Use of a fundus photograph matching program in

imparting proficiency in ophthalmoscopy

Étienne Bénard-Séguin, BSc, Jason Kwok, MD,1 Walter Liao, MD, FRCSC,
Stephanie Baxter, MD, FRCSC
ABSTRACT ● RÉSUMÉ
Objective: To determine whether practice using an online fundus photograph program results in a long-term increase in proficiency
with direct ophthalmoscopy skills in medical students.
Design: This study was a prospective medical education trial. Students were enrolled to participate in a voluntary Objective
Structured Clinical Examination (OSCE) using 5 patients with ocular findings. Students who matched a minimum of 6 discs 16
months before the study were assigned to the intervention group and were compared with students who did not participate in the
exercise.
Participants: Forty-six second-year medical students at Queen’s University: 15 in the intervention group and 31 in the control group.
Methods: Students were evaluated using the Queen’s University Ophthalmoscopy OSCE Checklist (QUOOC). Students were asked
to calculate the cup-to-disc ratio, comment on disc margins, and whether there was any macular pathology. Students participated
in a summative OSCE as part of the curriculum in which all students attempted to match fundus photographs.
Results: Students in the intervention group performed significantly better on the QUOOC, with a mean score of 78.3% (± 4.2),
compared with the control, who had a mean score of 69.4% (± 4.2) (p ¼ 0.005). The intervention group was significantly more
accurate at matching optic nerve photographs, with 100% (15/15) of the students correctly identifying the correct optic nerve on
first attempt compared with 53.3% (16/30) in the control group (p ¼ 0.0014).
Conclusions: The use of an online peer fundus photograph program leads to a long-term increase in examination technique,
proficiency in ophthalmoscopy, and accuracy at matching optic nerve photographs.

Objet : Déterminer si le recours à un programme de photographie de fond d’œil en ligne se traduit par une augmentation à long
terme des aptitudes pour l’ophtalmoscopie directe chez des étudiants en médecine.
Nature : Étude prospective d’enseignement médical, menée auprès d’étudiants prenant part à un examen clinique objectif structuré
(ECOS) facultatif portant sur 5 cas d’atteinte oculaire. Les étudiants qui réussissaient à apparier au moins 6 disques 16 mois
avant l’étude ont été affectés au groupe évalué, qu’on a comparé au groupe témoin (formé d’étudiants n’ayant pas pris part à
l’exercice).
Participants : 46 étudiants en médecine de deuxième année de l’Université Queen’s, dont 15 dans le groupe évalué et 31 dans le
groupe témoin
Méthodes : Les étudiants ont été évalués à l’aide de la liste QUOOC (Queen’s University Ophthalmoscopy OSCE Checklist). Ils
devaient calculer le rapport cupule /disque, décrire les bords de la papille optique et indiquer s’il y avait ou non atteinte maculaire.
L’ECOS sommaire faisait partie du programme de base, où tous les étudiants devaient tenter d’apparier des photographies de
fond d’œil.
Résultats : Les étudiants du groupe évalué ont obtenu des scores significativement plus élevés au QUOOC, leur score moyen étant
de 78,3 % (± 4,2) comparativement à celui des patients témoins, soit 69,4 % (± 4,2; p = 0,005). L’appariement des photographies
du nerf optique a également été significativement plus exact dans le groupe évalué; 100 % (15/15) des étudiants ont repéré le bon
nerf optique du premier coup, comparativement à 53,3 % (16/30) de ceux du groupe témoin (p = 0,0014).
Conclusions : L’utilisation d’un programme de photographie de fond d’œil en ligne se traduit par une amélioration à long terme de
la technique d’examen, de l’aptitude pour l’ophthalmoscopie et de l’exactitude d’appariement de photographies du nerf optique.

