Objective:

To evaluate the diagnostic capabilities of two AI chatbots, ChatGPT 5.1 and DeepSeek V3.2, in providing preliminary diagnosis and triage for musculoskeletal disorders, specifically focusing on low back pain.

Approach:

• Study Design: A retrospective comparative study was conducted at a tertiary academic teaching hospital in Beijing, enrolling outpatients with low back pain. The study assessed the performance of the AI chatbots in triage and differential diagnosis through two phases: chief …
• Ethical Considerations: The study adhered to ethical principles outlined in the Declaration of Helsinki and was approved by the institutional ethics committee. It involved minimal risk to participants, and informed consent was waived.
• Population Selection: Patients presenting with low back pain as the primary symptom during their first visit to the orthopedic outpatient clinic were retrospectively analyzed.

Key Findings:

• Musculoskeletal disorders have a rising prevalence, with a 21.71% increase in the US from 2000 to 2021.
• AI chatbots like ChatGPT and DeepSeek have potential in assisting with clinical diagnosis and triage but require systematic evaluation in real-world contexts.

Interpretation:

The study highlights the need for effective preconsultation triage systems to improve resource allocation and patient outcomes in managing musculoskeletal disorders.

Limitations:

• The study is retrospective and relies on existing clinical records.
• The evaluation of AI chatbots in real-world clinical scenarios remains an emerging area of investigation.

Conclusion:

The study aims to assess the utility of AI chatbots in improving the diagnostic process for low back pain, addressing the complexity of musculoskeletal disorders.