Real-time surveillance to support dynamic and predictive models of MDR/XDR TB

点击翻译按钮获取中文摘要

Mathematical models provide mechanisms to evaluate the potential efficacy of public health interventions. While traditional models of tuberculosis have relied on parameters obtained from historical observation, there is a pressing need to develop models that depend more directly on actual surveillance and contact-tracing data. Additionally, treatment outcomes of patients with drug-resistant tuberculosis are often compromised by the delay before detailed drug resistance profiles are available. In practical terms, this means that patients with drug-resistant disease are often treated with inappropriate or ineffective drugs for many months. We have previously developed an Electronic Medical Record (EMR) to facilitate clinical care of tuberculosis patients; here we propose to extend this platform to function as a research tool. The specific aims of this project are to: (1) develop a surveillance system for multi-drug resistant (MDR) and extensively drug resistant (XDR) tuberculosis based on the EMR established in Lima, Peru; (2) develop realistic models of tuberculosis transmission that capture important individual and community heterogeneities and use these models to assess the impact of specific control policies; and (3) use surveillance data to develop an individual prediction model for drug resistance profiles that will serve as a decision support tool for choosing appropriate tuberculosis drug regimens.