Since the invention of the ophthalmoscope in 1851 by
Hermann von Helmholtz, direct ophthalmoscopy has
provided clinicians with a simple tool to assess the fundus
of patients.1 If performed correctly, this skill enables
physicians to obtain crucial information about ocular
and systemic health. It is essential that all medical students
have exposure to this skill to allow for correct and timely
diagnoses once in clinical practice. In 2006, the Interna-
tional Task Force of Ophthalmic Education of Medical
students labeled the use of the direct ophthalmoscope as a
basic skill necessary to the education of medical students
worldwide. Nonetheless, there is rarely more than
3 hours of dedicated clinical training in ophthalmoscopy
within the Canadian medical curriculum,2 which could
contribute to a growing cohort of health care professionals
with insufficient confidence in performing this skill,3
causing a decrease in its use.4,5
The proficiency of medical students with direct oph-
thalmoscopy is based on 2 pillars: examination techniques
and clinical diagnosis. In 1990, George Miller described a

& 2017 Canadian Ophthalmological Society.

Published by Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.jcjo.2017.11.022
ISSN 0008-4182/18

480 CAN J OPHTHALMOL — VOL. 53, NO. 5, OCTOBER 2018

 Online fundus photograph program for ophthalmoscopy proficiency—Bénard-Séguin et al.
framework for the clinical assessment that stipulated that
the most effective substitute for examining real patients
was through the simulation of encounters using stand-
ardised patients.6 The Objective Structured Clinical
Examination (OSCE) is a well-studied tool to evaluate
the clinical skills of medical students; however, its format
relies on standardised patients, which often do not have
physical findings.7 Therefore, proficiency in physical
examination techniques assessed by the OSCE may be
used as a surrogate measure of clinical proficiency. Haque
et al. developed the Queen’s University Ophthalmoscopy
OSCE Checklist (QUOOC) in 2012.8 This checklist,
which evaluates both ophthalmoscopy technique and
diagnostic accuracy, correlates to the ability of the learner
to correctly identify pathology.
In a previous study, Kwok et al. described the use of an
online peer fundus photograph matching software program
as a method of teaching ophthalmoscopy to medical
students.9 The learning activity consisted of medical
students examining their colleagues with an ophthalmo-
scope and subsequently matching what they visualised with
a series of online fundus photographs, one of which
corresponded to the examinee. This study concluded that
medical students who matched at least 6 fundus photo-
graphs using the software demonstrated an increase in their
confidence and ability to accurately match fundus photo-
graphs. Furthermore, students who matched at least 9 eyes
were significantly quicker at making matches than their
peers. Nevertheless, it has yet to be demonstrated that this
novel fundus photograph-matching program translates to
increased clinical proficiency or diagnostic accuracy.
Herein, we determine whether the use of an online peer
fundus photograph matching software program correlates
to increased clinical proficiency with ophthalmoscopy in
medical students. The study will describe the use of an
ocular pathology–based OSCE to determine whether
using an online disc and fundus photograph matching
program results in increased clinical proficiency with the
direct ophthalmoscope.
This study’s objectives are to assess whether self-
directed ophthalmoscopy practice using an online disc
matching program improves the participants’:
