The discharge importer tool (DIT) uses natural language processing (NLP) to read patient discharge summaries written by clinicians and automatically enter the data into RES-Q. Before finalising the data entry, the tool includes an interface for clinicians to verify what data the DIT will input and avoid entering incorrect information. 

 

Currently, entering data manually into registries takes large amounts of time and effort from medical staff that could otherwise be helping patients. Moreover, the manual data entry often results in incorrect or incomplete data entry due to clinicians’ heavy workload and time demands. This tool aims to alleviate the burden of data entry to both relieve clinicians and also ensure more complete data entry.

 

The DIT can read and understand medical data from seven languages (English, Spanish, Czech, Greek, Polish, Romanian, and Bulgarian). We trained this tool based on how real clinicians from multiple hospitals performed data entry. The DIT’s interface was designed based on insights from multiple clinicians across Europe and evaluated with 22 clinicians across the Czech republic.