Artificial intelligence and telerehabilitation in distal radius fracture care for older adults: a narrative review for resource-constrained health systems

Abstract

Distal radius fractures are among the most common fragility injuries in older adults and account for a substantial share of orthopaedic trauma workload, particularly in high-volume and resource-constrained settings. In elderly patients the choice between nonoperative management and surgical fixation is rarely straightforward: long-term functional outcomes tend to converge across treatments, whereas the optimal approach for an individual is shaped by fracture pattern, bone quality, frailty, cognition, social circumstances and rehabilitation access. This narrative review, written for resource-constrained health systems and using Türkiye as a worked example, synthesises current evidence on conservative versus surgical treatment of distal radius fractures in older adults and examines where artificial intelligence (AI) and machine learning (ML) might realistically contribute along the care pathway. Sources were identified through PubMed and Web of Science searches of English-language literature conducted in 2025, prioritising randomised trials, guidelines and systematic reviews for the clinical questions and validation studies and reviews for the AI applications. The evidence base is uneven: automated fracture detection is the most mature application, whereas AI-supported treatment selection, prediction of loss of reduction, prediction of patient-reported outcomes and rehabilitation triage remain largely investigational in this population. Implementation barriers are substantial, including limited dataset quality, incomplete external validation, difficult workflow integration and the digital exclusion of many older adults. AI is therefore best understood as a potential decision-support layer rather than a replacement for clinical judgement. If rigorously validated, its most realistic near-term contributions are standardising assessment, flagging patients who need closer follow-up and extending follow-up capacity in high-volume environments. By separating established from investigational applications on an explicit readiness gradient, distinguishing prediction of surgical need from prediction of surgical benefit, and separating telerehabilitation from AI-based rehabilitation, this review offers a realistic, implementation-focused framework for using AI in high-volume, resource-constrained fracture care and for prioritising the validation still required.