Output 1
Assessment of new decentralized childhood TB diagnostic approaches at district health system level
Proper TB diagnosis evaluation is rarely implemented for children at district hospital level and primary health care level in high TB-burden settings. More evidence is needed to recommend the best approach to childhood TB diagnosis in district health systems. This operational research implemented in collaboration with the National Tuberculosis Programs (NTPs) will use an innovative diagnostic approach based on Xpert MTB/RIF (using the newly developed Xpert Ultra cartridge and battery-operated GeneXpert Omni platform) performed on a combination of easy to collect samples (nasopharyngeal aspirate and stool), standardized symptoms screening and chest radiograph interpretation.
Objectives
- This implementation research will assess the impact on childhood TB case detection of decentralizing childhood TB diagnosis compared to pre-intervention status.
- It will also compare the efficiency, feasibility and acceptability of two decentralisation approaches at district hospital and primary health care levels.






Participating countries
and partners:
Output Leader:
Eric Wobudeya – MUJHU Research Collaboration/MUJHU Care Ltd, Uganda
Output project Manager:
Agnes Barbara Kobusingye – MUJHU Research Collaboration/MUJHU Care Ltd, Uganda
Study Population
Children aged below 15 years seeking care at the district hospital and primary health care centres of selected districts.
Study intervention
- Assessment of the existing diagnostic capacities for childhood TB at district hospital and primary health care centres
- Capacity building on microbiological and clinical diagnosis at district hospital and primary health care centres and upgrading of chest radiography (CXR) services at the district hospital level in two selected districts.
- Implementation of two different decentralisation strategies:
- district level decentralisation including screening of children for TB at primary health center level and reference to district hospital for microbiological, clinical and radiological diagnosis
- primary health center (PHC) decentralisation with screening, microbiological and clinical diagnosis done at PHC and reference if needed to district hospital for CXR and further investigations
Output 2
Evaluation of an early TB detection strategy in children with severe pneumonia
TB usually presents as a chronic disease in children, with symptoms such as cough lasting for several weeks. Recent studies have shown that it can also present as a pneumonia, a rapidly progressing disease with acute symptoms including cough, fever, and shortness of breath. In children with pneumonia, tuberculosis is poorly recognized or diagnosed late, which contributes probably to increased mortality.
Objectives
- Output 2 is a diagnostic clinical trial evaluating whether an early tuberculosis detection strategy with the Xpert Ultra test on respiratory (nasopharyngeal aspirate) and stool samples has an impact on mortality in children with severe pneumonia.






Implementing countries
Output Leader:
Olivier Marcy, Université de Bordeaux, France
Output project manager:
Aurélia Vessière, Université de Bordeaux, France
Study population
3800 children aged below 5 years with severe pneumonia.
Study intervention
All children will be treated for pneumonia with the standard of care recommended by WHO which includes intravenous antibiotics and oxygen therapy if needed. Half of children will have immediately upon admission an Xpert test performed on nasopharyngeal aspirate and stool samples. If the result is positive, anti-tuberculosis treatment will be started immediately.
Expected impact
If the clinical trial shows that the strategy contributes to reduce mortality, it could be integrated in recommendations for management of childhood pneumonia and tuberculosis diagnosis by WHO.
Output 3
Validation of diagnostic tools and algorithms in HIV-infected and severely malnourished children with presumptive tuberculosis
HIV-infected children and children with severe acute malnutrition are highly vulnerable population with regards to tuberculosis. Immunodepression due to HIV or to malnutrition exposes them to a higher risk of developing tuberculosis disease and higher mortality. In these children, diagnosis of tuberculosis is challenging because they do not present a lot of symptoms, or their symptoms are hard to differentiate form other infections, especially when they are HIV-infected.
Objectives
- To improve diagnosis of tuberculosis, it is necessary to develop and validate specific algorithms and diagnostic tools in these children. Algorithms are standardized decision rules for diagnosis and treatment initiation. Output 3 is a diagnostic study aiming to develop and validate adapted algorithms for these children.






