Using Functional Neuroimaging Combined With a Think-Aloud Protocol to Explore Clinical Reasoning Expertise in Internal Medicine
Military Medicine ,
Sep 2012
Durning, Steven J. , Graner, John , Artino, Anthony R. , Pangaro, Louis N. , Beckman, Thomas , Holmboe, Eric , Oakes, Terrance , Roy, Michael , Riedy, Gerard , Capaldi, Vincent , et al.
Using Functional Neuroimaging Combined With a Think-Aloud Protocol to Explore Clinical Reasoning Expertise in Internal Medicine
MILITARY MEDICINE
Using Functional Neuroimaging Combined With a Think-Aloud Protocol to Explore Clinical Reasoning Expertise in Internal Medicine
Steven J. Durning 3
John Graner 3
CDR Anthony R. Artino Jr.
MC USN 3
Louis N. Pangaro 3
Thomas Beckman 0
Eric Holmboe 2
Terrance Oakes 3
COL Michael Roy
MC USA 3
Gerard Riedy 4
CPT Vincent Capaldi
MC USA 4
CPT Robert Walter
MC USA 4
Cees van der Vleuten
Lambert Schuwirth
0 Mayo Clinic , 200 First Street SW, Rochester, MN 55905 , USA
1 77 , Washington, DC 20307 , USA
2 American Board of Internal Medicine , 510 Walnut Street, Suite 1700, Philadelphia, PA 19106 , USA
3 Uniformed Services University of the Health Sciences , 4301 Jones Bridge Road, Bethesda, MD 20814 , USA
4 Walter Reed Army Medical Center , 6900 Georgia Avenue, NW , USA
Background: Clinical reasoning is essential to medical practice, but because it entails internal mental processes, it is difficult to assess. Functional magnetic resonance imaging (fMRI) and think-aloud protocols may improve understanding of clinical reasoning as these methods can more directly assess these processes. The objective of our study was to use a combination of fMRI and think-aloud procedures to examine fMRI correlates of a leading theoretical model in clinical reasoning based on experimental findings to date: analytic (i.e., actively comparing and contrasting diagnostic entities) and nonanalytic (i.e., pattern recognition) reasoning. We hypothesized that there would be functional neuroimaging differences between analytic and nonanalytic reasoning theory. Methods: 17 board-certified experts in internal medicine answered and reflected on validated U.S. Medical Licensing Exam and American Board of Internal Medicine multiple-choice questions (easy and difficult) during an fMRI scan. This procedure was followed by completion of a formal think-aloud procedure. Results: fMRI findings provide some support for the presence of analytic and nonanalytic reasoning systems. Statistically significant activation of prefrontal cortex distinguished answering incorrectly versus correctly ( p < 0.01), whereas activation of precuneus and midtemporal gyrus distinguished not guessing from guessing ( p < 0.01). Conclusions: We found limited fMRI evidence to support analytic and nonanalytic reasoning theory, as our results indicate functional differences with correct vs. incorrect answers and guessing vs. not guessing. However, our findings did not suggest one consistent fMRI activation pattern of internal medicine expertise. This model of employing fMRI correlates offers opportunities to enhance our understanding of theory, as well as improve our teaching and assessment of clinical reasoning, a key outcome of medical education.
INTRODUCTION
Expertise in clinical reasoning improves patient care1,2 and
involves establishing a diagnosis and an individual care plan; it
is an essential outcome of medical education. Clinical
reasoning encompasses the sum of the thinking and decision-making
processes within clinical practice.2 However, investigating
clinical reasoning involves making inferences about internal
mental processes. Despite being at the core of clinical
expertise, the difficulty in assessing clinical reasoning greatly
challenges our understanding.3
Until recently, researchers have largely avoided
investigating the functional activation of complex psychological
constructs such as cognition.4 However, functional magnetic
resonance imaging (fMRI) enables the assessment of regional
brain activity by detecting changes in oxygenation reflecting
perfusion. Regions that use more oxygen to process stimuli (an
“artifact” of thought) evoke a rapid feedback loop that dilates
blood vessels to more markedly increase oxygenation to such
areas. These changes can be reliably detected by fMRI (blood
oxygen level-dependent signal). Pairing fMRI with a
thinkaloud procedure, whereby participants are asked to vocalize
whatever comes to mind as they work through a task such as
solving a clinical case, could facilitate identification of links
between cognition and its neural substrate, providing insight
into the cognitive processes underlying clinical reasoning.
Clinical reasoning, like many outcomes in medical
education, is difficult to measure. One current gold standard for
physician performance is multiple-choice exams. Indeed, to
become a practicing physician, one must pass a series of
multiple-choice question (MCQ) examinations.
A primary goal of the Long-Term Career Outcome Study is
tracking the performance (outcomes) of our graduates. Given
the limitations of MCQ examinations, we therefore
investigated fMRI neuroimaging patterns of our graduates and other
graduates who practice in the Department of Defense to
enhance our understanding of what these MCQ examinations
are measuring. Such understanding could ultimately help us to
establish the limits of these MCQ examinations and plan new
outcome measures that could be used in our and other systems.
Results (...truncated)
This is a preview of a remote PDF: https://academic.oup.com/milmed/article-pdf/177/suppl_9/72/21146228/milmed-d-12-00242.pdf
Durning, Steven J., Graner, John, Artino, Anthony R., Pangaro, Louis N., Beckman, Thomas, Holmboe, Eric, Oakes, Terrance, Roy, Michael, Riedy, Gerard, Capaldi, Vincent, Walter, Robert, van der Vleuten, Cees, Schuwirth, Lambert.
Using Functional Neuroimaging Combined With a Think-Aloud Protocol to Explore Clinical Reasoning Expertise in Internal Medicine ,
Military Medicine,
2012, pp. 72-78, Volume 177, Issue suppl_9, DOI: 10.7205/MILMED-D-12-00242