded, (3) ultrasonography excluded, and (4) QFT and ultrasonography excluded. We internally validated the models using resampling. We built a score on the basis of the model with the best area under the receiver operating characteristic curve and parsimony.

Results: A total of 438 children were enrolled in the study; 251 (57.3%) had tuberculosis, including 55 (12.6%) with culture- or Xpert-confirmed tuberculosis. The final 4 models included Xpert, fever lasting >2 weeks, unremitting cough, hemoptysis and weight loss in the past 4 weeks, contact with a patient with smear-positive tuberculosis, tachycardia, miliary tuberculosis, alveolar opacities, and lymph nodes on the chest radiograph, together with abdominal lymph nodes on the ultrasound and QFT results. The areas under the receiver operating characteristic curves were 0.866, 0.861, 0.850, and 0.846, for models 1, 2, 3, and 4, respectively. The score developed on model 2 had a sensitivity of 88.6% and a specificity of 61.2% for a tuberculosis diagnosis.
Conclusions: Our score had a good diagnostic performance. Used in an algorithm, it should enable prompt treatment decision in children with suspected tuberculosis and a high mortality risk, thus contributing to significant public health benefits.

Authors
Olivier Marcy, Laurence Borand, Vibol Ung, Philippe Msellati, Mathurin Tejiokem, Khanh Truong Huu, Viet Do Chau, Duong Ngoc Tran, Francis Ateba-Ndongo, Suzie Tetang-Ndiang, Boubacar Nacro, Bintou Sanogo, Leakhena Neou, Sophie Goyet, Bunnet Dim, Polidy Pean, Catherine Quillet, Isabelle Fournier, Laureline Berteloot, Guislaine Carcelain, Sylvain Godreuil, Stéphane Blanche, Christophe Delacourt

Infectious Disease Advisor material Pediatrics abstract

Receive Medical Brief's free weekly e-newsletter



Related Posts

Thank you for subscribing to MedicalBrief


MedicalBrief is Africa’s premier medical news and research weekly newsletter. MedicalBrief is published every Thursday and delivered free of charge by email to over 33 000 health professionals.

Please consider completing the form below. The information you supply is optional and will only be used to compile a demographic profile of our subscribers. Your personal details will never be shared with a third party.


Thank you for taking the time to complete the form.