Keywords
educational tools - medical student education - direct ophthalmoscopy - eye model
- clinical skills
Proper use of the direct ophthalmoscope to view the fundus is an important skill not
only for ophthalmologists, but also for all primary care physicians.[1]
[2] However, current medical school curriculums may not permit satisfactory training
of medical students in this critically important skill.[1]
[3]
[4] Literature suggests that medical students are failing to perform or even attempting
to perform adequate eye examinations, including proper direct ophthalmoscopy.[1]
[4] The need for additional attention to ophthalmology in the medical curriculum has
been made clear.[1]
[4]
[5] Other studies have shown that increased practice with direct ophthalmoscopy can
increase self-confidence and ability in its use.[4]
[6]
Eye models have been previously developed as an educational tool for direct ophthalmoscopy.[4]
[7]
[8]
[9]
[10] The models described by Dodaro and Lippa are similar in that they use a Styrofoam
head (The Dow Chemical Company, Midland, Michigan), a slide holder to house retinal
images, a model eye, and a plastic face cover.[4]
[7] McCarthy used the “EYE Exam Simulator” by Kyoto Kagaku Co., Ltd.,[11] which retails at $1,300 but reported low user satisfaction.[10] Bradley and Hoeg designed eye models using painted table tennis balls and canisters
with photos, neither of which was mounted into a model head.[9] The model we describe is significantly less expensive than the Kyoto Kagaku model,
more easily assembled than those of Dorado, Lippa, and McCarthy, and has the added
advantage of functioning as a stand-alone unit or being mounted in a Styrofoam head
to simulate a patient encounter. While several studies evaluated the effectiveness
of using eye models as an educational tool, none have done so using a prospective
cohort with a control group.[4]
[8]
[9]
The primary objectives of this study were to (1) create a new artificial eye model
that is simple and inexpensive, (2) develop an objective evaluation tool assessing
medical students' clinical skills in direct ophthalmoscopy, and (3) use this objective
evaluation to determine if practice with the eye model translates into improved clinical
examination skills.
A secondary objective was to determine whether access to the model impacts the frequency
with which students practice direct ophthalmoscopy on either the model or human subjects.
Methods
Eye Model
The FAK-I model was designed to approximate certain anatomic properties of the human
eye, using low-cost and easily assembled materials. An opaque, plastic bottle cap
measuring 24 mm in depth was used to replicate the anterior-posterior dimension of
the human eye. High-resolution, digital retinal images were created from montage color
fundus photography taken by the Topcon 50ix camera and processed by Ophthavision 3.0
(Escalon Medical Corp., Ardmore, Pennsylvania), with the optic nerve scaled to a vertical
diameter of 1.5 mm ([Fig. 1]). These photographs, printed on matte paper to minimize glare, were cut to size
and placed into the base of the bottle cap ([Fig. 2], inset). Black card stock with a central opening was used to cover the cap, creating
an “iris” and “pupil.” Superior poles of eye models were marked with an L or R for
placement orientation. The FAK-I model may be used as a stand-alone unit or mounted
into a Styrofoam head (available at beauty supply stores) with orbit cutouts, simulating
the facial anatomy and practical challenges of patient examination ([Fig. 2]).
Fig. 1 Sample image of fundus photograph.
Fig. 2 Assembled FAK-I with fundus photo inside bottle cap and black card stock cover (inset).
FAK-I mounted in Styrofoam head.
Student Participants
This study was approved by the Northwestern University IRB office. Students were recruited
from a group of 40 third-year Northwestern University Feinberg School of Medicine
students rotating through their surgical clerkship at the time of the study. As first-year
medical students, all had previously participated in a 30-minute lecture on eye examination
techniques followed by small-group, 90-minute practical training that included direct
ophthalmoscopic examination of their peers and a standardized patient. No objective
measurement of students' ability to use the direct ophthalmoscope was made during
this session, and no other educational interventions for research purposes had been
performed prior to their surgical clerkship.
Eight of the 40 students had been previously assigned an elective 2-week ophthalmology
rotation during their surgical clerkship, and were excluded from the study. For the
remaining 32 students, enrollment was voluntary and did not affect their surgical
clerkship evaluations.
Prior to commencement of the study, all 40 students were briefed on its nature and
purpose, and all participated in a 15-minute hands-on direct ophthalmoscopy training
session with three instructors. Of the 32 potential enrollees, 19 students volunteered
to participate in the study, 12 of whom were randomly selected. These 12 students
were further divided by randomly assigning de-identified participant forms into the
control arm (no FAK-I model intervention) and study arm (FAK-I model intervention)
with 6 students in each group. Students enrolled into the study arm received a FAK-I
set consisting of a Styrofoam head mount and five FAK-I eye models depicting different
posterior pole findings. The investigators did not provide students with direct ophthalmoscopes,
but they were told where direct ophthalmoscopes were available throughout the medical
school campus.
