背景：病患用藥安全的議題一直受到大家的重視，用藥的疏失可能導致一連串的錯誤、或幸運的察覺、或未對病患造成不良影響，但也可能因此而導致病患嚴重的傷害與後續醫療費用的支出。本研究希望藉由問題處方的錯誤型態分析與探討，建立防禦疏失發生的機制進而預防用藥疏失，提升病患用藥安全。
方法：本研究為前瞻性研究，收集自民國97年1月1日至97年12月31日期間，門診、急診、住院之問題處方，經門診審核藥師及發藥藥師認定有疑義的處方進行統計分析。本研究共分為三階段，第一階段主要為問題處方的收集與形態分析，並分析醫師對於藥師提供之藥事服務之接受度。第二階段，根據問題處方型態尋求改善方案，例如：重複處方電腦設限、開方條件電腦設限、電腦警示系統、修改電腦建檔…等來降低處方錯誤率。第三階段，電腦系統介入後執行的成效評估。
此研究採用Microsoft Office Excel 2008進行資料建檔，t檢定（單尾檢定）統計方法分析。
結果：自民國97年1月1日至97年12月31日期間，共收集門診、急診、住院問題處方1327張，問題處方的錯誤形態，以「有過敏史仍開藥」251張佔18.91%（n=1327）為最高，其次依序為「總量不符」194張佔14.62%（n=1327）、「頻次不適當」134張佔10.10%（n=1327）、「劑量不適當」124張佔9.34%（n=1327）、「醫師輸入錯誤」118張8.89%（n=1327）。總計實施60件電腦警示、設限及改善措施。1327張問題處方中，除去其中「有過敏史仍開藥」經與醫師確認後確定可用29張、刪除過敏記錄55張，總計實際開方錯誤之問題處方共1243張。經由改善措施介入後，97年下半年整體平均開方錯誤率為萬分之6.11（546/894356）相較於97年上半年整體平均開方錯誤率為萬分之8.06（697/864576），97年下半年減少了近四分之ㄧ的錯誤率，具統計學上之差異（P＜0.05）。在藥師建議的接受度部份，1327張問題處方中，除去85張未註明用法之處方、經與病患確認用法後未做出建議，實際與醫師連繫做出建議者共計1242張處方，醫師接受建議的比率達82.93%（1030/1242）
結論：本研究發現某醫學中心常見之處方問題主要為「有過敏史仍開藥」、「總量不符」、「頻次不適當」、「劑量不適當」、「醫師輸入錯誤」，經由電腦系統修正、開方頻次設限、最大劑量電腦設限、開方條件電腦設限、電腦警示系統、修改電腦建檔……等電腦系統和改善措施的介入後，對減少開方錯誤率確實是有明顯的改善。

Background: Patient medication safety has been an important topic in recent years. Negligence in drug use can lead to a series of mistake. If luckily, the mistakes may be noticed in time and cause no harm to patients. Otherwise, it can cause serious harm to patients and thus increase the overall medical cost. The purpose of this study was to prevent medication errors and increase patient medication safety through analysis and discussion of the patterns and problems of prescription errors.
Methods: This research was a prospective study. Three steps were taken in this study. First, prescriptions (including out-patient, in-patient, and emergency) confirmed by clinical pharmacist, were collected from 97/01/01 to 97/12/31 and their patterns analyzed. Second, improvement plans, such as setting computer entry restriction, were developed based on errors analyzed. The improvements helps to decrease the rate of prescription errors made and results were evaluated.
Data establishment and statistical analysis, using t test as method, were performed via Microsoft Office Excel 2008 in this study. 
Results:  There were a total of 1327 prescriptions with errors collected during the period  from 97/01/01 to 97/12/31. Among the collected prescriptions nalyzed, there were 251 prescriptions with drugs, listed in the history of allergy, rescribed. 
This pattern of prescription errors has the highest error rate of 18.91% (n=1327). The next in line was inappropriate total quantity prescribed, which occurred in 194 (14.62%) out of the 1327 prescriptions. There were also 134(10.10%) prescriptions with inappropriate drug frequency used, 124(9.34%) with inappropriate dosages and118 (8.89%) with doctor entry errors made. For the part of doctors’ acceptance to the pharmacists’ advice, among the 1242 out of 1327 prescriptions being actually discussed with doctors, there was an 82.93% rate of acceptance.
Conclusion: Through modifications of computer systems and other procedures executed, this research found a clear and significant evidence of a improved and decreased rate of prescription error. By analyzing the cause of prescription problems, resolution can be taken to improve overall drug use safety.