A new Australian-led study has found that artificial intelligence chatbots such as ChatGPT may be almost as effective as consulting a doctor for advice on lower back pain.
The research showed the potential of AI to address one of the world’s leading causes of disability which affects millions globally, often prompting sufferers to seek information online.
Associate Professor Bruno Tirotti Saragiotto, research co-author and Head of Physiotherapy at the University of Technology Sydney, said that as AI-powered chatbots like ChatGPT became increasingly common in offering health recommendations, understanding the accuracy of their recommendations was important.
“The findings show that AI chatbots can offer advice with accuracy levels comparable to those reported by healthcare professionals in Australia,” Associate Professor Saragiotto said.
AI chatbots excelled in answering questions related to suggested treatment and self-management, such as “What complementary therapies like massage or acupuncture could alleviate lower back pain?” while risk factors had the most inaccuracies.
The study also noted that AI chatbots accurately and consistently recommended exercise for preventing and managing low back pain.
However, chatbots provided inaccurate recommendations to other commonly asked questions, such as indicating that poor posture was responsible for 88% of lower back pain.
Another key observation was the ability of AI chatbots to recognise situations requiring medical referrals. In cases where professional care should be recommended, the AI systems advised users to consult a healthcare provider in 70-100% of instances.
Overall, out of 120 responses yielding 1069 recommendations, 55.8% were accurate, 42.1% inaccurate and 1.9% unclear.
“Our research indicates that AI chatbots have the potential to be a valuable resource for those seeking initial guidance on managing low back pain,” said Dr Giovanni Ferreira, co-author and Research Fellow at the University of Sydney Institute for Musculoskeletal Health.
“It’s important to note that these tools should complement, not replace, professional medical advice. While the accuracy of the AI-generated advice was impressive, we must consider the accessibility of this information.”
Notably, the responses were often complex, with a readability level suitable for individuals with a 10th to 12th grade or university-level education.
“Ensuring that health guidance is understandable to a broad audience remains an important challenge in the development of AI health tools,” Associate Professor Saragiotto said.
“As AI technology continues to evolve, further research will be necessary to refine these tools and ensure they can provide accurate, accessible, and safe health information to the public.”
The study was recently published in the journal Annals of the Rheumatic Diseases.