Web intelligence, Multi-agent, Evidence based medicine, Medical decision support
The purpose of this study was to develop and experiment EBMwiz medical information management system. The research and development were applied. This study consists of developmental phase and experimental phase. In developmental phase, the EBMwiz was developed by using multi-agent technique. The EBMwiz consist of eight agents that work together simultaneously: 1) Keyword performing agent, this agent works with PICO formulation. 2) Search engine agent, this agent working as a search engine. 3) Ranking agent, this agent ranks article by level of precision (level of evidence & level of recommendation). 4) Fulltext retrieving agent, this agent retrieve available online full-text. 5) Summarizing agent, this agent summarize user-selected article. 6) EBM Reporting agent, this agent prepares final report of overall evidence in term of level of evidence & level of recommendation. 7) Translator agent, this agent working as embedded dictionary. And, 8) Sharing tools agent, this agent working as web board. In experimental phase, the sample was 60 medical students that divided to two groups. First group (n=30) was using EBMwiz then using PubMed. Second group (n=30) performed vise versa. The descriptive statistics and Mann Whitney U-Test were applied. The finding show that medical student who used the EBMwiz searching evidence had statistically higher clinical decision effective than medical student who used routine searching both first exam and second exam at p-value 0.002 and 0.000, respectively. In conclusion, the EBMwiz will be provided an effective and efficient way of organizing medical information that helps medical student to retrieve medical information to improve the quality of health care.