Principal Investigator: Richard G. White

The Modeling Project was funded in September 2007 as an adjunct to CREATE, with the goal of reducing tuberculosis (TB) incidence in HIV prevalent areas by developing mathematical models to maximize the impact of CREATE study results.

The Modeling Core will provide long-term projections, compare the likely effects of alternative strategies to TB control in high prevalent areas, and investigate population attributes that may affect generalizability of CREATE study results. In addition to the strategies under trial in CREATE, the project is taking alternative approaches to estimating impact of integrated HIV services and TB control strategies such as DOTS plus joint HIV/TB care, so that policy-makers can be informed on their relative strengths and weaknesses.

The specific objectives of the Modeling project are to:

1. Develop models that fit baseline epidemiologic data on TB and HIV from the CREATE trials, ZAMSTAR and Thibela TB, and data from African populations with differing TB/HIV epidemic characteristics. Data from two trials with designs and aims complementary to ZAMSTAR and Thibela TB will also be included: the ongoing DETECTB trial in Zimbabwe and the community-wide preventive therapy trial from 1957 in Bethel, Alaska;

2. Investigate key model assumptions through additional fieldwork in Southern Africa, and;

3. Finalize and validate models in parallel with trial interventions, to enable timely modeling, projection, and in-depth understanding of the CREATE trial results, in turn increasing the impact of trial outcomes through better communication to policy-makers of the likely short- and long-term impact of adopting similar strategies in other populations.