Measurement of Tuberculosis treatment adherence: A methodological scoping review of observational studies

Despite extensive conceptual development, the definition and measurement of Tuberculosis (TB) treatment adherence remain inconsistent. This ambiguity obscures what is being assessed and compared across studies. This scoping review examined how adherence is labelled, defined, and operationalised in observational studies of drug-susceptible TB, assessed consistency with contemporary frameworks, and identified methodological gaps. We conducted a scoping review of observational studies on drug-susceptible TB identified from eight databases up to February 2026. Data were extracted on terminology, measurement approaches, and operational definitions. Content analysis was used to map labels and indicators to the constructs they implicitly presuppose. One hundred and three studies were included. Treatment-taking behaviour was mostly assessed through clinician assessment or self-report. Explicit definitions were uncommon, and operational criteria served as de facto definitions. Multidimensional measures were frequently reduced to binary or ordinal outcomes, and most studies relied on cross-sectional designs, reinforcing static representations of a dynamic process. Thus, TB adherence research faces a fundamental problem of construct conflation: the same label is used to describe behaviours that are conceptually distinct, limiting comparability and weakening the interventions' evidence base. Standardising reporting requirements would strengthen the evidence base for programmes addressing structural and social drivers of adherence.