In addition to making accurate health data available to all, IHME strives to provide researchers and decision-makers with efficient, easy-to-use computational tools that they can use to derive results from their own data. Our aim is to expand the global evidence base on health so that a variety of audiences can have access to information when making critical decisions. Validation studies on the methods incorporated within the tools will always be published in peer-reviewed journals. Our goal is that our tools will produce comprehensible information without the user having to know all the methodological complexities.
Researchers at IHME are in the process of developing several new computational tools to analyze and estimate health data. They include:
- A desktop application to apply the Random Forest Method to verbal autopsy data. This machine learning method proved to be the most successful of several methods rigorously developed and tested for performance as part of the Population Health Metrics Research Consortium (PHMRC) project. The desktop application would allow users to estimate population-level causes of death from their own verbal autopsy data collected using the newly developed PHMRC instrument.
- A more user-friendly version of DisMod-MR, a program designed to provide estimates of age-specific incidence and duration of specific diseases. It utilizes prevalence, incidence, remission, and relative mortality for a specific subpopulation localized in time, space, age, and sex and allows the user to modulate several different model assumptions to produce graphical results. A working version of the program is being used in the Global Burden of Diseases, Injuries, and Risk Factors Study 2010 and is being tested by multiple researchers. A more streamlined and easy-to-navigate version will be made available in the public domain after the study’s conclusion.