ACAD EMERG MED d July 2005, Vol. 12, No. 7 d www.aemj.
org
Does Simulator Training for Medical Students Change
Patient Opinions and Attitudes toward Medical Student
Procedures in the Emergency Department?
Abstract
Objectives: To determine how simulator training impacts
patients’ preferences about medical student procedures in
the emergency department. Methods: A questionnaire was
administered to a convenience sample of 151 of 185 patients
approached (82% participation) seen in the emergency
department of a midwestern teaching hospital. The ques-
tionnaire asked how many procedures they would prefer a
medical student have performed after mastering the proce-
dure on a simulator before allowing the medical student to
perform this procedure on them. The procedures included
venipuncture, placement of an intravenous line, suturing
the face or arm, performing a lumbar puncture, placement
of a central line, placement of a nasogastric tube, intubation,
and cardioversion. These results were compared with those
of a similar study asking about the same procedures without
the stipulation that the skill had been mastered on a
The days of ‘‘see one, do one, teach one’’ for medical
students in the emergency department (ED) are over.
Patients expect that students have experience before
they are willing to allow medical students to perform
procedures on them.1 This creates a dilemma for
academic EDs. Students must learn technical skills,
and patients need to be informed if a student is
treating them and performing a procedure. One pos-
sible solution to this dilemma is the use of simulator
training.2–5 Mastering a procedure on a simulator may
favorably predispose patients to allow medical stu-
dents to perform procedures. The purpose of this
study was to determine if mastering a procedure on a
simulator changes ED patients’ preferences about
having medical students perform procedures on
them as part of the students’ training.
From the Departments of Emergency Medicine (MAG, CW, LK) and
Family Medicine (MAG, CW, LK, YX), University of Iowa Carver
College of Medicine, Iowa City, IA.
Received August 3, 2004; revisions received September 27, 2004, and
January 7, 2005; accepted January 9, 2005.
Address for correspondence and reprints: Mark A. Graber, MD,
Departments of Emergency Medicine and Family Medicine,
Pomerantz Family Pavilion, University of Iowa, 200 Hawkins
Drive, Iowa City, IA 52246. Fax: 319-338-6283; e-mail: mark-graber@
uiowa.edu.
doi:10.1197/j.aem.2005.01.009
635
Mark A. Graber, MD, Christopher Wyatt, BS, MBA,
Leah Kasparek, BA, Yinghui Xu, MS
simulator. Results: A high of 57% (venipuncture) and a
low of 11% (placement of a central line) would agree to be
a student’s first procedure after simulator training. Except
for intubating and suturing, participants were more likely
(p , 0.05) to allow a medical student to perform a procedure
on them after simulator training than without simulator
training. Many patients prefer not to have a medical student
perform a procedure no matter how many procedures the
student has done (low of 21% for venipuncture, high of 55%
for placement of a central line). Conclusions: Patients are
more accepting of medical students performing procedures
if the skill has been mastered on a simulator. However,
many patients do not want a medical student to perform a
procedure on them regardless of the student’s level of
training. Key words: medical education; medical ethics.
ACADEMIC EMERGENCY MEDICINE 2005; 12:635–639.
METHODS
Study Design. This was a prospective study of a
convenience sample of patients who registered for ED
care from June to August 2003. Our institutional
human subjects review board approved this project.
Study Setting and Population. A convenience sam-
ple of ED patients was included in this study. After
consenting to participate, participants completed a
questionnaire asking about their willingness to have
a medical student perform a procedure on them after
mastering the procedure on a simulator. All patients
registering for ED care were eligible to participate, with
the exception of prisoners, psychiatric patients, pa-
tients in extremis, non–English-speaking patients, and
patients younger than 18 years of age. Prisoners were
excluded because they may feel an obligation to accept
care from any source, and their views may not reflect
the views of the general population. Psychiatric pa-
tients were excluded because they are generally in crisis
when seen in the ED. Non–English-speaking patients
were excluded because we do not have the ability to
provide real-time translation. Additionally, a very
small minority of our patients are non–English speak-
ing. The study was performed in the ED of a tertiary
care teaching institution with 35,000 visits a year.
636 Graber et al.
Survey Content and Administration. Participants
were asked how many times a medical student should
have performed a procedure on others, after having
mastered the procedure on a simulator, before the
participant would allow the student to perform the
procedure on him or her. The procedures included
venipuncture, starting an intravenous (IV) line, sutur-
ing the face, suturing the arm, performing a lumbar
puncture, starting a central line, inserting a nasogas-
tric tube, intubation, and cardioversion. With the
exception of intubation and starting a central line,
these are procedures that a medical student could
expect to do in our ED. Intubation and starting a
central line were included because they represent
procedures that can be particularly hazardous if not
done properly.
