Supplementary Material for: Targeted Searches of the Electronic Health Record and Genomics Identify an Etiology in Three Patients with Short Stature and High IGF-I Levels
datasetposted on 20.12.2019 by Cabrera-Salcedo C., Hawkes C.P., Tyzinski L., Andrew M., Labilloy G., Campos D., Feld A., Deodati A., Hwa V., Hirschhorn J.N., Grimberg A., Dauber A., the Genomics Research and Innovation Network
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
Introduction: Short stature is one of the most common reasons for referral to a pediatric endocrinologist and can result from many etiologies. However, many patients with short stature do not receive a definitive diagnosis. Objective: To ascertain whether integrating targeted bioinformatics searches of electronic health records (EHRs) combined with genomic studies could identify patients with previously undiagnosed rare genetic etiologies of short stature. We focused on a specific rare phenotypic subgroup: patients with short stature and elevated IGF-I levels. Methods: We performed a cross-sectional cohort study at three large academic pediatric healthcare networks. Eligible subjects included children with heights below –2 SD, IGF-I levels >90th percentile, and no known etiology for short stature. We performed a search of the EHRs to identify eligible patients. Patients were then recruited for phenotyping followed by exome sequencing and in vitro assays of IGF1R function. Results: A total of 234 patients were identified by the bioinformatics algorithm with 39 deemed eligible after manual review (17%). Of those, 9 were successfully recruited. A genetic etiology was identified in 3 of the 9 patients including 2 novel variants in IGF1R and a de novo variant in CHD2. In vitro studies supported the pathogenicity of the IGF1R variants. Conclusions: This study provides proof of principle that patients with rare phenotypic subgroups can be identified based on discrete data elements in the EHRs. Although limitations exist to fully automating this approach, these searches may help find patients with previously unidentified rare genetic disorders.