Prevalence and Nomogram-Based Prediction Model of Neuropathic Pain After Knee Arthroplasty

Neuropathic pain (NP) following knee arthroplasty significantly impairs patient quality of life, yet current research lacks objective predictors and predictive models. This study aimed to identify postoperative NP associated factors and develop a nomogram prediction model.

We enrolled 128 patients undergoing primary unilateral knee arthroplasty (February 2023-August 2024) with 6-24 month follow-up. NP was assessed using PainDETECT (PDQ) and DN4 scales; pain-related quality of life was evaluated via the Oxford Knee Score pain subscale (OKS-PS). Clinical records were systematically reviewed. Variables were screened using univariate logistic regression, optimized via LASSO regression, and incorporated into a multivariate logistic regression nomogram, followed by validation analyses.

NP incidence was 14.06% (18/128). OKS-PS confirmed a significant association between postoperative NP and reduced quality of life. The nomogram identified female sex, decreased preoperative range of motion (ROM), greater preoperative mechanical axis angles (pMAA) difference, and higher postoperative C-Reactive protein (CRP) as factors significantly associated with NP. Bootstrap validation yielded an AUC of 0.764 (95% CI: 0.760-0.770), sensitivity 0.707 (0.694-0.718), and specificity 0.737 (0.724-0.747). 10-fold cross-validation showed comparable results: AUC 0.813 (0.721-0.906), sensitivity 0.778 (0.5-0.938), specificity 0.764 (0.676-0.833). Calibration curves indicated agreement (MAE=0.037), and DCA demonstrated strong clinical utility across threshold probabilities of 0.003-0.391.

Postoperative NP was significantly associated with female sex, reduced preoperative ROM, increased pMAA difference, and elevated postoperative CRP. The validated nomogram showed good predictive performance, establishing a reliable tool for early identification and risk assessment of post-arthroplasty NP.