Expert Systems with Applications: An International Journal
Designing adaptive feedback for improving data entry accuracy
UIST '10 Proceedings of the 23nd annual ACM symposium on User interface software and technology
| Hi-index | 0.00 |
This paper describes an effort to make computer interfaces more intelligent in facilitating the coding of clinical information. We believe the interface should be sufficiently efficient and easy-to-use that a physician can code information during the consultation without detracting from doctor-patient interaction. In this way, the benefits of a "clinical workstation" setting, such as best practices guidance and drug interaction detection, are maximised. We pursue the strategy of applying machine learning to existing databases of electronic medical records to develop probabilistic models of general practice. Based on this model, we have simulated and prototyped data entry interfaces with "hot lists" (short pick-list menus of relevant items) and dynamic graphical depictions of contextually-likely clinical data.