Modeling and Optimization of Clinical Processes Using EHR Data



 DESCRIPTION (provided by applicant): The goal of this K99/R00 award is to apply methods from biomedical informatics, operations research, and computer science to develop, implement, and evaluate stimulation and optimization tools for improving clinical processes. This is an important topic because electronic health record (EHR) systems have potential to improve quality and cost of health care, yet providers have raised concerns that EHR implementation has negatively impacted real-world productivity and efficiency. Methods for improving workflow based on automated time-motion data collection will have significant real-world impact. This proposal involves two phases: (A) K99 training and mentored research phase, which will include training in advanced operations research and analytics techniques at Oregon Health & Science University (OHSU) and Portland State University. Under the mentorship of two experienced informaticians and clinicians, a statistician, and a systems engineer, this training will be applie to research in improving clinical workflows. The K99 phase of research will focus on studying ophthalmology outpatient clinic workflows and the first two Specific Aims of the overall project: (SA#1) Develop tools for automated time-motion workflow data collection tools based on EHR timestamps and indoor positioning systems and (SA#2) Create simulation models to improve the efficiency of clinical workflow and validate these models by testing them in ophthalmology outpatient clinics. Ophthalmology will be an ideal clinical domain for performing these initial workflow studies because it is a fast-paced ambulatory specialty that includes medical and surgical patients, imaging tests, multiple examination stages (e.g., before & after eye dilation), and multiple ancillary staff members (e.g., technicians, photographers). (B) R00 research phase and transition to independence. The R00 phase of research will focus on the third Specific Aim of the overall project: (SA#3) Develop, implement, and evaluate data collection, simulation, and modeling techniques to broader inpatient and outpatient medical domains. This will generalize methods from the mentored project phase by collecting EHR data (e.g., surgery type, length of inpatient stay, demographics) to create probability density functions representing patient demand for resources. These data will be used to develop simulation models for different scheduling strategies based on patient classifications, and to evaluate them in the clinical setting. This project will benefit from a PI who has a strong background in mathematics and computer science, from an outstanding collaborative team of mentors with complementary experience across all areas of the proposed project, and from an outstanding academic informatics environment at OHSU.