Rapid Distinguishment of Drug Consumption Patterns by the Seasonal Decomposition Method / 中国药房

ABSTRACT

OBJECTIVE:To establish a simple and feasible method for rapid analysis of drug consumption patterns,and to im-prove the level of hospital pharmacy management. METHODS:Taking Motilium tablet as an example,the actual order amount da-ta of 24 months were analyzed by time series seasonal decomposition method;several key components of change trend were extract-ed by SPSS software,i.g. smoothed trend-cycle(STC),seasonal factors(SAF)and seasonally adjusted series(SAS). The varia-tion of components data series was observed to obtain the conclusion,and then compared with direct-vision method. RESULTS:The seasonal decomposition method had high prediction accuracy,the mean absolute percent error (MARE)(0.17) was less than that of error value by direct-vision method (0.24) significantly. CONCLUSIONS:This method can found drug consumption pat-terns quickly and accurately with the help of software. Based on this aspect,stock adjustment strategy can combine daily manage-ment and long-term dynamic management conveniently so as to improve the level of drug management.