Predicting persistent pain after total knee arthroplasty using different machine learning algorithms
DOI:
https://doi.org/10.2340/17453674.2026.45965Keywords:
Arthroplasty, Artificial intelligence, Knee, Machine learning, Oxford Knee ScoreAbstract
Background and purpose: After total knee arthroplasty (TKA), 10–20% of patients remain unsatisfied. Well-performing clinical prediction models can provide individualized risk estimates and stratification in terms of poor outcomes, resulting in unnecessary surgeries being avoided and patients being counseled preoperatively. We aimed to create a precise, well-performing prediction model for clinical application using different machine learning algorithms to predict those patients who will have residual pain, a low total Oxford Knee Score (OKS) and the patient group who do not achieve minimally clinical important difference (MCID) in OKS 1 year after TKA.
Methods: We conducted a retrospective cohort study based on patients who had undergone primary TKA at our institution combined with 751 patient-related variables. The multivariable models used were based on the results of univariate analysis. We used the machine learning method Extreme Gradient Boosting (XGBoost). The discrimination capability of the models was measured with the area under the curve (AUC).
Results: 11,755 patients were included in this study. There were 850 (7.2%) patients who experienced persistent pain 1 year after TKA. The AUC was 0.67. For the secondary outcomes, the AUC values were similar. The most important variables in the model were lower preoperative OKS, younger age, valgus malalignment, lower preoperative pain OKS, use of mild opioid, neuropathic pain medicine and thyroxine, and higher body mass index.
Conclusion: The prediction models achieved poor AUCs. It seems clear that the prediction of pain and functional outcome after TKA is difficult, even with a large patient cohort combined with 751 patient-related variables and sophisticated machine-learning algorithms.
Downloads
References
Kahlenberg C A, Nwachukwu B U, McLawhorn A S, Cross M B, Cornell C N, Padgett D E. Patient satisfaction after total knee replacement: a systematic review. HSS J 2018; 14: 192-201. doi: 10.1007/s11420-018-9614-8. DOI: https://doi.org/10.1007/s11420-018-9614-8
Laigaard J, Aljuboori S M, Nikolajsen L, Mathiesen O, Lunn T H, Lindberg-Larsen M, et al. Chronic pain after primary total and medial unicompartmental knee arthroplasty for osteoarthritis: a Danish nationwide cross-sectional survey. Acta Orthop 2025; 96: 814-21. doi: 10.2340/17453674.2025.44898. DOI: https://doi.org/10.2340/17453674.2025.44898
Revaz M, Perneger T, Barea C, Miozzari HH, Hannouche D, Lübbeke A. What aspects of pain and functional outcomes contribute to patient satisfaction 1 year after surgery within total hip and total knee arthroplasty populations? A registry-based cohort study. Acta Orthop 2026; 97: 265-71. doi: 10.2340/17453674.2026.45732. DOI: https://doi.org/10.2340/17453674.2026.45732
Gunaratne R, Pratt D N, Banda J, Fick D P, Khan R J K, Robertson B W. Patient dissatisfaction following total knee arthroplasty: a systematic review of the literature. J Arthroplasty 2017; 32: 3854-60. doi: 10.1016/J.ARTH.2017.07.021. DOI: https://doi.org/10.1016/j.arth.2017.07.021
Harris A H S, Kuo A C, Bowe T R, Manfredi L, Lalani N F, Giori N J. Can machine learning methods produce accurate and easy-to-use preoperative prediction models of one-year improvements in pain and functioning after knee arthroplasty? J Arthroplasty 2021; 36: 112-7.e6. doi: 10.1016/j.arth.2020.07.026. DOI: https://doi.org/10.1016/j.arth.2020.07.026
