Grant: CIHR Fellowship

One third of the world population harbors a latent Mycobacterium tuberculosis infection. Efforts in controlling the spread of tuberculosis both in developing and developed countries have stalled in recent years because of the ongoing activation and dissemination of infection in individuals with latent infection. It is impossible to screen the whole population for tuberculosis. As such, policy makers must focus on targeting high risk groups and providing them with preventive therapy. In collaboration with British Columbia Center for Disease Control (BCCDC), which hosts a longitudinal database of screened individuals, this study aims at developing a rigorous decision rule that will individualize tuberculosis screening for at risk subjects and recommend the screening/preventive treatment strategy that will maximize the benefit for each dollar spent on screening. This study will be carried out in the following steps. Data from BCCDC will be used to derive a predictive model that will calculate the probability of future tuberculosis reactivation based on subject's characteristics at the time of screening (age, gender, ethnicity, comorbid conditions, socioeconomic status etc). Next, results will be used to populate a computer model of tuberculosis in order to calculate the societal costs and health-related quality of life of hypothetical individuals if they are not offered preventive therapy, offered preventive therapy after positive screening, or offered preventive therapy without screening. Finally, the results of the previous step will be used to derive a decision rule that based on the characteristics of individuals will recommend the most cost-effective screening strategy. Such customized screening approach has the potential to reduce the costs of screening while improving its benefit, which, given the massive number of people affected, can result in substantial benefits to the society.