Abstract
BACKGROUND: A discharge letter analyzer was developed in the Bielefeld Geriatrics Living Laboratory at Bielefeld University of Applied Sciences in cooperation with the Catholic Hospital Association East Westphalia (Katholische Hospitalvereinigung Ostwestfalen). This can support quality assurance in the area of discharge management. METHODS: The discharge letter analyzer uses text mining methods to find information in the respective discharge letters. Data are extracted from the free text as well as numerical values from the documented assessments. In the project the cases were assigned to diagnostic categories through both deductive and inductive category development and the collection of keywords by the medical staff involved. RESULTS: With the help of the analyzer patients could be assigned to different diagnostic categories. It was found that the majority of cases were treated due to multifactorial or fall-related indications. In addition, the patients improved in various tests over the course of treatment. Most patients could be discharged back home after treatment. CONCLUSION: The results show that discharge management requires a special focus on fall-related and neurodegenerative diseases in the subsequent treatment. Due to the predominant discharge of patients to their homes, there is a significant focus on outpatient care. In principle, a measurable improvement can be achieved through inpatient treatment.