1. competence with the ophthalmoscope as measured with
a modified QUOOC,
2. ability to detect fundus pathology when present,
3. long-term retention of the fundus photograph match-
ing skill, and
4. confidence levels using the ophthalmoscope.
METHODS
This study was a prospective medical education trial.
This research protocol adhered to the tenets of the
Declaration of Helsinki and was approved by the Queen’s
Health Science Research Ethics Board committee.
CAN J OPHTHALMOL — VOL. 53, NO. 5, OCTOBER 2018
From April to May 2016, students were enrolled, on a
voluntary basis, to participate in an OSCE on May 14,
2016, whereby they would examine patients with a variety
of posterior segment findings. Given the voluntary nature
of enrolment, a convenience sample was used. Students
who enrolled in the OSCE and had matched at least
6 discs using the online peer fundus photograph software
program in the preceding academic year (16 months
before the study) were assigned to the intervention group.
This group was compared with students who had not
practiced with the peer fundus photograph software, or
who had successfully matched less than 6 discs using the
online peer fundus software the preceding year (Fig. 1).
Any participants who had additional exposure to ophthal-
moscopy other than having participated in the peer fundus
photograph matching exercise or the standard curriculum
(observerships, electives, etc.) were excluded. All students
had the opportunity to receive approximately 35 hours of
ophthalmology teaching from the standard curriculum.
The timing of the standard curriculum teaching was in the
2 weeks just before the study voluntary OSCE and 1 year
after the students had participated in the online peer
fundus photograph matching competition (Fig. 1).
On the day of the voluntary OSCE, the students
completed a questionnaire assessing any additional expe-
rience they had with ophthalmoscopy since their first year.
The questionnaire inquired about their level of confidence
performing ophthalmoscopy and used a 5-point Likert
scale, with the responses ranging from 1 (not confident at
all) to 5 (very confident). The questionnaire also contained
knowledge-based multiple-choice questions pertaining to
direct ophthalmoscopy. A 5-patient OSCE was used to
assess clinical competency in identifying pathology using
an ophthalmoscope. In each room, students could choose
to use the conventional direct ophthalmoscope or the
panoptic ophthalmoscope based on their preference. Each
student examined 1 dilated eye of 5 volunteer patients. For
each patient, students were evaluated using a modified
QUOOC by an ophthalmology resident. Students were
asked to calculate the cup-to-disc ratio, comment on the
blurring of the disc margins, and presence of macular
pathology for each patient.
Before the voluntary OSCE, a fourth- and a fifth-year
ophthalmology resident each independently performed a
fundus examination on each patient and documented their
findings on the previously stated parameters to obtain
consensus on the correct answers for these parameters. The
senior author assessed the patient, if required, to obtain
consensus to establish the correct answers for these
parameters for each standardised patient.
Each participant had a modified QUOOC (Appendix 1,
available online) score calculated based on his or her
performance on the OSCE for each of the 5 standardised
patient eyes examined. Modified QUOOC scores, as well as
scores for correct identification of macular pathology,
blurred disc margins, and cup-to-disc ratio, were recorded.
481
Online fundus photograph program for ophthalmoscopy proficiency—Bénard-Séguin et al.

