סמינר בניהול טכנולוגיה ומערכות מידע

מר לירן פוליטי מאוניברסיטת בן גוריון

Abstract

Billions of dollars have been invested worldwide in health information exchange systems (HIEs), which integrate patient-level data that originate in multiple information sources. These systems have been promoted as having great potential to change healthcare, particularly emergency medicine, by improving the quality of care and reducing costs. Despite the increase in the use of these systems, they have been explored in a limited way: their use is typically defined in a dichotomous manner (use versus no use), prominent antecedents of use (e.g., user attributes and information availability) are not explored, and actual, user-level patterns of use are seldom associated with clinical decisions. Moreover, although HIEs are often deployed in tandem with electronic medical records (EMR), studies have not examined how the availability of the latter affects HIE use.

This research focuses on critically-ill patients, treated at the internal emergency department of Soroka University Medical Center, in which Ofek HIE is deployed. Use patterns are described using system log files, which record the use of the HIE by physicians. The data recorded in the log files are complemented by data about the focal patients, physicians, situation, information availability, and system use dynamics. This rich dataset is leveraged to examine multiple research models, which describe use patterns as configurations (clusters), analyze how use patterns are associated with clinical decisions, and examine how use patterns are affected by a broad set of variables. In particular, I will present an analysis of how HIE use patterns have changed following the introduction of EMR alongside the HIE. I will also discuss the key implications for research on health information systems and for their design and implementation.