Using text prediction for facilitating input and improving readability of clinical text.

Area of Science:

• Health Informatics
• Medical Documentation
• Natural Language Processing

Background:

• Healthcare documentation is time-consuming, reducing patient care time.
• Clinical text often contains errors like misspellings and abbreviations, impacting readability.
• Efficient text input methods are needed to streamline clinical workflows.

Purpose of the Study:

• To explore the application of text prediction for clinical text input.
• To address the unique challenges of text prediction within the healthcare domain.
• To evaluate the effectiveness of a text prediction prototype in a clinical context.

Main Methods:

• Developed a text prediction prototype.
• Utilized data from a medical journal and medical terminologies for prototype training.
• Evaluated the prototype on simulated authentic clinical text.

Main Results:

• The text prediction prototype achieved 26% keystroke savings.
• Evaluation on realistic clinical text demonstrated significant efficiency gains.
• Results indicate feasibility of text prediction in clinical settings.

Conclusions:

• Text prediction offers a viable solution to reduce documentation burden in healthcare.
• The developed prototype shows promising results for improving clinical documentation efficiency.
• Further development can enhance text prediction's role in patient care.