A new physician-scientist at Cincinnati Children’s is combining his interests in user-centered design, electronic health records and neonatology to improve the way clinicians interact with software.
When Kevin Dufendach, MD began his neonatal fellowship, he was handed an 8 ½ by 11-inch piece of paper. Nurses and physicians in the neonatal intensive care unit (NICU) used these papers to share information about patients between day and night shifts. The printed text covered the basics: patient name, age, medical record number, and date of admission. But more pressing details for the sickest babies, like ventilator settings and medications, were jotted down by hand throughout the day.
The papers were as small as the babies, and the pens were even smaller—0.3mm, super fine tipped, and special ordered. It was the only way to fit handwritten notes of 16 babies’ care information on a single sheet.
“It killed me, seeing all of this information copied down,” he says. “My concern is that we spend so much time just writing information that we aren’t focusing on the higher-level thought that it takes to put a complex situation together and come up with an appropriate treatment plan.”
Dufendach researches ways to design smarter. He recently completed his medical training at Vanderbilt and joined Cincinnati Children’s last year as a clinician in Neonatology/Pulmonary Biology and a researcher in Biomedical Informatics. Aiming to offer users a quick and easy way to help design software, he has developed an open-source tool called VandAID focused on user-centered design. He is already applying this tool to solve a variety of care and research challenges.
Better software, better care
Medical software can be a big asset to clinicians, allowing them to streamline data management and care for patients with greater precision. But design is everything—if the software doesn’t meet clinicians’ unique needs, it could become more of a hindrance than a help.
This is why early involvement of end users in the design process for medical software is a major part of every project. But users often have a difficult time communicating what they want—it can be challenging to identify your design needs if you aren’t a designer yourself.
“It started as this idea that we need to be able to get input on human-centered design,” Dufendach says. “In order to make software that is going to be useful for people, you need to involve those people in the design process.”
VandAID—short for Vanderbilt Active Interface Design—is a web-based crowdsourcing platform. The application involves an interactive canvas where users can select the colors, fonts, locations, and other design choices that best match their needs. Each choice appears on the users’ screen in real-time—if they select a preference for diagrams over tables, they can immediately see how this would affect the display. If they change their mind after seeing the diagrams, they can easily switch back to tables.
Remembering his experience with the tiny notes in the NICU, Dufendach tested VandAID in designing neonatal patient lists. Neonatal clinicians from Vanderbilt University Medical Center participated in the study, where some of the group took part in traditional one-on-one participatory design (PD) groups, and others utilized the VandAID software. With its tangible examples and customization options, VandAID significantly outperformed the PD sessions in speed and ease-of-use.
“I see VandAID as a multifunctional tool,” he says. “Its use is not limited to a specific point in the design process—it just provides a link between a survey and a dynamic canvas.”
His tool is already proving useful at Cincinnati Children’s in a variety of applications, including a project Dufendach is working on with the eMERGE genetic research group. Researchers involved in the eMERGE project wanted to survey teenagers and parents about the type of information they would like to be given—or not given—in genetic testing results.
With the VandAID software, users can choose which genetic conditions will appear in their results. Teenagers can choose not to see results for conditions that usually don’t emerge until adulthood, like breast or colon cancer. Parents can choose to see results for conditions that are preventable or treatable only. Users save their preferences, which are then sent to REDCap, an existing survey application that can seamlessly integrate with VandAID for any project.
In the NICU, it’s crucial that parents are engaged with their baby’s care. Dufendach is also working on a neonatal-specific inpatient portal application that will make it easier for parents and NICU staff to communicate effectively. His research highlights the differences in adult care and pediatric care from a software standpoint—what works for adults doesn’t always work for children. A child’s evolving physiology, maturity, and relationship with their caregivers are very different from an adult’s. Dufendach seeks to create pediatric-specific functionalities in the electronic health record that will meet these unique needs.
When design begins with user needs, it ends with successful outcomes. His research with the biomedical informatics department will continue to ensure that designers and users are able to work smarter—super fine pens not required.
- Read Dufendach’s published articles on the VandAID trials and pediatric Electronic Health Records
- Access the open-source VandAID software