Development and validation of a nomogram to predict the risk of post-stroke complex regional pain syndrome

TL;DR

This study creates a tool to predict which stroke patients are at risk of developing complex regional pain syndrome, helping doctors identify high-risk patients early.

Contribution

A novel nomogram model was developed and validated for early prediction of post-stroke complex regional pain syndrome.

Findings

The nomogram model achieved an area under the curve (AUC) of 0.858 for predicting post-stroke CRPS.

Key predictors included gender, age, NIHSS score, cervical spondylosis, sleep disorders, fasting blood glucose, and albumin.

Decision curve analysis and clinical impact curves confirmed the model's clinical utility.

Abstract

This study aims to assess risk factors and build a nomogram model to facilitate the early recognition of post-stroke complex regional pain syndrome (CRPS).

A total of 587 stroke patients admitted to Dongguan Hospital of Guangzhou University of Traditional Chinese Medicine from September 2021 to October 2024 were initially included in this study. After exclusions, 376 patients were selected. Among these, there were 90 patients with post-stroke CRPS, while the non-stroke CRPS group consisted of 286 patients. Feature selection and optimization to generate the predictive model and nomogram were performed using LASSO regression and multivariable logistic regression analysis. We also utilized calibration plots, receiver operating characteristic (ROC) curves, decision curves (DCA), and clinical impact curves (CIC) for model validation.

LASSO regression analysis and multivariate logistic regression identified gender, age, NIHSS score, cervical spondylosis, sleep disorders, fasting blood glucose (FBG), and albumin (ALB) as significant predictors. The nomogram model showcased reliable predictive effectiveness, achieving an area under the curve (AUC) of 0.858 (95% CI, 0.801–0.915). Both DCA and CIC demonstrated that the nomogram model holds substantial clinical utility.

This study has developed a novel predictive model for post-stroke CRPS, providing a valuable tool to facilitate the early detection of high-risk patients in a clinical environment.

Figures