The student intervention period was 45 days in length, and concluded with the objective
evaluation. During this period, all 40 students attended a required ophthalmic practical
skills session, consisting of two hours of lectures followed by small-group sessions
designed to practice complete eye examinations. Study participants received no additional
training, and investigator contact with student enrollees was limited to email communication
regarding logistics for the final evaluation.
Patient Participants
Patients were recruited on a voluntary basis from the Northwestern Medical Faculty
Foundation Ophthalmology Clinic. Inclusion criteria were patients over the age of
18 with a clear view of the posterior pole. Pregnant patients and patients unable
to tolerate 12 consecutive fundus exams for any physical, emotional, or psychological
reason were excluded.
Evaluation Tool
To objectively evaluate students' ability to examine a patient with the direct ophthalmoscope
in a clinical setting, an evaluation tool similar to that used in United States Medical
Licensing Examination Step 2 Clinical Skills was designed. Students enrolled in this
study had previous exposure to similar methods during their medical school training.
After 12 patient subjects were recruited, one of two investigators (RM or AB) performed
a bilateral dilated exam, selecting one eye per patient to be photographed for later
use in the objective evaluation. On the day of the evaluation, patients had only the
predesignated eye dilated using tropicamide 1% and phenylephrine 2.5%. Each patient's
fundus photograph was printed among five other unrelated photographs selected from
the Northwestern departmental photographic database. Students evaluated the dilated
eye of each patient in successive 90-second encounters. For each patient encounter,
students were instructed to introduce themselves, wash their hands, examine the posterior
pole using direct ophthalmoscopy, select the matching fundus photo from six possible
photographic choices, and record their selection on their answer sheet. No patient
information was provided to student examiners and no other examination techniques
were permitted or performed. Additionally, students were not asked to diagnose ocular
pathology and no photographs were labeled.
Following the session, students completed an anonymous questionnaire inquiring about
prior experience and use of the direct ophthalmoscope, estimated time spent using
the FAK-I models over the course of the study, estimated time spent practicing direct
ophthalmoscopy on a patient or peer over the course of the study, ownership of a direct
ophthalmoscope, intended field of postgraduate training, and expected use of the direct
ophthalmoscope as a practicing physician.
Students were scored one point for the correct identification of a patient's fundus
photograph, and percent correct answers were then calculated.
Statistical Analysis
The study was powered to detect a 23% difference in the ability of the study cohort
to correctly identify patients' fundus photos from their exam when compared with the
control cohort. A total of 144 assessments were made by 12 students (6 study cohort
students and 6 control cohort students) examining 12 patients. The decision to power
the study as such was in part affected by the number of dilated examinations the investigators
felt a typical patient could tolerate within a relatively short period of time. Fisher
exact test was used to compare the demographic characteristics of the study and control
groups as well as the secondary study outcome. Assessment of the primary outcome was
performed using a generalized linear model.
Results
Student Demographics
Of the students in the study, all 12 (100%) reported prior use of direct ophthalmoscopy
during previous clinical rotations. In both cohorts, five (83%) of six students anticipated
using direct ophthalmoscopy in the future. There was great variability among both
cohorts regarding their future medical specialty choice, with no students citing a
traditional surgical field. More students owned a direct ophthalmoscope in the study
group (two, 33%) compared with the control group (zero, 0%), but more students had
prior exposure to ophthalmology in the control group (two, 33%) compared with the
study group (zero, 0%). The differences between the two cohorts with regard to both
ownership and prior exposure did not reach statistical significance (p > 0.05) ([Table 1]).
Table 1
Demographic data
|
Control (n [%])
|
Study (n [%])
|
p value
|
|
Prior training
|
3 (50)
|
4 (67)
|
NS[a]
|
|
DO ownership
|
0 (0)
|
2 (33)
|
NS
|
|
DO use in prior rotation
|
6 (100)
|
6 (100)
|
NS
|
|
Future medical specialty
|
|
|
|
|
• Internal medicine and subspecialty
|
2 (33)
|
3 (50)
|
|
|
• Pediatrics
|
0 (0)
|
1 (17)
|
|
|
• Family medicine
|
1 (17)
|
1 (17)
|
|
|
• Emergency medicine
|
1 (17)
|
0 (0)
|
|
|
• Anesthesiology
|
1 (17)
|
1 (17)
|
|
|
• Unknown
|
1 (17)
|
0 (0)
|
|
|
Anticipated DO use
|
5 (83)
|
5 (83)
|
NS
|
|
Prior exposure to ophthalmology
|
2 (33)
|
0 (0)
|
NS
|
Abbreviation: DO, direct ophthalmoscopy.
a NS indicates p > 0.05 between control and study groups.
