Clinical user interfaces that learn from experience

Abstract

Clinical data entry is one key to success in health information systems that is not a matter of technology alone, but of appropriateness and usability of design. We review the technology of adaptive user interfaces and learning agents. In these technologies we see the potential to improve the usability of general practice clinical workstations through machine-learnt adaptation to the user, the patient and the specific situation. Use of intelligent split menus that adapt based on past clinical encounters is one specific adaptive interface method that has shown potential by simulation. We are undertaking research in 'expert in the loop' use of data mining for iterative refinement of clinical workstation adaptation with an eye to significantly improving general practice data entry quality.

Keywords

• adaptive user interfaces
• clinical workstation
• data entry
• learning agents