Evaluating clinician response to interruptive alerts in intensive care units

The National Heart, Lung, and Blood Institute awarded Regents of the University of Michigan $193,952 on December 1, 2026, under the Cardiovascular Diseases Research program (CFDA 93.837) to evaluate clinician response to interruptive alerts in intensive care units.

The research, led by Dr. Hechtman, adapts human factors engineering metrics to measure alert engagement and response in clinical decision support systems implemented through electronic health records. Aim 1 evaluates clinician behavior during and after interruptive alerts in seven Michigan Medicine ICUs. Aim 2 identifies patient, clinician, alert design, and contextual clinical factors influencing alert response using multilevel cross-classified modeling to understand relationships between interconnected factors. Deliverables include 1–2 first-author publications yearly and presentation at 1–2 national conferences yearly. Results will serve as foundational work for a K23 career development application.

Performance occurs in Ann Arbor, Michigan, with a period of performance ending May 31, 2027. This is a Project Grant assistance type.

Abstract

Clinical problem to be addressed: clinical decision support (CDS) systems have great potential to improve management of critically ILL patients by reducing errors, expediting diagnosis and treatment, streamlining clinician workflow, and disseminating information when rapid practice change is necessary. CDS is often implemented via interruptive alerts in electronic health records (EHRS). Studies indicate up to 96% of these alerts are ignored or overridden. However, we lack standard metrics to fully evaluate clinician responses, typically categorizing them simply as overridden or accepted. This dichotomization fails to capture the complexity of responses, such as not reading the alert, disagreeing with the recommendation, or following the alert recommendation after closing the alert. Accurate measures of alert engagement and response and better understanding of the factors influencing alert response are essential to improving alert effectiveness. Specific aims: in aim 1, DR. Hechtman will evaluate clinician behavior during and after interruptive alerts in seven michigan medicine icus (2019-2024). She will adapt standard metrics from human factors engineering to accurately measure alert engagement and response. In aim 2, she will identify patient, clinician, alert design, and contextual clinical factors influencing alert response. She will use multilevel cross-classified modeling to understand the relationships between these interconnected factors. Deliverables: DR. Hechtman's aims will lead to 1-2 first-author publications yearly and presentation at 1-2 national conferences yearly. The results will serve as foundational work for her K23 application, which will develop a necessary framework for evaluating alert effectiveness and burden. Candidate's career goal and development plan: DR. Hechtman's goal is to be a physician-scientist leading multidisciplinary research focused on understanding and optimizing clinical decision support to improve management of critically ILL patients. Through the training outlined in this F32 award, she will build on her foundational master's training in quantitative methods to develop expertise in electronic health record (EHR) data analysis, human factors engineering, and advanced statistical methods. Candidate background: DR. Hechtman is a T32-sponsored research fellow and early career physician- scientist in the university of michigan's division of pulmonary and critical care medicine. She has clinical expertise in critical care medicine and a foundation in quantitative research methods through an MSC in health and healthcare research. She has published 9 manuscripts (4 first-authored original research publications). Her division has a strong commitment to supporting her career progression. They selected her to join their T32 grant in june 2024 and protected her time to complete a master's in health care research. Her primary mentor, DR. Hallie prescott, has a strong record of successful mentorship and has mentored 4 clinician- scientists on K awards. DR. Hechtman has selected a multidisciplinary CO-mentoring team that is well suited to support successful completion of her research and training aims.