Munn J S, Lanting B A, MacDonald S J, Somerville L E, Marsh J D, Bryant D M, et al. Logistic regression and machine learning models cannot discriminate between satisfied and dissatisfied total knee arthroplasty patients. J Arthroplasty 2022; 37: 267-73. doi: 10.1016/J.ARTH.2021.10.017. DOI: https://doi.org/10.1016/j.arth.2021.10.017
Fontana M A, Lyman S, Sarker G K, Padgett D E, MacLean C H. Can machine learning algorithms predict which patients will achieve minimally clinically important differences from total joint arthroplasty? Clin Orthop Relat Res 2019; 477: 1267-79. doi: 10.1097/CORR.0000000000000687. DOI: https://doi.org/10.1097/CORR.0000000000000687
Batailler C, Lording T, De Massari D, Witvoet-Braam S, Bini S, Lustig S. Predictive models for clinical outcomes in total knee arthroplasty: a systematic analysis. Arthroplast Today 2021; 9: 1-15. doi: 10.1016/j.artd.2021.03.013. DOI: https://doi.org/10.1016/j.artd.2021.03.013
Katakam A, Karhade A V, Collins A, Shin D, Bragdon C, Chen A F, et al. Development of machine learning algorithms to predict achievement of minimal clinically important difference for the KOOS–PS following total knee arthroplasty. J Orthop Res 2022; 40: 808-15. doi: 10.1002/jor.25125. DOI: https://doi.org/10.1002/jor.25125
Klemt C, Uzosike A C, Esposito J G, Harvey M J, Yeo I, Subih M, et al. The utility of machine learning algorithms for the prediction of patient-reported outcome measures following primary hip and knee total joint arthroplasty. Arch Orthop Trauma Surg 2023; 143: 2235-45. doi: 10.1007/s00402-022-04526-x. DOI: https://doi.org/10.1007/s00402-022-04526-x
Kunze K N, Polce E M, Sadauskas A J, Levine B R. Development of machine learning algorithms to predict patient dissatisfaction after primary total knee arthroplasty. J Arthroplasty 2020; 35: 3117-22. doi: 10.1016/j.arth.2020.05.061. DOI: https://doi.org/10.1016/j.arth.2020.05.061
Farooq H, Deckard E R, Ziemba-Davis M, Madsen A, Meneghini R M. Predictors of patient satisfaction following primary total knee arthroplasty: results from a traditional statistical model and a machine learning algorithm. J Arthroplasty 2020; 35: 3123-30. doi: 10.1016/j.arth.2020.05.077. DOI: https://doi.org/10.1016/j.arth.2020.05.077
Dawson J, Fitzpatrick R, Murray D, Carr A. Questionnaire on the perceptions of patients about total knee replacement. J Bone Joint Surg Br 1998; 80: 63-9. doi: 10.1302/0301-620X.80B1.7859. DOI: https://doi.org/10.1302/0301-620X.80B1.0800063
Murray D W, Fitzpatrick R, Rogers K, Pandit H, Beard D J, Carr A J, et al. The use of the Oxford hip and knee scores. J Bone Joint Surg Br 2007; 89-B: 1010-4. doi: 10.1302/0301-620X.89B8.19424. DOI: https://doi.org/10.1302/0301-620X.89B8.19424
Ingelsrud L H, Terluin B, Gromov K, Price A, Beard D, Troelsen A. Which Oxford Knee Score level represents a satisfactory symptom state after undergoing a total knee replacement? Acta Orthop 2021; 92: 85-90. doi: 10.1080/17453674.2020.1832304. DOI: https://doi.org/10.1080/17453674.2020.1832304
Beard D J, Harris K, Dawson J, Doll H, Murray D W, Carr A J, et al. Meaningful changes for the Oxford hip and knee scores after joint replacement surgery. J Clin Epidemiol 2015; 68: 73-9. doi: 10.1016/j.jclinepi.2014.08.009. DOI: https://doi.org/10.1016/j.jclinepi.2014.08.009
Price A J, Kang S, Cook J A, Dakin H, Blom A, Arden N, et al. The use of patient-reported outcome measures to guide referral for hip and knee arthroplasty. Bone Joint J 2020; 102-B: 941-9. doi: 10.1302/0301-620X.102B7.BJJ-2019-0102.R2. DOI: https://doi.org/10.1302/0301-620X.102B7.BJJ-2019-0102.R2
