DESCRIPTION (provided by applicant): Background/Rationale: According to the Institute of Medicine, over 1 million preventable adverse drug events (ADEs) occur in the U.S. each year. To reduce the risk of patient harm, the VA's computerized provider order entry (CPOE) system includes automated medication alerts, such as drug-drug interaction alerts, to warn prescribers about potential problems. Unfortunately, compelling evidence indicates that VA alerts do not provide sufficient information to assist prescribers as they make decisions while ordering medications. Alert design can be informed by human factors research; human factors engineers focus on how to design technologies, processes, and work systems to match inherent human capabilities and limitations. Objectives The research objective of this career development award (CDA-2) is to assess the strengths and weaknesses of VA alert designs compared to other systems and identify strategies to enhance VA medication alerts for clinical decision-making. The specific aims are as follows: Aim 1: Identify the cognitive strategies that prescribers use for medication decision-making when faced with medication alerts, to provide evidence on how to improve alert design. Aim 2: Develop and test new strategies for alert content and display that aid decision- making at the earliest possible stage of the order process, in order to promote safer prescribing decisions. Aim 3: Evaluate whether the promising new alert design improves medication management. Each of these aims incorporate human factors techniques. There are few human factors researchers in VA; this CDA-2 experience will prepare the candidate to become an independent VA human factors researcher, with specialized expertise in information technologies for prescribing, medication management, and patient safety. Methods For Aim 1, cognitive task analysis, including critical decision method and simulation interviews, will be conducted with prescribers to uncover their decision-making process for drug-allergy alerts, drug-drug interaction alerts, and drug-disease alerts. For Aim 2, two new alert designs will be developed and evaluated to assess which design most effectively supports prescriber comprehension, since this is essential to support decision-making. Finally, for Aim 3, a simulation study will be conducted with prescribers to compare the new alert design against the VA CPOE system to assess whether the redesign improves clinical decision-making outcomes. As part of Aim 3, we will also test the feasibility of implementing design changes and collecting clinical outcomes at one VAMC. Significance This CDA-2 research will identify VA alert design strategies to better support clinical decision- making. This research is essential to reduce preventable harm and advance VA health information technology for the future.

The VA's computerized provider order entry (CPOE) system is used by VA prescribers nationwide and includes automated medication alerts, which are intended to prevent medication errors during order entry. This research will help the VA understand how to better integrate medication alerts into patient care and improve alert designs for prescriber decision-making and patient safety. Findings will be shared with local and national VA informatics and patient safety leaders. This research will provide a foundation for developing design guidelines for the medication alert interface and may ultimately be used to improve medication safety for veterans.