Design and implementation of a common drug information database for a university hospital

Background: Providing high quality problem-oriented drug information relies on the ability to easily collect appropriate background information on clinical cases, to access relevant information from published sources by defined search strategies and to store and retrieve previously answered questions. To do this efficiently, an easy-to-use, flexible and reliable drug information database is necessary. Methods: We designed and implemented an Intranet-based drug information database for a major university hospital in Germany. The overall design and the technical details of its design are discussed. We developed a generic, XML-based data model for pharmaceutical inquiries including a MeSH-oriented system of 99 pharmaceutical qualifiers to enable efficient indexing of questions and the searching of indexed questions. Results: The system provides query statistics and various search algorithms. The software implementation takes into account recent FDA recommendations for software used in clinical trials; internal review for quality control is supported. The database currently consists of 4224 records after 3.5 years of operation. Each inquiry consists of 50 items, 18 of 50 are categorized; 135 text elements support data entry. Our evaluation is focused on technical feasibility, user acceptance and query patterns. Conclusion: The intensive use and widespread acceptance of the database indicates a need for a computerized drug information system and suggests that Intranet technology can perform this task.