JMIR Publications

ABSTRACT

Background:

The burden of type 1 diabetes (T1D) in children youth is growing. However, the current approach for identifying pediatric T1D is costly, because it requires substantial manual efforts. The purpose of this study was to develop a computable phenotype for accurately and efficiently identifying T1D in youth.

Objective:

The objectivel of the project was to use a set of criteria known to be associated with pediatric patients with Type 1 diabetes, and apply this knowledge in the form of a phenotype to query the electronic health record and identify members who met the criteria.

Methods:

This retrospective study utilized a dataset from the University of Florida Health Integrated Data Repository (IDR) to identify 300 patients <18 years of age with T1DM, T2DM, or were healthy based upon a developed computable phenotype. Three endocrinology residents/fellows manually reviewed medical records of all probable cases to validate diabetes status and type. This refined computable phenotype was then used to identify all cases of T1DM and T2DM in the OneFlorida Clinical Research Consortium.

Results:

A total of 295 electronic health records were manually reviewed. A total of 126 cases were found to have type 1 diabetes, 35 with type 2, and 130 with no diagnosis. After running analyses, the positive predictive value was 94.7 percent, the sensitivity was 96.9 percent, specificity was 95.8 percent, and the negative predictive value was 97.6 percent. Overall, the computable phenotype was found to be an accurate and sensitive method to pinpoint pediatric patients with type 1 diabetes.

Conclusions:

We developed a computable phenotype for identifying T1D correctly and efficiently. The computable phenotype that was developed will enable researchers to identify a population accurately and cost-effectively. As such, this will vastly improve the ease of identifying patients for future intervention studies.