- A scaled-up study of the intervention within a full RCT should be conducted.
- Improvements in clinical outcomes such as treatment outcomes, relapse, disability, and general lung health should be documented and analysed in future studies.102–104
- Efforts to increase the generalisability of the population enrolled in such a future study should be undertaken.
- Use of components in isolation or a package with a smaller number of components may reduce administration time and increase feasibility – this could be explored in future evaluations.
- Groups attempting similar exercises might consider alternative study designs, for example, a stepped-wedge design.105
- To recruit people at greatest risk of non-adherence, it may be effective to randomise at clinic- rather than individual-level, but the potential ethical implications of this approach require exploration.
- Future investigators may want to consider if adherence (with all the associated issues of measurement in an NHS setting) is the right outcome measure. Potential alternatives include measures of well-being and stigma.106,107
Manualised package intervention to achieve treatment adherence in people with tuberculosis: the IMPACT pilot cluster-RCT · Research recommendations
Research recommendations
Attribution. Marc Lipman; Ibrahim Abubakar; Colin NJ Campbell; Caroline Clarke; Andrew J Copas; Patricia Costello et al. (2026). Manualised package intervention to achieve treatment adherence in people with tuberculosis: the IMPACT pilot cluster-RCT. National Institute for Health and Care Research. Copyright © 2026 Lipman et al. Licensed under CC BY 4.0. NCBI Bookshelf NBK621906. CC BY 4.0. Original source ↗. Text reproduced from the Bookshelf XML without changing wording; figures, images and supplements are not reproduced.