Burden of Diseases and Comparative Risk Assessment

The Global Burden of Diseases, Injuries, and Risk Factors Study 2010 involves collaborators from Harvard University, Johns Hopkins University, the University of Queensland (UQ), and the World Health Organization (WHO). More than 800 condition-specific experts from around the world are involved in the study. As part of its leadership of the Global Burden of Disease Study 2010, IHME is contributing data and conducting analyses to produce new sets of disease, injury, and risk factor burden estimates for 1990, 2005, and 2010 for 21 regions across the globe. To do this, IHME researchers will use outputs from the Mortality, Causes of Death, and Functional Health Status work groups.

Burden of disease regions

Major activities in Burden of Diseases and Comparative Risk Assessment:

Conduct systematic reviews of epidemiological evidence for diseases, injuries, and risk factors:

The expert groups involved with the GBD Study 2010 are undertaking systematic reviews of epidemiological evidence from both published and unpublished sources, including datasets and epidemiological articles. These reviews are intended to find information by disease and injury on incidence, prevalence, case-fatality, and mortality. The experts are noting sample size, population of interest, regional scope, definitions of conditions, measurement approach, and other critical details for each source found. Together, they are investigating approximately 300 conditions. In parallel, select expert groups are reviewing risk factors that lead to particular health outcomes, such as obesity, smoking, and diabetes. Currently, they are investigating approximately 40 such risk factors and establishing a precise definition of risk factor exposure and the variable used to measure exposure in the population. They will complete a systematic review of all published and available unpublished epidemiological studies, health surveys, health examination surveys, and other data sources that can be used to estimate risk factor exposure.

The expert group leaders convened in 2009 to compare and discuss the limitations of the existing data and to learn more about the methods that will be centrally applied across the data to produce comprehensive estimates. They plan to convene again to review and determine the best methods and estimates for the burden of diseases, paying special attention to consistency and comparability across groups. Many groups are completing their epidemiological reviews and are working with the relevant GBD subteams to analyze the data. The results of this extensive study will be published in a series of peer-reviewed articles.

Produce estimates for all conditions using DisMod III:

Expert groups will send data to IHME, WHO, and UQ, who will jointly analyze the information. GBD researchers will use an analytic modeling tool developed by IHME called DisMod III. It produces consistent epidemiological parameters (e.g., incidence, case-fatality, remission, prevalence, duration) of a given disease using a generic model of disease dynamics in a population using any data available for that disease. For each condition, all available data will be entered into the software program, and results will be compared and critiqued iteratively by the central core team of researchers and the expert groups. This process will be exceptionally time-consuming but critical to ensure consistency across the final burden estimates. The final epidemiological parameters generated provide a critical input for the overall estimates.

Produce a new analytic engine:

A number of different quantitative components are inputs into final burden estimates. They include overall cause-of-death estimates by cause, overall mortality numbers by age, the epidemiological parameters mentioned above, and disability weights. An analytic engine must be created that seamlessly combines these different components to produce final estimates of disability-adjusted life years, a time-based measure that combines years of life lost due to premature mortality and years of life lost due to time lived in health states of less than ideal health.

Top of Page