Participants could respond in one of three ways:
they could specify a number of procedures they would
prefer a student do before performing it on them, they
could state that they would be willing to be a medical
student’s first patient, or they could state that they
would never allow a medical student to perform the
procedure on them. Participants were provided with
written descriptions of the procedures as well as a
picture of the procedure and the simulator. Complica-
tions of the procedures were described in easily
understandable terms if relevant. For nasogastric tube
placement, for example, the possibility and conse-
quences of nasogastric tube placement in the lung were
discussed. For intubation, the possibility and conse-
quences (including death) of esophageal intubation
were discussed. Remedies for the complications were
also mentioned (e.g., reintubation). The complications
of phlebotomy, IV line insertion, and suturing were
considered minimal and not listed. It was stipulated on
the questionnaire that the student had mastered the
skill on a simulator and that ‘‘a supervising staff
physician (’attending physician’) is always available
when you are being treated by a medical student.’’ We
did not state that the physician would be at the bedside
for every procedure. This reflects the reality in the ED
that an attending physician is not always present at the
bedside when every procedure is performed.
The remaining questions asked the patient’s age,
gender, and type of insurance. Ethnicity was excluded
because a small proportion of the population in our
state is minority and it is likely that there would not
be enough minority patients for a valid analysis.1
Questionnaires were administered by two of the
investigators (CW and LK), and pilot data were
collected by one of the investigators (MAG). Partic-
ipants could elect to either answer the questions on
paper or have the questions read to them and answer
verbally. The primary data collectors (CW and LK)
were medical students at the University of Iowa
Carver College of Medicine. Their participation in
this study, including helping with design, analysis,
and drafting of this report, was part of a summer
d PATIENTS AND MEDICAL STUDENT PROCEDURES IN THE ED
research internship that included other duties such
as participation in other projects, didactic lectures,
an introduction to research networks, etc. The spo-
radic nature of their available time, along with the
variable flow of patients in the ED, precluded a
randomized design. Neither data collector was in-
volved in the care of the patient while the patient was
in the ED. We specifically chose to not include the
investigators in the patients’ care to minimize any
possible accountability bias.
Measures. The main outcome measure was how
many times participants believed that a medical
student should have done a procedure on other
patients after mastering the procedure on a simulator
before the participant would allow the student to
perform it on him or her. These results were compared
with the results of a separate study1 performed one
year previously in our institution that asked the same
questions without the proviso that the skill had been
mastered on a simulator.
Data Analysis. Analysis was performed using SAS
software (version 8.0; SAS Institute Inc., Cary, NC).
Because the data were not normally distributed,
we used the Wilcoxon rank sum test to assess the
effect of age and gender on patient preferences. The
Kruskal–Wallis test was used to examine univariate
associations between the number of prior procedures
required by the patient and type of insurance (pri-
vate, Medicaid/Medicare/State Papers, and self-pay).
For purposes of reporting and statistical analysis,
responses were grouped into the following categories:
never allow, willing to be a student’s first patient, one
to four prior procedures, five to nine prior proce-
dures, and more than ten prior procedures. These
categories were derived in the previous study by
using the overall median of the quartiles from all
nine procedures after excluding the ‘‘never allow’’
responses. Differences between the two studies were
evaluated using the chi-square test.
A power analysis showed .85% power to detect a
small to moderate difference between the groups at an
a of 0.05. A measure of the differences between
groups is the overlap index.6 This index is zero if
patients in one group all require more previous
procedures than all patients in the other group, and
it is 1.00 if the median values of the two groups are
identical. Our study has the power for detecting an
overlap index of 0.6, which indicates that we have a
.85% chance of demonstrating a small to moderate
difference between the two groups.
RESULTS
A total of 151 of 185 patients approached had usable
data, for a participation rate of 82%. Twenty-five pa-
tients refused to participate, and another nine supplied
ACAD EMERG MED d July 2005, Vol. 12, No. 7 d www.aemj.org 637

TABLE 1. Demographics of Participants DISCUSSION

Gender Male
Female
Age (yr) 18–19
20–29
30–39
40–49
50–59
60 or older
Insurance Medicare/Medicaid/
State Papers*
Private
None
*State Papers is a program in the state of Iowa that allows
qualifying indigent patients to be seen at the University of Iowa
free of charge.
ambiguous data (such as marking two answers for all
questions). Participant demographics are shown in
Table 1. After simulator training, a high of 57% of
patients would be willing to be a student’s first
venipuncture. Fifty-eight percent of patients would
still never allow a student to perform a lumbar punc-
ture regardless of the degree of training.
When compared with no simulator training, pa-
tients required fewer ‘‘live’’ procedures after simula-
tor training for all procedures with the exception of
suturing (arm and face) and intubation. When com-
paring ‘‘never’’ and ‘‘allow first procedure’’ between
the two studies, more patients were willing to be a
student’s first procedure, and fewer would never
allow a student to perform a procedure for all
procedures except for lumbar puncture, placement
of a central line, and intubation (p , 0.05). There were
no statistical differences between any of the demo-
graphic groups in their responses. Participants who
had no insurance or government insurance were no
more likely to be accepting of medical student
procedures than were those with private insurance.
The results of this study (but not the data from the
prior study) are summarized in Table 2. A summary
of the results from the comparison between the two
studies1 is presented in Table 3.
48%
Simulator training can mitigate patients’ unwilling-
52%
6% ness to allow medical students to perform procedures
26% on them but does not eliminate it. While more patients
23% are willing to be a student’s first procedure (high of
19% 57% for venipuncture) and fewer would prefer never
13%
to have a student perform a procedure on them (high
18%
24% of 58% for lumbar puncture), many patients are still
not willing to be a student’s first procedure. In
55% addition, a significant percentage do not want medical
27% students to perform procedures on them regardless of
how much practice the student has had, including
mastering the skill on a simulator. Because patients
have a right to know the level of training of their
provider and have a choice of providers,7 this leaves
us with the dilemma of how to teach procedures to
medical students.
Patients have two mutually exclusive goals: they do
not want procedures done by a student, but they want
a health care system with well-trained physicians. The
burden of training physicians has traditionally fallen
on the poor and uninsured even though, as this study
and a previous study show, they feel no differently
about student procedures than do those with insur-
ance.1 Limiting the risk of training students to the
poor and uninsured violates the principle of justice,
which dictates that we should treat all patients in the
same way regardless of demographic or other factors.
While the burden of educating students will never be
equally distributed among all in society, we should
distribute the risk as equitably as possible. This may
mean that by agreeing to participate in the benefits of
the medical system, patients have to agree to give up
some autonomy and acquiesce to being treated by
properly supervised students. While unpopular, this
would avoid placing an undo burden on one segment
of society to the benefit of other segments.
LIMITATIONS
The first limitation is that patients did not have to
commit to having a procedure. However, this design

TABLE 2. Number of Procedures before a Patient Would Allow a Medical Student to Perform
Procedure on Self after Simulator Training
No. of Suture Suture Lumbar Central Line Nasogastric Tube
Procedures Venipuncture Start IV Arm Face Puncture Placement Placement Intubation Cardioversion
None* 57 48 44 25 13 11 34 19 40
1–4 16 15 17 14 7 9 15 11 13
5–9 10 14 13 11 8 9 16 7 7
$10 11 15 11 18 14 17 13 17 11
Nevery 10 9 15 33 58 54 21 44 30
No. of answered
questions 151 151 151 151 151 151 151 151 151
All values are expressed as percentages unless stated otherwise. Totals may not equal 100% because percentages were rounded.
*Would allow a medical student to perform this procedure for the first time.
yWould never allow a medical student to perform this procedure.
638 Graber et al.
TABLE 3. Number of Procedures after Simulator
Training versus No Simulator Training1 (Excluding
None and Willingness to Be First Procedure)
Procedure p-value Significant
Venipuncture ,0.001 Yes
Start IV ,0.001 Yes
Suture arm 0.12 No
Suture face 0.10 No
Lumbar puncture 0.011 Yes
Central line placement 0.03 Yes
Nasogastric tube placement 0.002 Yes
Intubation 0.13 No
Cardioversion 0.014 Yes
was chosen to limit accountability bias. A second
limitation is that at least part of the unwillingness of
participants to allow medical student procedures is a
‘‘symbol’’ reaction rather than a ‘‘signal’’ reaction.
Participants are reacting to the connotation of the term
‘‘medical student’’ (the symbol) and what it implies to
them (not trained, not competent, etc.) rather than
reacting to the skill of the particular student (the
signal) and the fact that the student had mastered the
procedure on a simulator.8,9 Semiotics is the study of
symbols, including words. Words are symbols that
represent an object or concept. An example of a
symbol reaction is the reaction to the term ‘‘drug
seeker.’’ This term (a symbol that is used to represent
a patient) evokes an emotional reaction and a set of
assumptions about the patient that may or may not be
real. In our case, the term ‘‘medical student’’ may
evoke a negative reaction in some people (e.g., young,
not fully trained). This does not invalidate our
conclusions, because this will be a factor in actual
medical student/patient interactions. A third limita-
tion is that framing bias likely played a role in
participants’ answers. A related concept is the ‘‘fram-
ing effect.’’ The framing effect is responsible for
different responses to a question depending on how
the question is phrased. For example, patients will
have a very different reaction to ‘‘Do you want to have
a diagnostic procedure to see if you have meningitis?’’
than to ‘‘Do you want to have a painful spinal tap to
see if you have meningitis?’’ If, instead of using the
term ‘‘medical student,’’ we had asked, ‘‘How many
procedures do you think a highly trained individual
who has gone through at least six years of training
after high school should do before they do it on you?’’
our results would likely have been different. Framing
bias is unavoidable. What is important is to use
terminology that is as consistent with the reality of
the situation as possible. Thus, we used the term
‘‘medical student,’’ which is used in common parlance
and represents how students are (or should be)
introduced.
Even though framing bias was a possible con-
founder, there is no way to remove this bias from
either theoretical or real-world patient/student inter-
d PATIENTS AND MEDICAL STUDENT PROCEDURES IN THE ED
actions. Thus, as with the symbol reaction, the pres-
ence of framing bias does not invalidate the findings
of this study. Finally, we used a convenience sample
rather than a random selection of patients. Thus, our
results may not be as robust as with randomized
patient selection. Given the nature of our ED, it would
have been difficult, if not impossible, to adhere to a
randomization scheme.
It can be argued that reality seems to contradict our
data; patients in the ED generally do allow medical
students to perform procedures on them. This does
not necessarily invalidate our findings. Patients may
not know that a medical student is involved in their
care, and students may not feel obligated to inform
patients that they are students.10–12 Patients also may
be unwilling to confront providers in order to assert
their right to choose the level of training of their
physician.
Not all studies have found the same results. In a
small study by Fletcher et al., ED patients were
willing to be a student’s first procedure. However,
that study included only a small range of relatively
noninvasive procedures.13 Additionally, unlike in our
study, the person administering the survey instru-
ment was involved in the participant’s care. This may
have introduced accountability bias into their study as
patients tried to avoid or minimize conflict with
someone involved in their care.
CONCLUSIONS
Simulator training helps to mitigate the unwillingness
of patients to allow medical students to perform
procedures on them. However, a significant percent-
age of patients still do not want to be a student’s first
procedure, and a significant proportion would prefer
not to have a medical student perform procedures on
them at all.
The authors thank Arthur Hartz, PhD, MD, for assistance with
statistical analysis.
References
1. Graber MA, Pierre J, Charlton M. Patient opinions and
attitudes toward medical student procedures in the emergency
department. Acad Emerg Med. 2003; 10:1329–33.
2. Gordon JA, Wilkerson WM, Schaffer DW, Armstrong EG.
‘‘Practicing’’ medicine without risk: students’ and educators’
responses to high-fidelity patient simulation. Acad Med. 2001;
76:469–72.
3. Morgan PJ, Cleave-Hogg BA, McIlroy J. Simulation
technology: a comparison of experimental and visual learning
for undergraduate medical students. Anesthesiology. 2002; 96:
6–10.
4. Lynch TG, Steele DJ, Johnnson JE, et al. Learning preferences,
computer attitudes and test performance with computer-aided
instruction. Am J Surg. 2001; 181:368–71.
5. Letterie GS. How virtual reality may enhance training in
obstetrics and gynecology. Am J Obstet Gynecol. 2002; 187:
S37–40.
ACAD EMERG MED d July 2005, Vol. 12, No. 7 d www.aemj.org
6. Hartz AJ. Overlap index: an alternative to sensitivity and
specificity in comparing the utility of a laboratory test. Arch
Pathol Lab Med. 1984; 108:65–7.
7. The Council on Ethical and Judicial Affairs of the American
Medical Association; Policy E-8 0.87 Medical Student
Involvement in Patient Care, adopted June 2001. Available at:
http://www.ama-assn.org/ama/pub/category/8491.html.
Accessed Apr 9, 2004.
8. Woodhead I. Neural Mechanism, Beast Consciousness and
Symbol Trading, 1995. Available at: http://www.cogsci.ed.
ac.uk/;ira/papers/paper.html. Accessed Jul 31, 2004.
9. Berman SI. Opening the closed mind: making assumptions,
jumping to conclusions. Etc: Et Cetera: A Review of General
Semantics. 2001; 58(4).
639
10. Silver-Isenstadt A, Ubel PA. Erosion in medical students’
attitudes about telling patients they are students. J Gen Intern
Med. 1999; 14:481–7.
11. Hemphill RR, Santen SA, Rountree CB, Szmit AR. Patients’
understanding of the roles of interns, residents, and attending
physicians in the emergency department. Acad Emerg Med.
1999; 6:339–44.
12. Santen SA, Hemphill RR, Prough EE, Perlowski AA. Do Patients
understand their physician’s level of training? A survey of
emergency department patients. Acad Med. 2004; 79:139–43.
13. Fletcher ND, McCloskey CB, Santen SA, Hemphill RR,
Isenhour JL. ‘‘Sorry, it’s my first time!’’ Will patients consent
to medical students learning procedures? [abstract]. Acad
Emerg Med. 2002; 9:495.