Fig. 1

Conceptual overview of the model, without screening for subclinical TB (main panel) and with screening (inset). Within each cohort, we identify two subgroups whose outcomes may be affected by TPT: those with a latent infection that will, if untreated, progress to active TB at some point in the future (“latent progressors”), and those who have prevalent subclinical TB (asymptomatic but detectable through additional testing such as chest X-ray or sputum testing) that will progress to active rather than resolving spontaneously (“subclinical progressors”). These subgroups are further stratified (not shown) by the drug resistance present at t = 0 and by CD4 count (people living with HIV) or age (household contacts). These latent and subclinical progressor groups are followed through the following steps: (1) testing to detect and treat subclinical TB in some modeled scenarios [inset], (2) TPT (if given), and (3) TB treatment (if subclinical TB is diagnosed, or if progression to active TB occurs despite the intervention). For the TPT and treatment steps, the probabilities of cure, and of acquiring new or additional resistance if not cured, depend on the regimen and the drug resistance present. When testing for subclinical TB is required prior to TPT (Inset), it reduces access (such that a proportion of individuals with latent and subclinical TB receive neither the additional testing nor TPT), but a proportion of the remaining individuals with subclinical TB are identified and immediately treated for active TB. Proportions in the figure are not to scale. All modeled scenarios include symptom screening to detect and treat active TB, so the outcomes of those with active (symptomatic) TB at enrollment are not compared between scenarios