Objective To analyze the occurrence and influencing factors of residual back pain after percutaneous vertebroplasty (PVP) for osteoporotic vertebral compression fractures (OVCF) patients, and to establish a risk predictive nomograph model to verify its efficacy.

Methods A total of 302 patients with single segment OVCF who received bilateral PVP treatment were retrospectively selected as study objects. According to the pain Visual Analogue Scale (VAS) score one month after surgery, they were divided into residual back pain group (≥4 points, n=43) and non-pain group(< 4 points, n=259). The clinical data and changes of magnetic resonance imaging (MRI) parameters before and after treatment were compared between the two groups. Univariate and multivariate Logistic regression analysis was used to screen risk factors, the receiver operating characteristic (ROC) curve was plotted and the area under the curve (AUC) was calculated, the nomogram model was constructed with R software, and the calibration curve and decision curve were plotted.

Results Univariate and multivariate Logistic regression analysis showed that the intraspinal vacuum gap (IVC) (OR=2.680; 95%CI, 1.429 to 5.029; P < 0.001), posterior fascia edema (OR=2.863; 95%CI, 1.584 to 5.175; P < 0.001), paravertebral muscular degeneration grade Ⅱ (OR=3.762; 95%CI, 1.477 to 9.582; P=0.004), paravertebral muscular degeneration grade Ⅲ to Ⅳ (OR=5.801; 95%CI, 2.098 to 16.042; P < 0.001) and massive bone cement distribution (OR=1.578; 95%CI, 1.064 to 2.340; P=0.012) were independent risk factors for residual low back pain. A nomograph model based on the regression results was established, with the highest total score of 200 points. ROC curve showed that the area under the curve (AUC) of residual low back pain predicted by the model was 0.845, indicating a good discrimination. The calibration curve and decision curve showed that the model had a good fit and net benefit ratio.

Conclusion There is still a high incidence of residual back pain in OVCF patients after PVP. IVC, posterior fascia edema, severe paraspinal muscle degeneration and blocky cement distribution are important predictors of residual back pain. The nomogram risk prediction model has a good potential to guide clinical identification of high-risk patients with residual back pain.