Prediction tool development: creation and adoption of robust predictive model metrics at the bedside for greatly benefiting the patient, like preterm infants at risk of bronchopulmonary dysplasia, using Shiny-R

The purpose of this chapter is to discuss prediction tool development. Steps of a general data pipeline are discussed, including data collection, data tidying, exploratory data analysis, modeling, and model validation. We also endorse the incorporation of the R Shiny web app and connect to SQL databases to enhance data workflows and efficiency. The need for clinically relevant prediction models is highlighted with an emphasis on the importance of transparency by following the standardized reporting methods of the TRIPOD guidelines.