Key result
ChatGPT-5 reliably distinguishes post-operative mortality risk between survivors and non-survivors with highly reproducible estimates.
Why the study?
Existing scoring systems for predicting post-operative mortality after femoral trochanteric fractures show only moderate predictive ability, prompting exploration of ChatGPT to synthesize diverse pre-operative multimodal data.
Can ChatGPT-based multimodal integration of pre-operative data generate clinically interpretable patterns for 2-month post-operative mortality risk in elderly pertrochanteric fracture patients?
Population
134 patients with pertrochanteric fractures
Comparison
GPT-4o vs GPT-5 multimodal assessment across two repeated evaluations
Design
Retrospective exploratory study
Follow-up
2 months
Authors
Loading...
May aid mortality risk estimation in elderly trauma; hypothesis-generating and requires prospective validation before adoption.
Observational (n=134)
Can ChatGPT-based multimodal integration of pre-operative data generate clinically interpretable patterns for 2-month post-operative mortality risk in elderly pertrochanteric fracture patients?
Mean Difference: 0.18 (95% CI -0.41–0.76)
ChatGPT-5 demonstrated potential in separating generated mortality risk values between survivors and non-survivors after integrating diverse pre-operative data, suggesting potential clinical relevance for AI-assisted multimodal assessment.
Noda et al. (2026) conducted an observational in Pertrochanteric fractures (n=134). ChatGPT (GPT-4o and GPT-5) multimodal pre-operative assessment was evaluated on Agreement between repeated estimates of 2-month post-operative mortality risk (MD 0.18%, 95% CI -0.41-0.76). ChatGPT-5 demonstrated separation of generated mortality risk values between survivors and non-survivors, with high reproducibility between repeated estimates (mean difference 0.18%; 95% CI -0.41-0.76).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: