If You Build It, They Will Be Identified

2013 - 1 February - Pain Management in the ICU
Vision Grant recipient aims to build a database that can better identify an individual patient’s risk for postoperative complications, such as acute kidney injury.
 
Azra Bihorac, MD, MS, FCCM, FASN, had an interest in improving postoperative outcomes, knowing that reducing these preventable complications by 20% could save thousands of lives and reduce costs by $5 billion annually within the United States. Among the most common and deadly postoperative complications is acute kidney injury (AKI). Postoperative AKI, characterized by even small increases in routinely measured serum creatinine levels, is associated with up to a 10-fold increase in hospital mortality and decreased survival for up to 15 years after surgery. The Society of Critical Care Medicine (SCCM) awarded Bihorac its Vision Grant to propel research on a novel automated database aimed at predicting and identifying postoperative AKI risks.
 
"AKI injury occurs in patients who are usually relatively healthy before they experience an event such as trauma or surgery. AKI appeared to be an important complica­ tion that predisposed patients to unfavorable short- and long-term outcomes," noted  Bihorac. "I was frustrated with our inability to identify  patient's risk, in real time, for developing such complications."
 
Today, clinicans and scientists have few, if any, tools to quantify a patient's  individual risk  for postoperative complications. Bihorac  and her team, with funding from the Society, are seeking  to develop an integrated perioper­ ative electronic  medical record.  This software would apply a  novel, computational approach, utilizing machine learning algorithms,  to develop and validate a  probabilistic risk score for postoperative AKI in the intensive care unit.
 
"The major obstacle to advancing the development of an automated process for 'on-the-fly' risk assessment has been the lack of available technologies for real-time data integration and appropriate analytic methods. Our long-term goal is to build a partnership among compu­ting, engineering, health  informatics, and  clinical experts to develop, evaluate and implement an 'intelligent perioperative system.' This system will integrate electro­nic medical records and data-driven, model-based algorithms with clinical reasoning to improve decision-making," noted Bihorac during the application process.
 
This open-source, computational algorithm may be applied within databases across institutions, eventually offering widespread benefits and opportunity for improving patient care.
 
The $50,000 Vision Grant is awarded to an SCCM member whose work supports education, multiprofessional teamwork, outcomes measures, and reporting and continuous improvement. Interest is broad, offering a unique opportunity for investigators at every level. For example, the grant may be awarded to those who focus on technical aspects, such as electronic surveillance systems, but also to those exploring staff cultural and educational factors that either impede or facilitate the promotion of best practice and error reduction. Priority is given to projects that have broad relevance. Bihorac is an Assistant Professor of Anesthesiology and Surgery and Associate Division Chief for Research in the Division of Critical Care Medicine at the University of Florida School of Medicine in Gainesville, Florida, USA. She has been supported by the National Institutes of Health and several industry grants. The Society is proud to fund the efforts of Bihorac and her team this year through the Vision Grant.