Essential questions: accuracy, errors and user perceptions in a drag/drop user-composable electronic health record.

In previous work we have described the creation and user testing of a drag/drop user-composable electronic health record, MedWISE. Any new design poses new potential problems; here we discuss the accuracy, potential for new types of errors, and user reactions to the approach, including their perceptions of ease of use and usefulness. Our results come from a mixed methods laboratory study with 18 clinicians at a large academic medical center. 12 used MedWISE and 6 used the conventional system. MedWISE users had comparable assessment accuracy to use of the conventional system, low (1/10) risk of diagnosis momentum error, and overall favorable and even enthusiastic user perception. 10/12 said the system eased their cognitive process, Ease of use and usefulness were rated 3.79 and 4.00 on a 5-point Likert scale. Users were unconcerned about the possibility of increased errors due to their trust in colleagues and system similarities to current practice. After first describing the issues, we suggest a method to elucidate risks in innovative and current systems.