Diagnosis of tuberculosis in children

Background

Under diagnosis and subsequent under treatment of tuberculosis (TB) in children is a major public health issue in many parts of the world. To provide more structured guidance to primary health care staff assessing children with presumptive TB, the World Health Organization has recently recommended the use of two integrated treatment decision algorithms (for contexts with and without radiography). These algorithms allocate evidence-based scores to clinical and radiological features to allow clinicians to make decisions regarding the initiation of TB treatment in children in children < 10 years of age with presumptive pulmonary TB. The WHO recommendation was formulated as an interim, conditional recommendation and the WHO has called for the generation of evidence through external validation and other studies, to inform a future review of the recommendation on the diagnosis, management of TB in children.

Aim of the study and objectives

The overall aim of the study is to evaluate the diagnostic performance, describe the diagnostic cascade and assess the feasibility of implementing the new 2022 WHO treatment decision algorithms for pulmonary TB in children under 10 years old under programmatic conditions.

The study had several secondary objectives related to the use of the algorithms but also to assess the added value of TB-LAM in HIV-negative children and of Point-of-care ultrasound (POCUS) for the diagnosis of TB in children.

Primary objectives

1. To assess the diagnostic accuracy of the 2022 WHO treatment decision algorithms against a reference standard. 
2. To document the diagnostic cascade using the 2022 WHO algorithms (frequency of children undergoing each step of the algorithms, started on TB treatment etc). 
3. To assess health workers and patients/caregivers’ perspectives on the 2022 WHO treatment algorithms including challenges and enablers to the introduction of the recommended algorithms. 

Secondary objectives

1. To assess clinical evolution, TB treatment initiation, and mortality at 2 months of follow-up.
2. To assess the diagnostic performance of TB-LAM in HIV-positive and HIV-negative children (only specific study sites).
3. To assess the diagnostic performance of POCUS (only specific study sites).
4. To compare the actual TB treatment algorithms’ recommendations with the potential recommendation that would have been reached if chest X-ray or laboratory results were not available.
5. To assess the number and the proportion of children potentially eligible for a 4-month treatment regimen.
6. To report changes on paediatric aggregated case notification rates 12 months prior to, and 12 months following the implementation of the algorithms.

Countries 

Nigeria, Niger, Guinea, Uganda, South Sudan

Status

Patient recruitment and follow-up have been completed at the Nigeria site. The other sites - Niger since November 2023, Guinea & Uganda since December 2023, South Sudan, since January 2024 - are open and currently recruiting. 

Tentative end Date 

December 2025

Epicentre role  

Study Coordination  

Partners 

Médecins Sans Frontières, France, Médecins Sans Frontières, Belgium, Médecins Sans Frontières, Spain, Ministry of Health, Nigeria, Ministry of Health, Niger, Ministry of Health, Guinea, Ministry of Health, Uganda, Ministry of Health, South Sudan, University of Tübingen, Germany, ANRS, France

Methods

Performance of the algorithms (accuracy and cascade of care)

An observational prospective diagnostic multi-country study will be conducted to assess the diagnostic performance (accuracy and cascade) of the 2022 WHO treatment decision algorithm. Children included in the study will be followed up to two months after the initial algorithm assessment. Study sites and countries will be: Niger (Madarounfa), Nigeria (Maiduguri), Guinea (Conakry), South Sudan (Malakal), and Uganda (Mbarara).

Children presenting with signs or symptoms suggestive of pulmonary TB (presumptive pulmonary TB), age less than 10 years and whose representative provided informed consent will be invited to participate. Depending on the study site, the algorithms with or without chest X-ray will be assessed on either ambulatory and/or hospitalized children.

Some sites will only include malnourished children (Niger and Nigeria) or children living with HIV (Guinea). Participants will be included for a minimum of 6 months. The inclusion period is expected to last 6 to 12 months, aiming for a sample size of minimum 500 in Uganda, 230 in South Sudan, 600 in Niger, 350 in Nigeria, and 340 children in Guinea.

Process evaluation (documentation, feasibility and acceptability study)

In all five countries, the research team will document how the new 2022 WHO algorithms are implemented. The researchers will describe in detail each step of the intervention from the delivery of the training to the patient’s discharge. This will allow to highlight needs, priorities and challenges encountered when using the new TB algorithm across different countries.

A mixed-methods multi-country study will be conducted in three study sites (Niger, Guinea, and Uganda) to assess the process of the intervention, including the feasibility of implementing the 2022 WHO algorithm and the health care workers and patients/caregivers’ perspectives on it.

Caregivers of children with presumptive TB or with TB, key informants working/living with children with TB, health managers, policy makers and health workers 18 years and above will be purposively chosen. Data will be collected through in-depth and semi-structured interviews, focus groups discussions, health workers’ diaries and non-participant observations.

The process evaluation will follow the Medical Research Council (MRC)’s process evaluation framework adapted in 2010 for complex interventions. It will be conducted in three phases: (1) Before implementing the algorithms to analyse the contextual factors that may influence their implementation; (2) During training and within 6 months of implementation to analyse the impact mechanisms of the implementation, how the new practice of algorithmic diagnosis is perceived and implemented; (3) Within 12 months post implementation to analyse the impact mechanisms of the implementation over time.

Expected results

This study will contribute to a larger pool of evidence allowing the evaluation of the WHO recommended treatment decision algorithms. The study results will be shared with the WHO to support the revision of their algorithms. This will ensure that the results contribute to build the future WHO recommendations.

The mixed methods study will provide important information on the feasibility and acceptability of the algorithms that will be valuable for the implementation of the algorithms in other contexts and countries. This study will also assist in site specific implementation and adaption and may lead to improved acceptability amongst stake holders.

Evidence generated from this study is particularly pertinent as it will contribute ‘real-world’ data from a wide range of populations and contexts.

Upon completion of the study, the results will be shared as soon as possible with all partners and public health authorities.

Hence, the results are expected to improve the diagnosis process and the management of TB in children, both at the study sites by integrating the diagnosis process in routine practice, and on wider scale through the revised WHO recommendations.