The Role of ChatGPT in osteoporosis management: a comparative analysis with clinical expertise
- 9 April 2025
- journal article
- research article
- Published by Springer Nature in Archives of Osteoporosis
- Vol. 20 (1) , 1-7
Abstract
Summary This study evaluates the role of ChatGPT in osteoporosis management, demonstrating 91% diagnostic accuracy and significantly faster response times compared to clinicians. The findings highlight the potential for artificial intelligence (AI) to revolutionize clinical decision-making while emphasizing the critical need for professional oversight to ensure patient safety and comprehensive care. Objective Osteoporosis is a progressive skeletal disease that is characterized by increased bone fragility and an increased risk of fracture. Early diagnosis and effective treatment can significantly reduce healthcare costs; however, limited access to clinical expertise represents a significant challenge to patient care. This study evaluates the diagnostic and treatment recommendations provided by natural language processing (NLP)-based AI models for osteoporosis management and compares them with those of healthcare professionals. Methods A multicenter, cross-sectional study was conducted with the creation of 100 real scenarios from 206 patients with a diagnosis of osteoporosis. The data pertaining to bone mineral density (BMD) and the clinical parameters were subjected to analysis using ChatGPT-4.0. Thereafter, the recommendations proffered by this software were compared to those of five independent physiatrists. A statistical validation of the model’s accuracy was conducted through the use of categorical distribution analysis. Results ChatGPT exhibited a high degree of diagnostic accuracy, with 91% of responses being entirely accurate. It provided recommendations for both pharmacological and non-pharmacological interventions that were consistent with current clinical guidelines. Nevertheless, 8% of the responses were reported as incomplete. Furthermore, ChatGPT was able to produce diagnoses and treatment recommendations at a significantly faster rate than clinicians, while the mean answer time is 5.4 ± 2.45 min in clinicians and 2.3 ± 0.76 min in ChatGPT (p < 0.001). Conclusion These findings highlight the potential of AI tools like ChatGPT to improve efficiency in clinical decision-making while underscoring the necessity of collaboration with healthcare professionals to guarantee comprehensive patient care.Keywords
This publication has 16 references indexed in Scilit:
- Assessing the performance of ChatGPT in answering questions regarding cirrhosis and hepatocellular carcinomaClinical and Molecular Hepatology, 2023
- Quantitative CMR population imaging on 20,000 subjects of the UK Biobank imaging study: LV/RV quantification pipeline and its evaluationMedical Image Analysis, 2019
- High-performance medicine: the convergence of human and artificial intelligenceNature Medicine, 2019
- A guide to deep learning in healthcareNature Medicine, 2019
- Framing the challenges of artificial intelligence in medicineBMJ Quality & Safety, 2018
- Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus PhotographsJAMA, 2016
- Impact of the NOGG and NOF Guidelines on the Indication of Bone Mineral Density in Routine Clinical PracticeJournal of Clinical Densitometry, 2015
- The role of DXA bone density scans in the diagnosis and treatment of osteoporosisPostgraduate Medical Journal, 2007
- An estimate of the worldwide prevalence and disability associated with osteoporotic fracturesOsteoporosis International, 2006
- Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fracturesBMJ, 1996