The Sharpest Scalpel Interview with Dr. Lola Ogunyemi( continued)
How does this work make a convincing case for NIH and other funding sources to continue supporting the everincreasing mandate to eradicate health inequities? What results keep them interested in your efforts?
A lot of the work we do is focused on strategies for improving health outcomes in safety net settings, which often suffer from being underresourced or having a shortage of specialists. We partner with organizations and providers who are providing access to health care regardless of patients’ health insurance status or ability to pay.
The data and knowledge that are generated in such settings can be used to learn about patterns of disease and to figure out ways to use computational approaches such as machine learning to improve assessment of disease, patient outreach and ultimately, the quality of care that is provided. I believe that the NIH values research that involves partnerships outside of traditional academic medical centers, with health care organizations that are often understudied in research, even as they bring care to communities that suffer disproportionately from chronic diseases.
Artificial Intelligence has become an important focus area within medicine and science training curricula. Why should medical students, residents, clinicians and other professionals become aware of AI’ s potential benefits and drawbacks as they embrace the technology?
Artificial intelligence will be a component of the health informatics master’ s program I described previously and is something that medical students, residents and providers should be familiar with, whether we’ re talking about machine learning, natural language processing, or rule-based systems used to provide clinical decision support within electronic health records systems. Clinicians should understand the potential benefits and limitations and the very real problem of algorithmic bias.
AI tools developed from unrepresentative data may end up causing harm, as we’ ve seen recently in the non-medical space. For example, there have been facial recognition algorithms in software sold to and used by police that weren’ t trained using sufficient facial data from people of color and as a result, we’ ve had African American men erroneously flagged by such software and arrested for crimes they did not commit. Clinicians should understand enough to ask questions about the representativeness of the underlying data used to create the tools that they utilize.
Informatics is increasingly becoming incorporated into medical school( and nursing) curricula, the AAMC and CCNE have informatics requirements, the ACGME has informatics requirements and the American Board of Preventive Medicine voted to approve subspecialty certification in Clinical Informatics in 2011. The ACGME released program requirements for graduate medical education in Clinical Informatics in 2014. ACGME-accredited residency programs in Family Medicine and Internal Medicine can oversee Clinical Informatics fellowships. It would be wonderful to have Clinical Informatics Fellows at CDU in addition to the Clinical Informatics Fellows currently at UCLA.
What is the best way that a young person with a passion for computer science and informatics can prepare for a career in medicine?
Many of my academic mentors in biomedical informatics were medical doctors who had undergraduate degrees, master’ s degrees or doctorates in computer science, but there are multiple paths for pursuing these interests now. Computer science students can be pre-med and go on to medical school and non-computer science students who realize that they have an interest in informatics later on don’ t have to get additional degrees.
These days, medical students and residents are exposed to informatics methods through AAMC and ACGME requirements and can be boardcertified in Clinical Informatics if they so choose. The vast majority of students who take my biomedical informatics course at CDU intend to go on to medical or dental school and are often amazed at the ways in which modern medicine relies on health information technology such as electronic health record systems as well as informatics methods that help improve patient outcomes.
CDU College of Medicine | PG. 15