A Study on Birth Prediction and BCG Vaccine Demand Prediction using ARIMA Analysis

Background/Objectives: This study was conducted to solve the problem by predicting vaccine demand in advance through analysis of progress of birth of newborn babies in our country. Methods/Statistical Analysis: The deducted problem was defined and information, data, and use analysis method and planning procedures for creating alternatives for this issue were conducted. Afterwards, R which is an open source analysis tool was used to analysis and for visualization. In this analysis, a time series model (ARIMA model, Box-Jenkins methodology) was used to predict demand and perform the research to predict the number of births in Republic of Korea. Findings: The tuberculosis vaccines in Korea are currently being entirely of imported ones. However, the import volume often lacks meeting the demand. In this paper, research was performed to predict the demand of tuberculosis vaccines to secure vaccine stock. As result of analysis, the number of births next year was predicted to be 445,558 (in 2016). Also, analyzed results showed that approximately 388,251 to 502,864 babies will be born in reliability level of 85% and that approximately 357,915 to 533,200 babies will be born in reliability level of 95%. Vaccine should be prepared standard to the minimum value within error range because vaccines have expiration dates. Also, if more births occur than the predicted result, the issue can be coped in prior plans of preparing BCG seal-type vaccines by comparing with monthly predicted number of births. Application/Improvements: The results of this study will be applied to the ways to politically solve problems such as supply and demand of BCG vaccine for the expected newborns.