Background: The increasing bureaucratic burden in everyday clinical practice impairs doctor-patient communication (DPC). Effective use of digital technologies, such as automated semantic speech recognition (ASR) with automated extraction of diagnostically relevant information can provide a solution.
Objective: The aim was to determine the extent to which ASR in conjunction with semantic information extraction for automated documentation of the doctor-patient dialogue (ADAPI) can be integrated into everyday clinical practice using the IVI routine as an example and whether patient care can be improved through process optimization.
Methods: In a prospective monocentric study at the Sulzbach Eye Clinic, 50 patients were included in the period from 2020 to 2022. As part of a project funded by the "Zentrale Initiative des Mittelstandes" (ZIM), a demonstrator was developed with the consortium partners and integrated into the hospital information system (HIS). For qualitative and quantitative evaluation, a survey of patients was carried out using a questionnaire before and after implementation of the ADAPI module, supplemented by a determination of acceptance and possible time saving for users.
Results: The ADAPI module was successfully integrated into the HIS. The documentation of the findings and connection of the subsequent processes could be automated. An improvement in the DPC was reported for 13/50 patients (26%). The majority 35/50 (70%) did not notice any subjective change. The average duration of the conversation in the area of conventional documentation was 4.46 min (1-10 min) and 3.9 min with ADAPI (1-10 min).
Conclusion: With respect to the quality of the conversation within the DPC, no relevant differences could be shown with both types of documentation; however, larger case numbers and other areas of application still need to be evaluated. In the long term, the use of automated documentation solutions will sustainably improve the efficiency, completeness and consistency of clinical documentation.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1007/s00347-024-02165-8 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!