Pilot Study: ChatGPT-enabled Task Sharing and Team Building Improve Workflow in Lymphedema Care.

BACKGROUND

Lymphedema care requires multidisciplinary collaboration, substantial documentation, and interfacility communication. Large language models, including ChatGPT, may streamline these processes and support team-based care; however, concerns remain regarding accuracy, privacy, and governance.

METHODS

We conducted a cross-sectional, anonymous survey at a lymphedema specialty clinic in Tokyo 24 months after ChatGPT implementation (May 2023), following STROBE guidelines. All staff (N = 12; physicians, nurses, therapists, and administrative staff; mean age 39.6 y) participated. Responses were collected using 5-point Likert scales and multiple-choice formats, and descriptive statistics were summarized.

RESULTS

All staff reported using ChatGPT at least once, with frequent use reported by 58.3%. Efficiency improved in 83.3% (Likert 4-5), and preparation time for information-sharing materials decreased in all respondents (≤50% in 33.3%, ≤30% in 41.7%). Primary uses included drafting documents/emails (~92%) and referral reports (~83%). ChatGPT improved external communication (mean Likert score 4.42/5), reduced perceived work-related stress, and increased confidence (≥4 in 66.7%). Staff reported moderate trust, occasional hallucinations, and the need for training and guidelines. Privacy concerns were noted by more than half of respondents.

CONCLUSIONS

ChatGPT enhanced efficiency, communication, and teamwork in a lymphedema clinic while reducing stress. Safe use requires human oversight, verification processes, and privacy safeguards. Under appropriate governance, large language models may facilitate task sharing and support clinical education, including prelearning for lymphatic ultrasound. We believe this version retains the essential points while meeting the journal's requirements.