Data-driven optimization to improve mobile health care in disadvantaged communities
Led by health economist Rigoberto Delgado of the University of Texas Health Science Center at Houston, a team of researchers wants to help Texas Children's Hospital and other providers target their limited resources where they can do the most good. The team is mining public domain data in new ways, using geospatial mapping science and predictive analytics to forecast areas of highest risk for outbreaks. The researchers want to figure out not only where to send the mobile health clinics but, ultimately, how to prevent illness outbreaks in the first place and reduce the number of emergency room visits. Delgado's colleague on the project, optimization expert Jiming Peng of the University of Houston, is assisting with the modeling, algorithm design and analysis. "We're designing a way to predict the future demand for mobile clinic services in various communities, based on new modeling that uses many data sets from different sources," explains Peng. "And, a challenging issue is that all the optimization models developed in the project have binary variables with uncertain parameters and constraints due to the numerical errors in the process of data analysis and the dynamic nature of the target population." The researchers have already identified communities with high demand for some specific mobile clinic services in Houston and made recommendations to their hospital partners. Ultimately, they hope to reduce health care costs and disparities by identifying policies and strategies to encourage comprehensive and collaborative programs of mobile health units in Texas and across the United States.
Provided by National Science Foundation
More Science Nation videos