A Multicenter Prospective Study to Develop a Prediction Model for Postherpetic Neuralgia Using Clinical and Laboratory Indicators

INTRODUCTION: Postherpetic neuralgia (PHN) is the most common and difficult-to-treat complication of herpes zoster (HZ). This complication not only causes significant suffering but also imposes a substantial economic burden on patients. Early identification of patients at high risk and initiation of preventive interventions are crucial for reducing the burden of PHN. However, current clinical risk identification methods largely rely on subjective or semi-subjective clinical features and lack objective, easily accessible quantitative tools. This study aimed to develop and validate an efficient, objective, and clinically accessible prediction model for PHN risk by integrating routine clinical and laboratory indicators. METHODS: This prospective, multicenter study consecutively enrolled 722 patients with acute HZ who presented at four hospitals in Quanzhou between September 2, 2024, and July 12, 2025. Clinical data were collected, and multiple laboratory indicators were obtained through centralized testing. In the training set, univariate and multivariate logistic regression analyses were performed to identify independent predictors significantly associated with PHN. Subsequently, based on these factors, multiple machine learning algorithms were evaluated for performance to select robust models, with their generalizability subsequently assessed on an independent test set. RESULTS: The incidence of PHN was 18.84%. Multivariate analysis identified seven independent predictors of PHN: age, disease duration, prodromal pain, varicella-zoster virus immunoglobulin M (VZV-IgM), red cell distribution width-coefficient of variation (RDW-CV), urea, and serum sodium. The logistic regression model demonstrated the best predictive performance, achieving an area under the receiver operating characteristic curve (AUC) of 0.733 and accuracy of 0.765. CONCLUSION: This study provides the first report of PHN incidence among patients with HZ in Quanzhou. By integrating routine clinical and laboratory indicators, we successfully developed and validated a logistic regression model for predicting PHN risk. This model offers clinicians a practical tool for early identification of patients at high risk during the acute phase, thereby guiding timely preventive strategies.