Implementing countries
Output Leader:
Olivier Marcy, Université de Bordeaux, France
Output project manager:
Aurélia Vessière, Université de Bordeaux, France
Study population
- 550 HIV-infected children with suspected tuberculosis
- 720 hospitalized children with severe acute malnutrition
Diagnostic tools and tests evaluated in the study
- Xpert Ultra done on respiratory (nasopharyngeal aspirate, sputum), and stool samples
- Chest radiography
- Abdominal ultrasonography
- Clinical features validated in a previous study in HIV-infected children
Expected impact
If the study shows that developed algorithms have a good sensitivity to detect tuberculosis in severely malnourished or HIV-infected children, they could be integrated into recommendations for tuberculosis diagnosis by WHO.
Output 4
Identification of optimized, suitable, and affordable specimen processing and collection methods for childhood TB diagnosis in resource limited countries
This output involves microbiological and technological optimization work to identify and test simple and affordable specimen processing and collection methods for childhood TB diagnosis that can be deployed at low health care level in resource-limited countries. Indeed, stool is a very promising sample for diagnosis of TB in children but it requires to be processed before testing with Xpert MTB/RIF assay to remove factors that inhibit the molecular technique. This processing needs to be optimised and simplified to be used at low health care level. Nasopharyngeal aspirates, another promising method for specimen collection in children, requires equipment (aspirator) and device (mucus extractor) that can be expensive and difficult to procure. Therefore more work is needed to ensure secured access of robust and affordable equipment and devices for nasopharyngeal aspirate for limited resource countries.
Objectives
- This output aims to optimize stool specimen processing for Xpert MTB/RIF testing and nasopharyngeal aspiration method for childhood TB diagnosis in resource-limited countries at low health care level.






Output Intervention
Generation of a centrifuge-free stool processing method for Xpert MTB/RIF testing
- Optimized stool processing development and evaluation phase
- Generation of simple and affordable kits for stool processing
- Demonstration phase for Xpert MTB/RIF testing at low health care level
Optimization of nasopharyngeal aspirate sample collection
- Screening of the market for aspirator and mucus extractor
- Testing and optimisation of existing material or development of new material
Development of guidelines and training material for stool processing and optimized nasopharyngeal aspirate
Output Leader:
Maryline Bonnet, Institut de Recherche et Développement, UMI233 TransVIH MI, France
Output Project Manager:
Manon Lounnas, Institut de Recherche et Développement, MIVEGEC, UMR UM, CNRS, France
Technical partners:
Ryan Carroll, CAMtech, Uganda
Sylvain Godreuil, Institut de Recherche et Développement, MIVEGEC, France
Juliet Mwanga, Epicentre, Uganda
Output 5
Evaluation of cost-effectiveness of the proposed diagnostic approaches
Cost effectiveness and budget impact analyses are needed to evaluate the long-term impact of improving pediatric TB diagnostic, guide health authorities’ decisions and support the implementation of TB-speed approach in resource limited settings.
Objectives
- To evaluate the cost-effectiveness and budget impact of the TB-Speed interventions
- To assess the market of paediatric diagnostics tools
- To develop a forecasting tool for practical budget planning by NTPs






Output Intervention
Cost-effectiveness and budget impact analyses
- Development of a new model to evaluate cost-effectiveness of various diagnosis tools and approaches
- Collection of direct costs in the participating countries throughout the project implementation
- Cost-effectiveness and budget impact analyses based on project data
Market assessment and market impact analysis
- Assessment of the TB diagnostics market
- Modelling the impact on TB diagnostic tools market
Development of a forecasting and costing tool
Output Leader:
Eric Ouattara, Université de Bordeaux, France
Technical partners:
Pete Dodd, University of Sheffield, United Kingdom
Output 6
Dissemination, communication and stakeholders’ engagement
This includes communication on the project progresses, advocacy towards the relevant stakeholders, and dissemination of results to support future scaling-up.
Objectives
- To inform international/national TB stakeholders about the TB-Speed project
- To ensure policy change and further adoption of the TB-Speed diagnostic approach on a larger scale if supported by project’s results
- To increase project visibility among policy makers, scientific and health community
- To raise awareness about paediatric TB among health community and civil society






Output Intervention
- Communication plan and stakeholder’s engagement
- Awareness campaign targeting the citizens and the different communities
- Dissemination campaign targeting the scientific and health staff community
- Advocacy campaign targeting the Health authorities, policy makers and international donors
Output Leader:
Maryline Bonnet, Institut de Recherche pour le Développement, France
Project Communication Manager:
Nicolas Koskas, Université de Bordeaux, France