Internal Medicine Residents’ Attitudes Toward Simulated Depressed
Cardiac Patients During an Objective Structured Clinical
Examination: A Randomized Study
Kathleen Crapanzano, MD, MACM
, Rebecca Hammarlund, PhD
Eric P. Hsieh, MD
, and Win May, MBBS, PhD
Department of Psychiatry, Louisiana State University Health Sciences Center, Baton Rouge, USA;
Department of Medical Education, University of
Southern California Keck School of Medicine, Los Angeles, CA, USA;
Division of Academic Affairs, Our Lady of the Lake Hospital, Baton Rouge, LA,
Department of Internal Medicine, University of Southern California Keck School of Medicine, Los Angeles, CA, USA.
BACKGROUND: Physician biases toward mental condi-
tions such as depression have been shown to adversely
affect medical outcomes.
OBJECTIVE: To explore the relationship between resi-
dents’ explicit bias toward depressed patients and their
clinical skills on a cardiac case during an objective struc-
tured clinical exam (OSCE).
DESIGN: Prospective parallel randomized controlled
PARTICIPANTS: One hundred eighty-five internal medi-
cine residents from three residency programs in two
INTERVENTION: During October–November 2015, resi-
dents were randomized to either a depressed or non-
depressed standardized patient (SP) presenting with
acute chest pain.
MAIN MEASURES: The Medical Condition Regard Scale
(MCRS) assessed residents’ explicit bias toward patients
with depression. Their clinical skills (history-taking,
physical examination, patient counseling, patient–physi-
cian interaction (PPI), differential diagnosis, and workup
plan) and facial expressions were rated during an OSCE.
KEY RESULTS: No significant relationships were found
between resident explicit bias and clinical skill measure-
ments. Residents who examined the depressed SP scored
lower, on average, on history-taking (t  = −2.77,
p < 0.01, Cohen’s d = 0.41) and higher on PPI (t  =
2.24, p < 0.05, Cohen’s d = 0.33) than residents examining
the non-depressed SP. There were no differences, on aver-
age, between stations on physical examination, counseling,
correct diagnosis, workup plan, or overall SP satisfaction.
Facial recognition software demonstrated that residents
with a non-depressed SP had more neutral expressions than
depressed-SP residents (t  = −2.46, p < 0.05, Cohen’s
d = 0.46), and residents with a depressed SP had more dis-
gusted expressions than non-depressed-SP residents (t
[83.52] = 2.10, p < 0.05, Cohen’s d = 0.28).
CONCLUSIONS: Extrinsic bias did not predict OSCE per-
formance in this study. Some differences were noted in the
OSCE performance between the two stations. Further
study is needed to examine the effects of patient mental
health conditions on physician examination procedures,
diagnostic behaviors, and patient outcomes.
J Gen Intern Med 33(6):886–91
© Society of General Internal Medicine 2018
A prior review of literature linking depression and coronary
artery disease (CAD) found evidence of a bidirectional relation-
ship, with depression being an independent risk factor for CAD
and its complications, and CAD in turn being a cause of depres-
Unfortunately, evidence suggests a lack of awareness of
such links between mental and physical conditions among phy-
sicians. Indeed, studies have shown that physicians who attribute
physical symptoms to mental conditions are less likely to inves-
tigate or treat the physical condition,
especially if they are
unaware of strong links between specific conditions, such as that
between CAD and depression. This sort of diagnostic
overshadowing is particularly likely in cases where the mental
illness is complex and accompanied by communication or be-
havioral challenges that interfere with the clinical presentation.
Diagnostic overshadowing is defined as the Bprocess by which a
person with a mental illness receives inadequate or delayed
treatment because of the misattribution of their physical symp-
toms to their mental symptoms,^
and can lead to suboptimal
care for both the mental illness and any comorbid condition.
Research has shown that, when presenting to the emergency
department (ED) with chest pain, patients with depression
were more likely to be triaged to a lower level and to experi-
ence a longer wait time before receiving an EKG or fibrino-
lysis, and were nine times as likely to miss benchmark times in
receiving a balloon procedure, even with a prior history of
acute myocardial infarction (AMI), than someone without a
comorbid diagnosis of depression.
Beyond the ED, people
with a history of depression were more likely to be referred to
and lower-quality hospitals
less likely to receive percutaneous transluminal angioplasty or
coronary artery bypass graft surgery.
Patients with acute
myocardial infarction who also had mental illness were
Received June 16, 2017
Revised October 30, 2017
Accepted December 11, 2017
Published online January 16, 2018