Journal article
A Novel Machine Learning Model to Predict Revision ACL Reconstruction Failure in the MARS Cohort
Orthopaedic journal of sports medicine, Vol.12(11), pp.1-14
11/2024
DOI: 10.1177/23259671241291920
PMCID: PMC11565622
PMID: 39555321
Abstract
Background: As machine learning becomes increasingly utilized in orthopaedic clinical research, the application of machine learning methodology to cohort data from the Multicenter ACL Revision Study (MARS) presents a valuable opportunity to translate data into patient-specific insights. Purpose: To apply novel machine learning methodology to MARS cohort data to determine a predictive model of revision anterior cruciate ligament reconstruction (rACLR) graft failure and features most predictive of failure. Study Design: Cohort study; Level of evidence, 3. Methods: The authors prospectively recruited patients undergoing rACLR from the MARS cohort and obtained preoperative radiographs, surgeon-reported intraoperative findings, and 2- and 6-year follow-up data on patient-reported outcomes, additional surgeries, and graft failure. Machine learning models including logistic regression (LR), XGBoost, gradient boosting (GB), random forest (RF), and a validated ensemble algorithm (AutoPrognosis) were built to predict graft failure by 6 years postoperatively. Validated performance metrics and feature importance measures were used to evaluate model performance. Results: The cohort included 960 patients who completed 6-year follow-up, with 5.7% (n = 55) experiencing graft failure. AutoPrognosis demonstrated the highest discriminative power (model area under the receiver operating characteristic curve: AutoPrognosis, 0.703; RF, 0.618; GB, 0.660; XGBoost, 0.680; LR, 0.592), with well-calibrated scores (model Brier score: AutoPrognosis, 0.053; RF, 0.054; GB, 0.057; XGBoost, 0.058; LR, 0.111). The most important features for AutoPrognosis model performance were prior compromised femoral and tibial tunnels (placement and size) and allograft graft type used in current rACLR. Conclusion: The present study demonstrated the ability of the novel AutoPrognosis machine learning model to best predict the risk of graft failure in patients undergoing rACLR at 6 years postoperatively with moderate predictive ability. Femoral and tibial tunnel size and position in prior ACLR and allograft use in current rACLR were all risk factors for rACLR failure in the context of the AutoPrognosis model. This study describes a unique model that can be externally validated with larger data sets and contribute toward the creation of a robust rACLR bedside risk calculator in future studies. Registration: NCT00625885 (ClinicalTrials.gov identifier).
Details
- Title: Subtitle
- A Novel Machine Learning Model to Predict Revision ACL Reconstruction Failure in the MARS Cohort
- Creators
- Brian R. Wolf - University of Iowa Hospitals and ClinicsKinjal Vasavada - Yale UniversityVrinda Vasavada - New York University Langone Orthopedic HospitalJay Moran - Yale UniversitySai Devana - University of California, Los AngelesChanghee Lee - University of California, Los AngelesSharon L. Hame - University of California, Los AngelesLaith M. Jazrawi - New York University Langone Orthopedic HospitalOrrin H. Sherman - New York University Langone Orthopedic HospitalLaura J. Huston - Vanderbilt UniversityAmanda K. Haas - Washington University in St. LouisChristina R. Allen - Yale UniversityDaniel E. Cooper - University of California, Los AngelesThomas M. DeBerardino - The University of Texas Health Science Center at San AntonioKurt P. Spindler - Cleveland ClinicMichael J. Stuart - Mayo ClinicAnnunziato (Ned) Amendola - Duke UniversityChristopher C. Annunziata - University of California, Los AngelesRobert A. Arciero - University of Connecticut Health CenterBernard R. Bach - Rush University Medical CenterChamp L. Baker - Hughston ClinicArthur R. Bartolozzi - University of Pennsylvania Health SystemKeith M. Baumgarten - Orthopedic InstituteJeffrey H. Berg - University of California, Los AngelesGeoffrey A. Bernas - University at Buffalo, State University of New YorkStephen F. Brockmeier - University of VirginiaRobert H. Brophy - Washington University in St. LouisCharles A. Bush-Joseph - Rush University Medical CenterJay Brad Butler VJames L. Carey - University of PennsylvaniaJames E. Carpenter - University of Michigan–Ann ArborBrian J. Cole - Rush University Medical CenterJonathan M. Cooper - HealthPartnersCharles