Related Experiment Video
Updated: Aug 28, 2025
09:51
The Transition to an Anterior-Based Muscle Sparing Approach Improves Early Postoperative Function but is Associated with a Learning Curve
Published on: September 7, 2022
2.9K
External Validation of a Machine Learning Algorithm for Predicting Clinically Meaningful Functional Improvement After
Kyle N Kunze1,2, Austin Kaidi1,2, Sophia Madjarova2
1Department of Orthopedic Surgery, Hospital for Special Surgery, New York, New York, USA.
The American Journal of Sports Medicine
|September 22, 2022
Summary
This study validated a machine learning algorithm for predicting hip arthroscopy outcomes. The algorithm demonstrated reliable performance in an independent cohort, supporting its use in clinical decision-making for femoroacetabular impingement syndrome.
Area of Science:
- Orthopedics
- Medical Informatics
- Machine Learning
Background:
- Machine learning (ML) enables individualized risk prediction, potentially improving clinical decision-making.
- A previously developed ML algorithm predicts outcomes after hip arthroscopy for femoroacetabular impingement syndrome (FAIS).
- External validation is crucial for assessing the generalizability of prognostic models but is infrequently performed.
Purpose of the Study:
- To evaluate the external validity of a machine learning algorithm designed to predict clinically meaningful improvement following hip arthroscopy.
Main Methods:
- A cohort study (Level of evidence, 3) utilized data from an independent hip preservation registry.
- 154 patients undergoing hip arthroscopy for FAIS between 2015-2017 were included.
- Model inputs included patient demographics, radiographic parameters, and preoperative scores to predict a minimal clinically important difference (MCID) in the Hip Outcome Score-Sports Subscale at 2 years postoperatively.
Main Results:
- The ML algorithm achieved good to excellent discrimination in the validation cohort (concordance statistic = 0.80).
- Calibration was comparable to the derivation cohort (slope = 1.16, intercept = 0.13).
- Decision curve analysis indicated a net treatment benefit when using the algorithm.
Conclusions:
- External validation confirmed the ML algorithm's reliable performance in predicting meaningful improvement after hip arthroscopy.
- The algorithm demonstrates superior discrimination and comparable calibration in an independent population.
- This validated algorithm is a key step towards clinical implementation for enhanced decision-making and resource allocation.
