Development and Internal Validation of a Therapeutic Effect Predictive Model for Myofascial Pain Syndrome

Xiumei ZhuWanquan Cheng^*

• Guannan County First People's Hospital, Lianyungang, China

Background: Myofascial Pain Syndrome (MPS) is a common chronic pain disorder, and there are significant individual differences in its clinical efficacy. Currently, there is a lack of reliable prediction tools to guide individualized treatment decisions. This study aimed to construct and validate a prediction model based on clinical and biomarker data to evaluate the responses of MPS patients to different treatment regimens and optimize treatment strategies. Methods: A total of 340 MPS patients was retrospectively enrolled and randomly split into a training set (n = 238, 70%) and an internal validation set (n = 102, 30%). Baseline data (including pain characteristics, trigger point distribution, psychological status, and inflammatory markers) were collected. The patients received standardized treatment (including dry needling, physical therapy, and drug intervention), and the efficacy was evaluated after 8 weeks (primary outcome: pain relief ≥50%). Predictive factors were screened through multivariate logistic regression, and machine learning models (random forest, support vector machine, and K-nearest neighbor algorithm) were further constructed developed. Internal validation was performed using the Bootstrap resampling method. The model performance was evaluated by the area under the receiver operating characteristic curve (AUC), calibration curve, and decision curve analysis (DCA). Results: The final model included 6 key predictive factors (including disease duration, baseline pain intensity, Patient Health Questionnaire-9 depression score, pain catastrophizing score, interleukin-6, and high-sensitivity C-reactive protein levels). The AUC value of the support vector machine model reached (95%CI:0.840-0.950) in the training set and remained at a relatively high level of 0.873(95%CI: 0.794-0.953) in the validation set, and the calibration was good (Hosmer-Lemeshow test, P > 0.05). DCA showed that the model had a high clinical net benefit within the threshold probability range of 0.10-0.70. Conclusion: A MPS efficacy prediction model, which had good internal predictive efficacy and interpretability, integrating clinical, psychological and inflammatory indicators was successfully constructed and internally validated. In the future, multi-center external validation and model optimization are needed to further improve its clinical applicability and promotion value.