Theses and Dissertations

ORCID

https://orcid.org/0000-0001-7472-6099

Advisor

Babski-Reeves, Kari

Committee Member

Johnson, Jenna

Committee Member

Gonzalez-Vargas, Jessica

Committee Member

Wang, Haifeng

Date of Degree

12-12-2025

Original embargo terms

Visible MSU Only 2 Years

Document Type

Dissertation - Campus Access Only

Major

Industrial and Systems Engineering

Degree Name

Doctor of Philosophy (Ph.D.)

College

James Worth Bagley College of Engineering

Department

Department of Industrial and Systems Engineering

Abstract

Diagnostic errors occur in 10–15% of cases and cost the U.S. over $100 billion annually, driven by both cognitive and system-level factors. Sepsis exemplifies these challenges, causing more than half of hospital deaths and $52 billion in yearly costs, with outcomes worsened by delayed recognition and inconsistent adherence to protocols. Current CDS tools and EHR systems often rely on binary alerts or raw data, offering limited diagnostic value while contributing to alert fatigue. User-centered, evidence-based CDS that integrates validated severity scoring is urgently needed to improve timely diagnosis and patient outcomes. This research applies a user-centered approach to develop and evaluate clinical decision support (CDS) tools that improve sepsis detection and treatment while reducing provider burden across varying expertise levels. The study examines enhanced visual display models that integrate patient data with validated sepsis staging scores to support timely and effective clinical decision-making. Usability testing with physicians and nurses, stratified by experience and expertise, identifies preferences and performance factors that inform evidence-based design. Studies optimized EHR alert design, evaluated the role of clinician expertise in sepsis diagnosis, and piloted a nurse-focused CDS tool that increased usability, adherence to sepsis bundles, and patient outcomes. Results highlight the need for CDS systems that are workflow-integrated, adaptable to provider expertise, and designed to minimize cognitive burden. Collectively, this work establishes evidence-based design principles that advance human factors and informatics while offering broadly applicable strategies for decision support in high-stakes, safety-critical environments.

Download

Share

COinS