Prediction of Cardiac ICU Length of Stay after Infant Cardiac Surgery Using Exome Sequencing Data
Profound inflammatory response is a common complication after cardiac surgery with cardiopulmonary bypass. It has been shown that post-operative expression of pro-inflammatory cytokines is correlated with cardiac intensive care unit (CICU) length of stay (LOS). However, the underlying genetic mediators have not yet been fully characterized. In this study, we sequenced the exome of 54 subjects and identified genomic variants using The Genome Analysis Toolkit (GATK). We further built models for predicting CICU LOS using genomic variant information. Prediction performance was measured by area under a curve (AUC). We achieved AUC of 55?65% and 60?70% using all genomic variants and a subset of variants on inflammatory response-related genes, respectively.