Intelligent Clinical Decision Support for Perioperative Blood Management

Project summary/abstract Dr. Lou is a cardiothoracic anesthesiologist with a long-term career goal to develop, implement, and disseminate intelligent clinical decision support interventions that improve the care of surgical patients, with a current focus on perioperative blood management. Her prior research training has allowed her to develop expertise in data science and predictive model development using artificial intelligence (AI) techniques. This proposal builds on Dr. Lou’s prior experience by providing the protected time, mentorship, and training necessary to develop skills in the implementation of AI models as clinical decision support (CDS) tools, and in the evaluation and dissemination of such tools. The proposed career development plan will involve mentorship, didactic, and practical research training in the conduct of pragmatic clinical trials, the design and evaluation of implementation strategies for CDS tools, governance for CDS interventions, and foundational training in clinical research. She has brought together a multi-disciplinary group of mentors and advisors, including Thomas Kannampallil, PhD, Michael Avidan, MBBCh, and Sachin Kheterpal, MD, MBA, each of whom have expertise in clinical trials and the implementation of health technology innovations. Washington University provides the ideal environment for Dr. Lou’s training given its infrastructure and strengths in informatics, clinical trials, and implementation science. The proposed research plan provides the framework for Dr. Lou to acquire the skills she needs to achieve her career goals. Presurgical testing and preparation for surgical transfusion is essential for patient safety during surgery, yet excessive preparation is costly and contributes to blood product waste. Given recent blood shortages, presurgical blood orders should be placed only for patients who need it; there is an acute public health need for tools that accurately estimate the risk of transfusion to guide clinical decision- making. In preliminary work, Dr. Lou has developed a personalized AI model, named S-PATH, to estimate surgical transfusion risk, and demonstrated its validity across local and national datasets. The objective of this proposal is to evaluate the feasibility and preliminary effectiveness of S-PATH as a CDS system embedded within the Electronic Health Record (EHR) using a pragmatic cluster-randomized clinical trial. The expected outcome is a generalizable, EHR-agnostic personalized CDS system to guide presurgical blood orders that is feasible to deploy within preoperative workflow. This proposal is significant in its potential to change clinical practice, with considerable public health impact for patient safety, blood conservation, and reduced healthcare costs. The proposed research and training will provide Dr. Lou with the skills needed to launch an independent research program to improve the delivery science for AI innovations in healthcare.