Related Experiment Video
Updated: Jul 5, 2025
05:46
Establishing a Diaphyseal Femur Fracture Model in Mice
Published on: December 9, 2022
2.1K
Systematic review of machine-learning models in orthopaedic trauma
Hidde Dijkstra1,2, Anouk van de Kuit1, Tom M de Groot1,3
1Department of Orthopaedic Surgery, University Medical Centre Groningen, Groningen, Netherlands.
Bone & Joint Open
|January 16, 2024
Summary
Machine-learning models in orthopaedic trauma show promise but often lack transparent reporting and have a high risk of bias. Adhering to guidelines like TRIPOD and PROBAST is crucial for clinical trust and application.
Area of Science:
- Orthopaedic Trauma Surgery
- Medical Artificial Intelligence
- Clinical Prognostics
Background:
- Machine-learning (ML) prediction models offer potential for personalized risk stratification in orthopaedic trauma care.
- A comprehensive overview and critical appraisal of existing ML models against established reporting guidelines are currently lacking.
- This study addresses the need for evaluating the quality and transparency of ML prediction models in this field.
Approach:
- A systematic literature search identified 45 ML-based prediction models in orthopaedic trauma up to January 2023.
- The Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement was used to assess reporting completeness.
- The Prediction model Risk Of Bias Assessment Tool (PROBAST) was employed to evaluate the risk of bias.
Key Points:
- The majority of models (60%) were developed for hip fracture patients, with mortality and length of stay being common outcomes.
- Median reporting completeness according to TRIPOD was 62%, indicating substantial room for improvement.
- A high risk of bias (69%) was identified, primarily due to small datasets, incomplete data analysis, and lack of performance reporting.
Conclusions:
- Many ML prediction models in orthopaedic trauma suffer from incomplete reporting and a high risk of bias.
- Implementing guidelines such as TRIPOD and PROBAST is essential to enhance transparency and reliability.
- Addressing these issues is critical to bridge the gap between ML model development and their effective clinical implementation.