Primary Outcomes
All students from the study (FAK-I) cohort completed all 12 patient encounters as
planned, while only 5 of the control cohort were able to finish all encounters. One
student from the control group completed 11 patient encounters, but was unable to
complete the 12th due to malfunction of the direct ophthalmoscope. An intention-to-treat
approach was taken in the statistical analysis of the number of correct responses,
scoring the incomplete encounter as a correct response ([Table 2]).
Table 2
Primary outcome
|
Control (n [%])
|
Study (n [%])
|
p value
|
|
Correct responses,[a] intention to treat
|
29 (40)
|
32 (44)
|
0.61
|
a Numbers shown reflect total number of correct responses in the group out of 72 possible
responses.
The average number of correct answers from the control cohort was 5.0 compared with
5.3 (range 1–8) in the study cohort. There was a slight trend toward higher percentage
of correct responses in the study group compared with the control group (44 vs 40%,
respectively); however, the difference between the groups did not reach statistical
significance (p = 0.61).
Secondary Outcomes
More subjects in the study cohort reported independent practice using the direct ophthalmoscope
(three, 50%) compared with the control cohort (zero, 0%); however, these numbers did
not reach statistical significance (p = 0.1). Students in the study cohort reported spending more time practicing on the
FAK-I model (three, 50%) than on peers or patients (one, 16.7%). No student reported
spending more than 90 minutes (range 15–90 minutes) practicing on the FAK-I or on
peers or patients over the course of the study ([Table 3]).
Table 3
Secondary outcome
|
Control (n [%])
|
Study (n [%])
|
p value
|
|
Any independent practice during surgical rotation
|
0 (0)
|
3 (50)
|
>0.1
|
|
Practice on peers or patients
|
0 (0)
|
1 (17)
|
|
|
Practice on FAK-I
|
N/A
|
3 (50)
|
|
Discussion
We have employed two innovative methods to achieve our main objectives. First, we
created a new artificial eye model, FAK-I, which is simple and inexpensive to construct.
Using a bottle cap, printed fundus images, card stock, and a Styrofoam head, a training
set may be assembled for less than $20 (plus photographic costs). It also provides
anatomically realistic attributes for simulated fundus examination in the settings
of both individual practice and group instruction. Second, we employed a novel objective
evaluation tool to assess medical students' clinical skills in direct ophthalmoscopy
by using the same 12 patients for all subject testing within a controlled setting.
This study looked at the effects of providing a simple-to-use tool, the FAK-I, to
medical students in a clinical rotation, independent of any other intervention. Important
conclusions can be drawn from this study despite the fact that the results did not
reach statistical significance. First, the study suggests that the majority of medical
students anticipate using direct ophthalmoscopy in their future practice. Second,
simply providing the resources for independent practice of direct ophthalmoscopy on
an eye model does not necessarily translate into actual use of the model. Only 50%
of the students in the study cohort practiced with FAK-I without supervision. It may
be that perceived limitations in time and the inconvenience of obtaining a direct
ophthalmoscope with which to practice were barriers to model use. Although a majority
of students recognized the need to perform direct ophthalmoscopy in the future, varying
levels of motivation may be another contributing factor to the lack of practice. Third,
more students from the study group reported practicing on the FAK-I as opposed to
peers or patients, suggesting that students may find it more convenient and comfortable
to practice on a model. Finally, results also show a positive trend in the frequency
of independent practice with the direct ophthalmoscope among the students of the study
cohort in comparison to the control cohort. In addition, the study group also practiced
more on peers and patients than the control group, suggesting that availability of
the FAK-I model may have stimulated their interest to practice on human subjects.
However, this study was underpowered to detect a statistical significance for differences
on these points.
In this small pilot study, simply providing the FAK-I model to students without further
intervention did not significantly improve competency with direct ophthalmoscopy in
the objective evaluation. An investigation with a larger and more comprehensive study
population should be conducted before any further conclusions are made. While this
study was limited to students in their third-year surgical rotation, the FAK-I model
could certainly be an effective tool for medical students to practice their eye examination
skills at any point in their training.
This study had several limitations. The study size was limited due to what was considered
to be a tolerable number of dilated exams for a patient to undergo in a short period
of time. As a result, the study was underpowered to detect a small increase, if present,
in the ability of the study cohort to correctly identify patient fundus photography
when compared with the control cohort.
The performance of the students on the objective evaluation also may have been affected
by several confounding variables. Difficulty level of each set of photographs based
on similarity or dissimilarity between the correct and incorrect photographs is difficult
to standardize. Also, differences in the patient interactions, such as patient compliance
or variability of the equipment in the examination rooms, may have positively or negatively
impacted the students' performance during the objective examination.
While the findings of this study did not reach statistical significance, the trends
toward improved performance on the objective examination and increased likelihood
of independent practice in the study population support the utility of the FAK-I model
as a tool for practicing direct ophthalmoscopy.
