Abstract
BACKGROUND: The 3PM framework revolutionizes disease management by facilitating early risk prediction, disease prevention, and personalized treatment. For isolated incomplete lateral meniscal injuries (IILMI), where early diagnosis is challenging due to non-specific symptoms, 3PM's proactive approach is beneficial in preventing knee joint disease progression and maintaining patients' quality of life. AIMS: This study aimed to develop a predictive model within the 3PM framework, integrating knee MRI anatomical features with individual physical activity (PA) patterns to enhance early IILMI detection and treatment efficacy, improving patient outcomes and quality of life. METHODS: The training dataset comprised 254 patients. Using logistic regression analyses and least absolute shrinkage and selection operator (LASSO), IILMI was identified among various preoperative factors containing knee MRI and PA features. A dynamic nomogram was constructed and subjected to internal and external validations (91 patients). Validation encompassed C-index, receiver operating characteristic (ROC) curves, calibration curves, decision curve analysis (DCA), and clinical impact curves. ROC analysis determined the risk stratification cut-off. RESULTS: Six independent IILMI factors were identified, including PA intensity, PA type, degree of PA intensity, and MRI-derived anatomical parameters. The dynamic nomogram showed high predictive accuracy (C-index, 0.829 in training, 0.906 in validation). IILMI patients were divided into low-risk, medium-risk, and high-risk groups according to the cut-off value. CONCLUSION: In 3PM-guided management, the dynamic nomogram enables early IILMI diagnosis in patients while promoting IILMI stratification making personalized treatment feasible. With further development, it holds promise for effectively predicting IILMI risk, preventing severe knee pathologies, and enhancing the quality of life for at-risk patients. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s13167-025-00399-3.