The BigMouth Dental Data Repository is a comprehensive database of electronic health records provided by dental schools in a partially de-identified manner. Eleven major dental schools are currently participating with Iowa joining in 2020. This repository can be used for high-powered research using information collected from each of the participating institutions, and is a promising avenue for big data research.
At present, there are over four million electronic health records for subjects available from the repository, and Iowa provides almost 800K of them. These records can be used to discover the relationships between various oral health conditions, and it can explore risk factors for those conditions—whether they are related to demographics, prior treatments, or other health conditions.
For example, in Joshi, et. al.’s Skeletal malocclusion: a developmental disorder with a life-long morbidity, the researchers surveyed the correlations between malocclusion and other medical conditions in the available literature, including conditions as diverse as sleep apnea, cleft lip and palate, and hypertension. Using the BigMouth Dental Data Repository, they identified just over 3,000 patients diagnosed with Class I, II, or III malocclusion, and were, then, able to determine whether correlations found in significantly smaller data sets in the available literature held with this much larger group.
The amount and quality of information available in the repository is an exciting opportunity that can provide extensive evidence in response to important clinically-oriented research questions.
Here at Iowa, Brian Howe, director of clinics and assistant professor in the Department of Family Dentistry, and Ahmed Mahrous, assistant professor in the Department of Prosthodontics, have begun working on a project that will use the repository. They are piloting the project with data from Iowa’s own database in Axium, but plan to expand the study to include the larger data set from BigMouth.
The BigMouth team at Iowa includes Brian Howe, Chuck McBrearty, Kyungsup Shin, and Sara Miller. To learn more, including how to access the data, please review this presentation (pdf).