Predictive Factors for Postherpetic Neuralgia Using Ordered Logistic Regression Analysis

OBJECTIVES: To identify predictive factors for the occurrence of postherpetic neuralgia (PHN). METHODS: The participants were 73 herpes zoster patients who had been treated at the pain clinic of our hospital between January 2008 and June 2010. Variables present at the initial visit were extracted from the clinical records for regression analysis of factors related to the occurrence of PHN. The following scores for response were used: 0=no PHN after 3 months; 1=PHN present after 3 months but absent after 6 months; and 2=PHN present after 6 months. Multivariate ordered logistic regression analysis was performed to identify the predictive factors for PHN. RESULTS: Advanced age odds ratio (OR)=2.740, confidence interval (CI)=1.110-6.761; P=0.0288 and deep pain (OR=4.244, CI=1.114-16.163; P=0.0341) at the initial visit to our outpatient pain clinic were found to be significant predictive factors for the occurrence of PHN. Diabetes mellitus (OR=3.075) and pain reduced by bathing (OR=3.389) also had high OR, although they were not significant. DISCUSSION: Our study indicates that advanced age and deep pain at the initial visit are significant predictors for PHN. Our results are considered likely to contribute to the establishment of evidence-based medicine in the optimal treatment of PHN.