The DECIDE-TB Trial; Validation of Treatment Decision Algorithms for Childhood Tuberculosis

Validation of Treatment Decision Algorithms for Childhood Tuberculosis at District Health Care Levels in Mozambique and Zambia - the Decide-TB Cluster-randomized Pragmatic Trial

The Decide-TB project aims to generate evidence for the implementation of a comprehensive Treatment Decision Algorithms (TDA) based approach for TB in children living in high TB burden and resource-limited countries, at District Hospital (DH) and Primary Health Centre (PHC) levels, and to facilitate the integration of this evidence within practices and policies.

This programmatic pilot led by the National TB Programs (NTP) will test a TDA-based approach integrating TB screening, diagnosis, treatment decision-making, and disease severity assessment for shorter treatment eligibility, for use at a lower level of healthcare. This TDA-based approach will be evaluated in a hybrid effectiveness implementation study based on a pragmatic stepped wedge cluster-randomized trial. The Decide TB project will be implemented at the district level, targeting five districts in each country. Each cluster in a district will be made up of one district hospital and six primary health centers. The study will develop a Clinical Decision Support System (CDSS) to operationalize the use of TDAs, and strengthen District Health Information Systems (DHIS2) to collect individual data, which will contribute to monitoring and evaluation, clinical mentoring, and supervision by the country's NTPs.

The Decide-TB trial is a pragmatic cluster-randomized study utilizing a stepped wedge design to provide scientific evidence on a comprehensive TDA-based approach for TB screening, diagnosis, and treatment in children under 15 years at low healthcare levels in Zambia and Mozambique.

The trial's primary objective is to evaluate the effectiveness, feasibility, implementation, acceptability, costs, cost-effectiveness, and adoption of a TDA-based approach for childhood TB screening, diagnosis, and treatment decision-making under programmatic conditions at District Hospitals (DH) and Primary Health Centres (PHC) levels in these countries.

There are five specific objectives corresponding to the trial's research components:

1. To assess the effectiveness of the comprehensive TDA-based approach in increasing TB case detection in children as compared to the Standard of Care (SOC), and in providing good quality TB diagnosis and treatment decision (diagnostic accuracy and reliability of treatment decision for TB and shorter treatment).
2. To describe the implementation and the feasibility of using the comprehensive TDA-based approach, including associated digital tools, and to identify contextual determinants influencing implementation and contribute to improved implementation/adaptations throughout intervention delivery.
3. To assess preferences, acceptability, and perceived feasibility of using the comprehensive TDA-based approach, including associated digital tools, among end-users, beneficiaries, and key stakeholders.
4. To assess the costs from the health system and the beneficiary (parents/caregivers of children) perspective, the cost-effectiveness of using the comprehensive TDA-based approach, including associated digital tools, and the budget impact of scaling up the intervention.
5. To assess the factors and stakeholders that support or constrain the adoption of the comprehensive TDA-based approach as health policy at district level.

The intervention will be implemented as a programmatic pilot, following the National TB Programs' decision in Mozambique and Zambia to adopt a TDA-based approach in line with World Health Organisation's (WHO) conditional recommendations. Additionally, both the standard of care and the intervention will be implemented in a non-randomized district to document diagnostic accuracy throughout the trial. The Decide-TB trial is a hybrid effectiveness-implementation trial (type 2), assessing both the clinical intervention's effectiveness and the feasibility and utility of the implementation strategy.

Diagnostic accuracy of both the standard of care and the intervention will be evaluated in an additional district, which will transition to the intervention phase alongside the last district selected per the stepped wedge design. Aggregated data will be retrospectively collected over 12 months from all participating districts and health facilities.

The Decide-TB trial endpoints include effectiveness, acceptability, implementation, health economics, and health policy. Effectiveness endpoints include the proportion of children who started TB treatment, the time it takes to make a treatment decision, and the consistency of TDA results with final treatment decisions, among other factors. Acceptability endpoints consider user preferences and local social value, whereas implementation endpoints evaluate the practicality, fidelity, contextual aspects, and sustainability of the TDA-based method. Health economics endpoints include cost assessments, cost-effectiveness, and budget impact of scaling up the TDA-based approach vs the standard of care, while health policy endpoints look at key stakeholders' roles, practices, and policy translation processes.

Inclusion Criteria:

The effectiveness assessment will be conducted using aggregated or individual data from direct beneficiaries of the intervention:

• All sick children aged below 15 years entering the selected health facilities (DH and PHC) at either outpatient (OPD) or inpatient (IPD) departments, including children from high-risk groups, as well as children identified as contact of TB cases through community- or facility-based household contact tracing.
• Children with presumptive TB.

The WHO definition of presumptive TB will be used, as defined in the 2022 WHO Operational Handbook, namely: children are classified as having presumptive TB if they have unremitting symptoms lasting more than 2 weeks (any one of cough, fever, not eating well or anorexia, weight loss or failure to thrive, fatigue, reduced playfulness or decreased activity) .

The definitions of presumptive TB have been adapted locally for the programmatic pilot. All children with presumptive TB as defined locally will be considered in the intervention and in secondary effectiveness and sub-group analyses.

High-risk group will be defined using the definition in WHO-suggested TDAs A&B as children younger than 2 years, CLHIV or children with SAM. CLHIV will be defined per national testing strategy including positive PCR test for children below the age of 18 months. Children will be considered to have SAM (and thereby be eligible for the TB-Speed SAM algorithm) using WHO criteria. These include being <5 years with a weight-for-height Z score (WHZ) < -3 SDs or mid-upper arm circumference (MUAC) < 115 mm (in children over 6 months) or clinical signs of bilateral pitting oedema, and being aged ≥5 years with a body mass index (BMI) for age Z-score < -3SD.

Exclusion Criteria:

There will be no exclusion criteria for the programmatic pilot: all children will be offered the intervention.