Type 1 diabetes (T1D) is an immune-associated or immune-mediated chronic disease characterized by the progressive failure or targeted destruction of insulin-producing beta (β) cells in the pancreas. Management of the disease is challenging, involving lifelong exogeneous insulin replacement and 24/7 blood glucose monitoring. Although more than 1.6 million Americans are living with T1D, most diabetes research is currently focused on type 2 diabetes (T2D), which accounts for approximately 90% of diabetes cases. The expanding availability, granularity, and size of real-world health data, however, is opening unprecedented opportunities to use health informatics to advance T1D research that is computationally innovative and responsive to real-world needs. Conducting research that emerges from these opportunities necessarily occurs at the intersection of three interfacing domains: health informatics, clinical medicine, and biomedical research. Our work uniquely integrates these complementary disciplines to address distinct, yet overlapping, research foci. Our inaugural project involved implementation and refinement of a novel algorithm for discovery of clinically significant phenotypic heterogeneity in T1D. We first harmonized multiple years of comorbidity data for 16,000 individuals from the T1D Exchange Clinic Registry. We then used contrast pattern mining to examine individual and combined comorbidities that occur more frequently in affected individuals who have a family history, versus those who do not have a family history, of T1D. In our next project, we pivoted to using electronic medical record (EMR) data for nearly 500,000 individuals receiving care from 87 U.S.-based health systems. We first translated, refined, and implemented a complete EMR-based solution for diabetes case identification (T1D/T2D/No Diabetes). We subsequently used robust data imputation algorithms and statistical modeling methods to evaluate correlates of risk for various health outcomes based on diabetes status and type. We are now working with Cerner Real-World DataTM for 98 million individuals who have received care at nearly 120 U.S.-based health systems to develop a machine learning-based computational phenotyping methodology for automated case identification that will power nationwide recruitment efforts for T1D-focused clinical trials. In keeping with our commitment to Open Science, we are providing the tools and knowledge base developed through this work to the broader T1D and health informatics research communities. The totality of these projects demonstrates that informatics provides practical, scalable solutions for answering complex research questions and harnessing large-scale computational power that significantly expands current T1D research capabilities.
Advisor: Dr. Chi-Ren Shyu (MUIDSI PhD Core Faculty)
Dr. Sounak Chakraborty (MUIDSI PhD Core Faculty)
Dr. Grant Scott (MUIDSI PhD Core Faculty)
Dr. Camila Manrique (MU School of Medicine)
Dr. Mark Clements (Children’s Mercy Hospital at Kansas City)
For Zoom information, please contact Robert Sanders (firstname.lastname@example.org).