RESUMEN
<b>Objective</b>:Many epidemiological studies revealed that the varicella and herpes zoster have inverse correlation. This paper examines to confirm these associations by using Prescription Surveillance. <b>Design</b>: Multiple regression analysis about association between the incidence rate of herpes zoster and varicella adjusting for prefecture, fiscal year and week. <b>Methods</b>: We used the general public data from Prescription Surveillance. Prescription Surveillance estimates the number of patients who were prescribed anti-varicella zoster virus drug with age class. We assumed the patients under 15 years old were varicella patients, and over 65 years old were herpes zoster patients. Each the number of patient was aggregated by week and prefecture from April 5, 2010 to March 31, 2013. We regressed the incidence rate of herpes zoster on the incidence rate of varicella adjusting for prefecture, fiscal year and week. Moreover, we regressed it on the several lag of the incidence rate of varicella. <b>Results</b>: The incidence rate of varicella was significantly positive, however the estimated coefficient was 0.030. The lag of incidence rate of varicella were also significantly positive, however these estimated coefficient were less than 0.030. <b>Conclusion</b>: We found that the incidence rate of herpes zoster and varicella have positive correlation, however its magnitude was not so important. It needs attention that our finding means positive correlation when we remove seasonality and regionality of both incidences.