@article{Wang2026, 
author = {Ying Wang and Xi-you Wang and Ya-nan Sun and Chang-he Yu},
title = {Baseline predictors of treatment response to Chinese tuina and manual physical therapy in knee osteoarthritis: a secondary analysis of a randomized controlled trial},
year = {2026},
journal = {Evidence-Based Chinese Medicine and Technology Assessment},
volume = {2},
number = {2},
pages = {9570038},
keywords = {knee osteoarthritis, manual therapy, predictors, pain, mental health, logistic models},
url = {https://www.sciopen.com/article/10.26599/eCMTA.2026.9570038},
doi = {10.26599/eCMTA.2026.9570038},
abstract = {IntroductionAlthough manual therapy, including traditional Chinese tuina and manual physical therapy, is generally effective for knee osteoarthritis (KOA), individual responses vary considerably. This secondary analysis aimed to identify the quantifiable baseline characteristics that predict treatment response, thereby supporting personalized care. MethodsData from a randomized controlled trial comparing tuina and manual physical therapy for KOA were analyzed. At week 4, 59 of 127 patients (46.5%) were responders by Outcome Measures in Rheumatology–Osteoarthritis Research Society International (OMERACT–OARSI) criteria. Candidate variables were screened by univariate analysis (P < 0.25). Multivariate logistic regression identified independent predictors. Internal validation used 1000 bootstrap resamples, with sensitivity analyses including least absolute shrinkage and selection operator (LASSO) and stepwise regression based on Akaike information criterion or Bayesian information criterion. ResultsUnivariate analysis identified disease duration, Kellgren-Lawrence grade, prior treatment satisfaction, all Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) subscales, numeric rating scale (NRS) pain score, 12-item short form health survey (SF-12) mental health score, functional tests, and treatment expectations as associated with response (P < 0.25). Multivariate logistic regression revealed two independent predictors: higher baseline WOMAC pain score (odds ratio [OR] = 1.56, 95% confidence interval [CI]: 1.31, 1.87, P < 0.001) and higher baseline SF-12 mental health score (OR = 1.06, 95% CI: 1.01, 1.12, P = 0.023). The model showed good discrimination (apparent area under the curve [AUC] = 0.838). Internal validation yielded an optimism-corrected AUC of 0.763 and Brier score of 0.169. Calibration and decision curve analyses indicated acceptable fit and net benefit. ConclusionPatients with more severe baseline pain and better mental health are more likely to respond favorably to either tuina or manual physical therapy. These findings bridge pain and mental health as predictors, providing evidence for precision manual therapy in KOA; external validation in larger cohorts is required.Registration: This study was registered with ClinicalTrials.gov (https://clinicaltrials.gov/, NCT03966248).}
}