Analysis of Pediatric Respiratory Disease Trends Using the 2016 KIDs' Inpatient Database

Abstract

Many infants and young children worldwide have been affected by chronic respiratory conditions. In this paper, the authors performed an exploratory and predictive analysis of the 2016 KID data set to examine respiratory disease trends among children. They applied the multiple linear regression and random forest regression methods to build a predictive model for the length of stay (LOS) for children with respiratory problems. The tree approach implemented using random forest is found to be a better approach for predicting the length of stay (LOS). In addition, they performed an exploratory analysis of significant fields from the data set. From the analysis, it is found that the winter season has the highest number of inpatient admissions of children having chronic respiratory illnesses. Further, it is found that newborns and infants are more prone to respiratory diseases, with bronchitis being the leading cause of respiratory diseases among children.