Améliorer le diagnostic de la tuberculose: évaluation, modélisation et obstacles


Pai Madhukar

Université McGill


Domaine : maladies infectieuses et immunitaires

Programme chercheurs-boursiers - Senior

Concours 2014-2015

There are 9 million new cases of tuberculosis (TB) and 1.5 million deaths due to TB each year. In Canada, TB is an important health problem among immigrants, HIV-infected persons, and Aboriginal populations. However, existing diagnostic tests simply aren¿t good enough. Therefore, my research is focused on the validation and modeling of point-of-care tests (rapid easy-to-use tests) for TB and includes four main projects.

In our first project, we are evaluating how a new automated test (Xpert) could reduce delays in diagnosing TB and how cost-effective it is, the first study of this kind in Canada. This study is being done in Montreal (QC) and Iqaluit (NU), where more than 650 have been recruited so far.

In our second project, we are creating a mathematical model to understand the impact and cost-effectiveness of introducing new TB tests on a large scale in India and Southeast Asia. We'll look at the impact of TB routine testing for adults diagnosed with HIV, and for routine HIV testing for adults who have active TB.

In North America, healthcare workers are usually tested for latent TB (TB not yet active) every year, and in recent years, new blood tests have been used widely. However, when these tests are repeated, they sometimes vary over time: someone who tested positive one year might test negative the next, and vice versa. With our third project, we are creating a model to understand how and why test results change and how to interpret them.

Using focus groups, surveys and interviews of hundreds of people, our fourth project will examine which point-of-care tests are used and which aren't, why, and what an ideal point-of-care test should be.