When Naveed Afzal, Ph.D., worked at the Mayo Clinic, he helped build an algorithm that could read through a patient’s clinical notes and flag symptoms that were associated with a cardiovascular disease known as peripheral arterial disease. Say a patient came into the office with an irregular heartbeat; physicians could use the algorithm for assistance instead of reading through decades worth of patient history to see if symptoms were building up over time.
“If the patient is showing the symptoms, the algorithm points out that context window in the notes, so then the physician can recommend the lab test — reducing the time for diagnosis,” Afzal says.
Afzal, who now serves as the head of data science at Takeda Pharmaceuticals, recently visited Associate Professor and Healthcare Informatics Program Director Nafees Qamar Ph.D.’s “Electronic Health Records” (EHR) course and spoke with graduate students about how AI and data are changing healthcare. From discussing the hype around generative AI to ethical decision making, Afzal noted that AI integration is essential for a healthcare organization's sustainable growth.
“Data is crucial for business success, but you have to ensure that you have the right data framework in place to collect, govern, and manage that data,” says Afzal, adding that the Healthcare Informatics students will play critical roles in creating these frameworks.
From theoretical to hands-on learning
EHR, a digital repository that houses a patient’s medical information that can be shared across the individual’s providers with consent, is the backbone of the healthcare system today. To prepare Healthcare Informatics students to work with these systems, Qamar’s course focuses on EHR fundamentals — such as how EHRs are built, what architecture is needed, and what types of networks are required for set up.
Following this coursework up with hands-on learning, graduates practice using an EHR system — a fully-developed EHR system for academia that uses artificial patient data — to learn how to effectively operate EHR; the system includes demographic information, allergies, vitals, alerts, medical notes, and medical reports for each patient.
“One of the constraints in teaching EHR is that these systems are vendor based, which means that you cannot have access,” Qamar says. “At Bryant, we have an innovative way to teach electronic health records because we have integrated one that’s as powerful as any other system being used by a healthcare provider.”
RELATED ARTICLE: Grad student blends tech and healthcare with big data research
Using EHR, students start with simple clinical workflows like patient registration and move onto complex workflows such as medical billing, coding workflows, and data extraction and analysis. With intensive EHR activities each week, students also learn how to acquire and implement a new EHR. Beginning with a request for proposal to potential vendors, they look at different features for what each EHR provides and use software project management tools to develop an implementation strategy for the system that they plan to acquire.
The major project, Qamar notes, is having students choose a New England-based hospital and — using the organization’s website — investigate service offerings, needs, and current trends. Based on that information, they aim to implement a new EHR for that organization.
“Many of the students have a healthcare background, and they always share with me that they did not have a chance to work on an EHR before — despite the fact that they were working in healthcare as a dentist or as a pharmacist,” Qamar says.
Preparing for the future
Having wrapped up his presentation and taking questions from the crowd, Afzal offers advice on how the students could align with industry needs and prepare for forthcoming technological advancements.
“Get to know the industry you are working in. How does the current workflow work? Talk to folks and ask what are the pain points that they would like solutions to,” Afzal says, noting that a more collaborative approach combined with active listening can result in more effective data frameworks.
Highlighting elements that will help them stand out after graduating, Afzal recommends students conduct their own research, participate in data science competitions so they can see other people’s approaches for frameworks and solutions, and practice talking about a solution they built in lay terms.
Noting that EHR data sets can help students work on preventive care, Qamar adds that this technology will help provide better care to patients.
“The source of all prevention and treatment is EHR,” he says.