Clement N D, Afzal I, Demetriou C, Deehan D J, Field R E, Kader D F. The preoperative Oxford Knee Score is an independent predictor of achieving a postoperative ceiling score after total knee arthroplasty. Bone Joint J 2020; 102-B: 1519-26. doi: 10.1302/0301-620X.102B11.BJJ-2019-1327.R1. DOI: https://doi.org/10.1302/0301-620X.102B11.BJJ-2019-1327.R1
The Worcestershire CCGs. Musculoskeletal Surgery and Therapeutic Interventions Commissioning Policy. 2018. Available from: https://www.worcsacute.nhs.uk/images/CCG_Musculoskeletal_surgical_interventions_-_V1_9_7_-_June_2018
Christensen R, Ranstam J, Overgaard S, Wagner P. Guidelines for a structured manuscript: statistical methods and reporting in biomedical research journals. Acta Orthop 2023; 94: 243-9. doi: 10.2340/17453674.2023.11656. DOI: https://doi.org/10.2340/17453674.2023.11656
Chen T, Guestrin C. XGBoost. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, New York: ACM; 2016, p. 785-94. doi: 10.1145/2939672.2939785. DOI: https://doi.org/10.1145/2939672.2939785
Liu P, Fu B, Yang S X, Deng L, Zhong X, Zheng H. Optimizing survival analysis of XGBoost for ties to predict disease progression of breast cancer. IEEE Trans Biomed Eng 2021; 68: 148-60. doi: 10.1109/TBME.2020.2993278. DOI: https://doi.org/10.1109/TBME.2020.2993278
Hoerl A E, Kennard R W. Ridge Regression: biased estimation for nonorthogonal problems. Technometrics 1970; 12: 55. doi: 10.2307/1267351. DOI: https://doi.org/10.1080/00401706.1970.10488634
Zou H, Hastie T. Regularization and variable selection via the Elastic Net. J R Stat Soc Series B Stat Methodol 2005; 67: 301-20. doi: 10.1111/j.1467-9868.2005.00503.x. DOI: https://doi.org/10.1111/j.1467-9868.2005.00503.x
Hinterwimmer F, Lazic I, Suren C, Hirschmann M T, Pohlig F, Rueckert D, et al. Machine learning in knee arthroplasty: specific data are key-a systematic review. Knee Surg Sports Traumatol Arthrosc 2022; 30: 376-88. doi: 10.1007/s00167-021-06848-6. DOI: https://doi.org/10.1007/s00167-021-06848-6
Zhang S, Chen J Y, Pang H N, Lo N N, Yeo S J, Liow M H L. Development and internal validation of machine learning algorithms to predict patient satisfaction after total hip arthroplasty. Arthroplasty 2021; 3: 33. doi: 10.1186/s42836-021-00087-3. DOI: https://doi.org/10.1186/s42836-021-00087-3
Vega J, Emara A K, Orr M, Klika A K, Piuzzi N S, the Cleveland Clinic Arthroplasty Group. demographic and socioeconomic determinants are associated with poor preoperative patient-reported pain and function in primary TKA: a cohort study of 14,079 patients. J Bone Joint Surg Am 2023; 105: 286-92. doi: 10.2106/JBJS.22.00645. DOI: https://doi.org/10.2106/JBJS.22.00645
Hoelen T-C A, Heijnens L J M, Jelsma J, van Steenbergen L N, Schotanus M G M, Boonen B, et al. Socioeconomic status affects patient-reported outcome measures in total hip and knee arthroplasty: a retrospective Dutch registry study. J Arthroplasty 2024; 39: 2173-8.e2. doi: 10.1016/j.arth.2024.04.013. DOI: https://doi.org/10.1016/j.arth.2024.04.013
Additional Files
Published
How to Cite
License
Copyright (c) 2026 Anni Rajamäki, Aleksi Reito, Mari Karsikas, Mika Niemeläinen, Antti Eskelinen
This work is licensed under a Creative Commons Attribution 4.0 International License.
PlumX (by Elsevier) is an altmetrics platform that tracks and visualizes the online attention, usage, captures, citations, and social media engagement.
