Assistant Professor Jane Carrington Awarded NIH Grant to Develop and Test Electronic Health Record Algorithm

Nov 5, 2014

A clinical event, defined by Jane M. Carrington, PhD, RN, assistant professor at the University of Arizona College of Nursing, as a sudden and unexpected change in patient condition – including pain, bleeding, fever, or changes in output, respiratory status or levels of consciousness – is linked to a higher risk of unexpected patient death.

However, the vast amount of information contained in electronic health records (EHRs) and nurses’ reliance on patient information exchange through verbal, face-to-face handoffs can lead to these changes being overlooked.

To enhance nurse communication about these clinical events through augmentation of existing EHR technology, Dr. Carrington has been awarded an R01 (R01EB020395) grant from the National Institutes of Health (NIH), National Institute of Biomedical Imaging and Bioengineering, totaling $745,417 as part of the National Science Foundation/NIH Smart and Connected Health Program.

“We’re going to use the words and patient data nurses use to record patient care to trigger an alert when a clinical event is imminent or occurring and its severity,” said Dr. Carrington. “Our goal is to teach the computer how to read nursing language and recognize when an alert should be issued.”

Dr. Carrington, who earned her Doctor of Philosophy in Nursing in 2008 through the UA College of Nursing, says the alert will help ensure clinical events are not overlooked by nursing staff and support earlier intervention, if required.

Dr. Carrington is working with Mihai Surdeanu, PhD, associate professor at the UA School of Information Science, Technology and Arts (SISTA), and Angus Forbes, PhD, assistant professor in the Department of Computer Science at the University of Illinois at Chicago, to collect and translate the data into a smart algorithm using machine learning and natural language processing. They then will use innovative visualization strategies to enhance nurse-to-nurse communication of the clinical event.

“Once we’ve completed our first round of data collection, we will have a corpus of nursing words used to describe patients,” said Dr. Carrington. “We anticipate that a percentage of these data will contain the words used to describe a clinical event. Using these words, we’ll create a smart algorithm that will teach the computer to distinguish between an event and a non-event, to know when to trigger and not to trigger an alert.”

-------------------------------------

Faculty at the University of Arizona College of Nursing envision, engage and innovate in education, research and practice to help people of all ages optimize health in the context of major life transitions, illnesses, injuries, symptoms and disabilities. Established in 1957, the college ranks among the top nursing programs in the United States. For more information about the college, please visit its website, www.nursing.arizona.edu.