Can structured EHR data support clinical coding? A data mining approach

Structured data formats are gaining momentum in electronic health record systems and can be leveraged for decision support and research. Nevertheless, such structured data formats have not been explored for clinical coding, which is an essential process requiring significant manual workload in health organizations. This article explores the extent to which fully structured clinical data can support the assign- ment of clinical codes to inpatient episodes, through the design and application of a methodology that tackles high dimensionality issues, addresses the multi-label nature of coding and optimizes model parameters. The methodology encompasses transforming database entries to define a feature set and build a data matrix representation, and testing combinations of filter feature selection methods with machine learning models to predict code assignment. The methodology is tested with a real hospital dataset, with results showing varying predictive power across codes but demonstrating the potential of leveraging structuring data to reduce workload and increase efficiency in clinical coding.