Data scientist develops novel health ROI to optimize biomedical resource allocations

"We believe that greater returns on investment can be achieved for the entire society if we could more effectively distribute biomedical research resources, including both funding and human capital," said Yao, an assistant professor in the Department of Software and Information Systems in the College of Computing and Informatics. She worked with collaborators at Columbia University, GlaxoSmithKline and the University of Chicago to develop the health ROI.

Using the health research opportunity index (ROI), Yao and colleagues compared about 1,400 diseases during a 12-year period. Their study was based on the numbers of clinical trials and scientific literature and treatment costs computed from billions of claim records. The investigators uncovered a substantial imbalance between U.S. health needs and research investment. For instance, conditions such as breast and cervical cancers were over-studied while Hashimoto's thyroiditis (a condition in which one's immune system attacks the thyroid) and secondary bone cancer were under-studied.

The biomedical research ecosystem is really complex due to the long development cycle and the large number of stakeholders, including public and private funding agencies, the pharmaceutical industry, patient advocacy groups and the general scientific and medical communities.

"If a disease is not perceived as scientifically interesting or if there seems little chance of finding a viable or marketable therapy, scientists and pharmaceutical company will be unlikely to invest in it," said Yao, who completed a Ph.D. in biomedical informatics from Columbia University. "On the other hand, a biomedical breakthrough or extreme public attention from those affected or celebrities can tip funding toward one disease at the expense of others."

In the example of breast cancer, Yao's health ROI ranked it as the most over-studied condition due to the disproportionally large number of clinical trials being conducted. "Resource allocation in biomedical research can be impacted by politics, public media attention, legal environment, economic conditions and disease burden, et al. My team is still working on improving the mathematical model by incorporating more of these factors in a quantitative way. We expect to use the health ROI to monitor resource allocation for biomedical research much like analysts monitor the stock markets using the Dow Jones or NASDAQ indices," Yao stated.

Mirsad Hadzikadic, director of the Complex Systems Institute at UNC Charlotte, added, "In this era of Big Data, we have more information available that enables us to examine questions in a more systemic manner. Dr. Yao's research exemplifies the power of Big Data analysis in minimizing the gap between health needs and biomedical research resource allocation. Her methods provide decision-makers the tools to make unbiased investment decisions. In the end, we as a society can benefit from more efficient allocation of limited research resources."