Unlike other major infectious diseases such as HIV and malaria, global progress against tuberculosis (TB) ? now the world's leading agent of infectious mortality ? has been disappointingly slow. A major reason for this stagnation in global TB control efforts is that the current approach consists largely of broad guidelines that are difficult to implement in high-burden communities. What we need is an approach that can help prioritize between feasible, targeted TB control interventions in a way that uses limited resources to maximize impact on disease burden and transmission at the community level. We propose to address this knowledge gap by constructing a detailed snapshot of prevalent TB in a high-burden community: identifying every detectable case of active TB (including those with subclinical TB or not seeking care), using whole-genome sequencing to detect transmission networks in the community, and characterizing every prevalent case in terms of what specific targeted interventions could diagnose and treat that case sooner. These data will enable us to estimate intervenable fractions ? the proportion of prevalent TB (and of TB transmission potential) that could have been captured at an earlier time point by each of an array of feasible, targeted interventions. We will combine these data with data on cost and efficiency (ability to capture people with TB versus controls without TB) to help identify what combinations of TB interventions have greatest potential to reduce TB prevalence at the community level for a given budgetary outlay. We will also construct an agent-based mathematical model to help project how these findings might pertain to other settings with different epidemiological and economic characteristics. This research will provide a critical evidence base for policy-makers seeking to prioritize TB interventions with greatest impact on population-level disease burden, and it will also inform what interventions should be prioritized in the next generation of clinical trials.
In Specific Aim 1, we will not only estimate the prevalence of TB in a well-circumscribed urban Ugandan community, but also characterize that prevalence to a degree never before achieved at the population level ? including novel markers of infectiousness (e.g., cough frequency), transmission between cases that may not even have symptoms (in a second wave of case-finding), and cases' likelihood of being engaged by targeted interventions.
In Specific Aim 2, we will estimate and compare the intervenable fraction of TB associated with numerous targeted interventions, thereby helping to prioritize those interventions in terms of their likely community-level impact.
In Specific Aim 3, we will link these epidemiological priorities with cost-effectiveness analysis and agent-based simulation modeling, to guide priorities using constrained economic resources and to generalize such findings to a variety of epidemiological settings. This innovative approach will allow us to take an urgently-needed step, from simply measuring the burden of TB disease and offering ?one-size-fits-all? guidelines, to understanding how we can prioritize feasible, targeted interventions to effectively reduce that disease burden, one community at a time.
Tuberculosis (TB) is the world's leading single-agent infectious cause of mortality. Current efforts to control TB are stagnating because we do not understand which interventions are likely to have the biggest impact on TB prevalence at the community level. We propose to address this gap by finding every detectable TB case in a small, well-circumscribed urban Ugandan district, using whole-genome sequencing to identify which cases ? even those without symptoms ? are responsible for transmission in the community, and characterizing each case in depth to understand which types of targeted interventions would likely help the most people.
