本研究個案醫院的中藥局，其藥品品項多達700種，而在調劑過程中花費最多時間的步驟，是從藥櫃中尋找處方箋中的藥品。為了讓藥劑師有規則可循，須要有邏輯性的安排藥局中中藥藥罐的擺放位置。本研究試圖改善目前個案醫院中藥局調劑給藥流程，能夠針對藥劑師在「尋找藥罐調配藥品」這個最需要耗時費神的重要步驟加以改善。因此，結合資料探勘技術中的關聯規則法來對於大型中醫院數以萬計的處方箋進行探勘，找出處方箋中各種藥品與藥品之間的相關性，將常常搭配在同一處方箋的藥品罐集中擺放，這樣或許可以大量減少藥劑師尋找藥罐的時間，並進而減少因為疲勞或體力不繼發生不當調劑與作業疏失的機率。研究中進行三個主要部份的實驗為：依照「全年度」發掘關聯規則的結果與分析、依照「季節」發掘關聯規則的結果與分析、依照「醫師」發掘關聯規則的結果與分析。 最後本研究提供了一套流程與方法，讓中醫與藥劑師可以利用資料探勘技術來輔助處理中藥調劑。也試圖用更廣泛的面向與角度去探勘個案醫院的處方箋資料，再藉由專科中醫師來驗證分析探勘出來的處方箋藥品間的關聯規則，給予個案醫院建議來改善現況的不足。 ;The case study is a traditional Chinese medicine hospital phamarcy with items up to 700 species, and step pharmacist spend the most time in the drug administrating process is to find the prescription drugs from the medicine cabinet. In order to have pharmacists follow the rules, it needs to have a logical arrangements for gallipot in the pharmacy. This study attempts to improve the current medicine administration process in the case hospital. It may be better to the step of “Looking for formulations gallipot” which is the most important and time-consuming step for pharmacists. For a large Hospital, therefore, making use of data mining techniques with the association rule method for thousands of prescriptions identifies correlations of various prescription drugs. Cetralizing the frequent-prescriped gallipot in the neighborhood may be able to significantly reduce the time for pharmacist finding and getting gallipots. Furthermore, this could also reduce the opportunity of carelessness and improper actions due to pharmacist’s fatigue. There are three studies conducted, which are based on extracting the association rules for “full year, “season”, and “idividual doctors” Finally, this study provides a process and method and those let pharmacists are able to make use of data mining techniques supporting the traditional Chinese medicine administrating process. Also tries to use a broader perspective to mine prescription in the case hospital, and then validate the association rules between prescription drugs by the doctors of traditional chinese medicine.