L. Cox - Vanderbilt UniversityR. Alexander Creighton - University of North Carolina at Chapel HillTal S. David - University of California, Los AngelesWarren R. Dunn - Texas Orthopedic HospitalDavid C. Flanigan - The Ohio State UniversityRobert W. Frederick - Rothman InstituteTheodore J. Ganley - Children's Hospital of PhiladelphiaCharles J. Gatt - University Orthopaedic AssociatesSteven R. Gecha - University of California, Los AngelesJames Robert Giffin - Fowler Kennedy Sport Medicine ClinicJo A. Hannafin - Hospital for Special SurgeryNorman Lindsay Harris Jr - University of California, Los AngelesKeith S. Hechtman - University of California, Los AngelesElliott B. Hershman - Lenox Hill HospitalRudolf G. Hoellrich - Slocum Research and Education FoundationDavid C. Johnson - University of California, Los AngelesTimothy S. Johnson - University of California, Los AngelesMorgan H. Jones - Brigham and Women's HospitalChristopher C. Kaeding - The Ohio State UniversityGanesh V. Kamath - University of North Carolina at Chapel HillThomas E. Klootwyk - Methodist Sports MedicineBruce A. Levy - Orlando HealthC. Benjamin Ma - University of California, Los AngelesG. Peter Maiers - Methodist Sports MedicineRobert G. Marx - Hospital for Special SurgeryMatthew J. Matava - Washington University in St. LouisGregory M. Mathien - Tennessee Orthopaedic ClinicsDavid R. McAllister - University of California, Los AngelesEric C. McCarty - University of Colorado DenverRobert G. McCormack - Fraser HealthBruce S. Miller - University of Michigan–Ann ArborCarl W. Nissen - Connecticut Children's Medical CenterDaniel F. O’Neill - University of California, Los AngelesBrett D. Owens - Brown UniversityRichard D. Parker - Cleveland ClinicMark L. Purnell - University of California, Los AngelesArun J. Ramappa - Beth Israel Deaconess Medical CenterMichael A. Rauh - University at Buffalo, State University of New YorkArthur C. Rettig - University of California, Los AngelesJon K. Sekiya - University of Michigan–Ann ArborKevin G. Shea - Stanford UniversityJames R. Slauterbeck - Southeastern HealthMatthew V. Smith - Washington University in St. LouisJeffrey T. Spang - University of North Carolina at Chapel HillSteven J. Svoboda - University of California, Los AngelesTimothy N. Taft - University of North Carolina at Chapel HillJoachim J. Tenuta - Albany Medical Center HospitalEdwin M. Tingstad - University of California, Los AngelesArmando F. Vidal - Steadman ClinicDarius G. Viskontas - University of California, Los AngelesRichard A. White - Fitzgibbon HospitalJames S. Williams - Cleveland ClinicMichelle L. Wolcott - University of Colorado Anschutz Medical CampusRick W. Wright - Vanderbilt UniversityJames J. York - University of California, Los AngelesMARS Group
- Resource Type
- Journal article
- Publication Details
- Orthopaedic journal of sports medicine, Vol.12(11), pp.1-14
- Publisher
- SAGE PUBLICATIONS INC
- DOI
- 10.1177/23259671241291920
- PMID
- 39555321
- PMCID
- PMC11565622
- ISSN
- 2325-9671
- eISSN
- 2325-9671
- Grant note
- Allen F. Anderson, MD (Tennessee Orthopaedic Alliance)(Montana, USA)
The authors express their appreciation to the late Barton Mann, PhD (AOSSM, Rosemont, Illinois, USA), Allen F. Anderson, MD (Tennessee Orthopaedic Alliance, Nashville, Tennessee, USA), Jeffery R. Bechler, MD (University Orthopaedic Associates LLC, Princeton, New Jersey, USA), and Timothy M. Hosea, MD (University Orthopaedic Associates LLC, Princeton, New Jersey, USA), whose contributions to this work was of great significance. They sincerely appreciate Elizabeth Garofoli's years of dedicated work and effort on this study. The authors also extend their gratitude to John P. Albright, MD (University of Iowa Hospitals and Clinics, Iowa City, Iowa, USA), Jack T. Andrish, MD (Cleveland Clinic, Cleveland, Ohio, USA), John D. Campbell, MD (Bridger Orthopedic and Sports Medicine, Bozeman, Montana, USA), Diane L. Dahm, MD (Mayo Clinic, Rochester, Minnesota, USA), and Brett (Brick) A. Lantz, MD (Slocum Research and Education Foundation, Eugene, Oregon, USA) for their effort and leadership on this project. All are enjoying a well-deserved and happy retirement after many years of dedication to the advancement of orthopaedics.
- Language
- English
- Date published
- 11/2024
- Academic Unit
- Orthopedics and Rehabilitation; Physical Therapy and Rehabilitation Science
- Record Identifier
- 9984749495602771
Metrics
14 Record Views