Analyzing Visual Search Techniques using Eye Tracking for a Computerized Provider Order Entry (CPOE) Task

Abstract

A challenge with current Computerized Provider Order Entry (CPOE) systems includes patient identification errors, i.e. when an incorrect patient’s record is referenced. These types of errors can lead to patient safety issues such as administrating medication to the incorrect patient. Eye tracking technology can provide insights into the visual search patterns of healthcare professionals and shed light on how patient identification errors occur. This study investigates whether there are differences in visual search metrics, response time, and accuracy when searching for a patient by two identifiers – name or date of birth – from a list of patients with similar names. The findings revealed there was no effect of search strategy on speed or accuracy; however, there was an effect on fixation duration and number of fixations within specific areas of interest. Across both search strategies, there were more fixations on names. This demonstrates the importance of a patient’s name regardless of search strategy and is an important consideration to take into account if multiple patients share the same name. This study shows that eye tracking technology can be used to investigate the visual search patterns employed during patient identification and provide insights as to how patient identification errors occur. It also demonstrates a need to develop alternative methods to prevent patient identification errors apart from relying on healthcare professionals to verify patient identity.