© 2025 MJH Life Sciences™ , Patient Care Online – Primary Care News and Clinical Resources. All rights reserved.
If primary care clinicians are to meet the cognitive screening demands that are growing with an aging population, they need the sharpest tools in the box, this expert says.
Recent data on detection of mild cognitive impairment (MCI) in primary care suggests significant and widespread underdiagnosis. A study using administrative data for the full US population aged 65 and older revealed an average detection rate of 0.08 for both clinicians and primary care practices, translating to only about 8% of expected cases being diagnosed. Other findings suggest a rate of 11.4% and 15%.1
"We clearly need to empower primary care to be able to detect cognitive impairment. They are the folks on the front lines of this problem. And if you don't detect something, you can't start to manage it," said neurocognitive researcher Rodney Swenson, PhD, in a recent interview with Patient Care.©
Swenson and research colleague David Libon, PhD, spoke with editors about the growing field of precision neurocognition, a heuristic that is developing new methods that can operationally define early and very subtle neurocognitive changes that might be associated with cognitive decline. Swenson's research2 has led to digital assessment tools that can detect a wide range of the very earliest neurocognitive biomarkers that suggest declining brain health. These tools are what can make the difference in detection and diagnosis of MCI in primary care, Swenson emphasized during the interview.
In the video clip above, he highlights the need for the refined and actionable data available with the new technology, how the information can change the landscape of early detection in primary care, and the potential barriers to adoption that lie ahead.
The following transcript has been lightly edited for style and flow.
Patient Care: How do you envision the future for cognitive health assessment at the most basic level, in primary care?
Rodney Swenson, PhD: We clearly need to empower primary care to be able to detect cognitive impairment. They are the folks on the front lines of this problem. And yet, if you don't detect something, you can't start to manage it.
When we try to get cognitive health assessment implemented at the primary care level, there are a lot of roadblocks—and a lot of need for solutions to those roadblocks. Time is a big issue for primary care physicians and providers. I think they already are assessing for cognition—if you look at most annual wellness visits and so forth—but the instruments they're using are not sensitive enough or specific enough.
So, when you develop these digital types of assessments, they have to be brief and powerful. They need to give useful information to the primary care provider. Many of these digital assessments are more reliable, can be easily integrated into the electronic patient record, and can be available—ideally with a report—before the physician even steps into the room. That’s what you want to deliver: actionable, validated, and reliable information.
Along with these evaluations, it’s important to provide action plans that primary care providers can begin to discuss with their patients. The purpose isn’t just to determine if someone has a problem and then refer them to neurology or neuropsychology. The reality is that access to neuro specialists is limited. We want to make sure we're sending the right patients to the right specialists, and in the meantime, help primary care providers take action based on cognitive testing results.
Any effort like this has to start in primary care, but we also need to understand how to work through that person's clinic or health system—what’s realistic for them—so we can speed up getting people to the right place at the right time, and seeing the right professionals. Just as importantly, we need to equip providers with tools and guidance so they know what to do next. We've had feedback from primary care providers who say they’re unsure of the next steps after detecting a cognitive issue.
In rural areas—what we sometimes call ‘neural deserts’—the scenario is much different than in cities like New York or Boston. In these regions, primary care providers often have to handle these problems on their own. So we try to build digital platforms that give them that kind of support.
Rodney A Swenson, PhD is Clinical Professor of Psychiatry and Behavioral Sciences, at the University of North Dakota School of Medicine and Health Sciences in Grand Forks, ND.
David J Libon, PhD, is a Professor at the New Jersey Institute for Successful Aging, at Rowan University, in Glassboro, NJ.
For more from our conversation with Drs Libon and Swenson, see:
Related Content: