Developing an Intelligent Cross-Facility-Based Therapeutic Decision Reminder Engine

摘要

Background: The duplication of medication orders is a critical issue that causes medication errors for patients and wastes medical resources? National Health Insurance of Taiwan provided a NHI-PharmaCloud web-based database, and it contains the most complete personal drug history across different healthcare facilities in Taiwan. Currently, most healthcare facilities check the patients' cross facility medical information from NHI-PharmaCloud manually. However, new drugs are released quickly and frequently, and the same ingredients or the same pharmacological mechanisms can be hard to detect, therefore, it is difficult for clinical staff to prevent duplicated medication orders. Methods: To prevent duplicated medication orders, to add more flexible duplicate medication order checking rules, and to improve patient drug-use safety, we developed an Intelligent Cross Facilities Based Therapeutic Decision Reminder Engine (TDRE) [Fig. 1.] based on NHI-PharmaCloud in Taipei Medical University Hospital. In addition, to ensure TDRE performance in future implementations, we also conducted load testing to determine how the maximum workload TDRE would be able to support once it is deployed on a basic hardware system. Results and Discussion: The maximum affordable workload of TDRE in our deployed environment is 1100 requests per second. From the test results we can infer that TDRE would be able to provide 528,000 (1100 × 60 × 8 = 528,000) requests during an eight-hour working day. The test results show that TDRE worked properly and is able to fulfill our expected goal. Field-testing at Taipei Medical University Hospital indicates that the average service volume (inpatient plus outpatient) is approximately 6000 to 7000 encounters per day. Therefore, TDRE should be able to support this service quantity.