AI-Supported Documentation and Clinical Monitoring in Palliative Care: A Real-World Observational Study

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Palliative Care Medicine

Background:

• Palliative care patients often have prolonged hospital stays, necessitating comprehensive documentation and coordinated care.
• The integration of AI, specifically GPT-based language models, presents novel avenues for optimizing clinical workflows.

Purpose of the Study:

• To assess the practical application of a GPT-based AI tool in palliative care.
• To evaluate AI's utility in documentation, trend identification, and clinical decision support within a real-world setting.

Main Methods:

• A retrospective observational study involving 25 patients in a hospital-based palliative care unit.
• Utilized a GPT-based AI tool for clinical documentation, drug monitoring, and trend recognition.
• Physicians inputted data; AI generated drafts, which clinicians reviewed and finalized. Therapeutic suggestions were verified by specialists.

Main Results:

• AI significantly reduced discharge summary documentation time (20.4 min to 6.1 min).
• AI identified critical clinical trends in 8 patients, enabling timely re-evaluation.
• AI provided medication suggestions in 6 cases, confirmed by specialists. Physicians reported decreased cognitive load and enhanced record clarity.
• AI-generated summaries improved family communication.

Conclusions:

• GPT-based AI tools enhance documentation efficiency and clinical awareness in palliative care.
• AI serves as a valuable support tool for managing complex patients under clinical supervision, without replacing physician judgment.