the fundus matching skills for all study participants, and

the mean scores between groups were compared using the
t test for independent samples.
Due to a small sample size, the bootstrapping technique
was performed with a set repeated number of samples of
1000. All p-values were considered significant at the 5%
level.

RESULTS
A total of 46 second-year medical students were included
in our study. The intervention group consisted of 15
second-year medical students who had participated in the
peer fundus photograph competition the year before and
who had successfully matched at least 6 disc photos. The
control group consisted of 31 second-year students who had
not participated in the peer fundus photograph competition
or who had participated but who had successfully matched
less than 6 disc photographs. No students were excluded
from the final analyses. Results demonstrated no significant
difference in prior use of an ophthalmoscope (p ¼ 0.76),
considering a career in ophthalmology (p ¼ 0.99), com-
pletion of ophthalmology observerships (p ¼ 0.71), or the
use of an online module on how to perform ophthalmo-
scopy created by the senior author (p ¼ 0.99) between
groups. Students in the intervention group attended more
class time than the control group; however, this was not
statistically significant (p ¼ 0.54) and there was no
statistical difference in baseline knowledge between the
groups (p ¼ 0.44) based on the results of the knowledge-
based multiple-choice questions asked on the voluntary
OSCE day questionnaire (Table 1).

Voluntary OSCE: Posterior Segment Pathology

QUOOC Performance. Students in the intervention group

performed significantly better on the modified QUOOC,
with a mean score of 78.3% (± 4.2), compared with the
control group, who had a mean score of 69.44% (± 4.2)

Table 1—Participant demographics

Variables Intervention Control p

n 15 31
Fig. 1 — Flow diagram outlining the sequence of events of the Male 5 (33%) 15 (48%) 0.36
study. Considered a career in ophthalmology 1 (6.6%) 4 (13%) 0.53
Ophthalmology observership 4 (27%) 6 (19%) 0.45
Use of an ophthalmoscope (past 17 months)
The mean scores on the voluntary OSCE between both None 8 (54%) 18 (58%) 0.76
o1 hour 6 (40%) 13 (42%) 0.99
groups were compared using the t test for independent 42 hours 1 (6.6%) 0 0.33
samples. Each student’s confidence levels with the ophthal- Participation in ophthalmology research 2 (13%) 0 0.1
moscope were also analyzed for each group. % of didactic teaching attended
0% 0 3 (10%) 0.54
One week after the voluntary OSCE (May 20, 2016), o75% 2 (13%) 11 (35%) 0.17
all second-year medical students participated in a man- 75%–99% 2 (13%) 0 0.1
100% 11 (73%) 17 (55%) 0.34
datory summative OSCE. The online disc photograph
Completion of online ophthalmoscopy module 7 (47%) 14 (45%) 0.99
matching software was used for this exercise. Both the Pretest
number of attempts and the time to a correct match were Mean 68% 65% 0.44
Standard deviation 11% 14%
used from this OSCE to determine long-term retention of

482 CAN J OPHTHALMOL — VOL. 53, NO. 5, OCTOBER 2018

 Online fundus photograph program for ophthalmoscopy proficiency—Bénard-Séguin et al.
* accuracy of 24.7% (± 8.5) compared with 15.8%
Fig. 2 — Mean Queen’s University Ophthalmoscopy OSCE
Checklist (QUOOC) score and 95% confidence interval com-
paring the intervention group to the control group.
(p ¼ 0.005) (Fig. 2). The intervention group scored
significantly better at darkening the room (p ¼ 0.01), at
asking the patient to fixate in the distance (p ¼ 0.04), and
at finding the optic disc (p ¼ 0.04). Thirteen of the 15
(87%) participants in the intervention group were able to
find the optic disc for at least 4 of the 5 patients compared
with 19 of the 31 (61%) in the control group (OR ¼ 4.1
[0.8–21.5]).
Both the intervention and
Clinical Appreciation of Pathology.
the control group fared poorly when identifying ocular
pathology. When assessing for blurred disc margin, the
intervention group had a mean accuracy 56.0% (± 12.8)
compared with 51.7% (± 8.6) for the control group (p ¼
0.58). The intervention group was slightly more accurate
in identifying the presence of macular pathology, with a
mean of 65.3% (± 8.9), compared with 51.7% (± 10.1)
for the control group (p ¼ 0.06). When calculating the
* the number of hours allocated to teaching this skill has
Fig. 3 — Mean score and 95% confidence interval for 4 para-
meters comparing the intervention group (blue) to the control
group (green).
CAN J OPHTHALMOL — VOL. 53, NO. 5, OCTOBER 2018
cup-to-disc ratio, the intervention group had a mean
(± 5.0) for the control group (p ¼ 0.07). A subgroup
analysis of students (n ¼ 7) who had made at least
9 matches with the fundus photograph software program
the previous year revealed a mean accuracy of 32.9%
(± 12.7) when calculating the cup-to-disc ratio, which was
significantly greater compared with the control group’s
15.8% (± 5.0) (p ¼ 0.008). Both the intervention and the
control groups were unable to accurately diagnose con-
ditions in the eye, with a mean score of 22.6% (± 15.7)
for the intervention group and 26.8% (± 10.4) for the
control group (p ¼ 0.67) (Fig. 3).
Confidence. The mean confidence level when asked, “How
confident are you using an ophthalmoscope?” was 2.73 for
the intervention group versus 2.19 for the control group
(p ¼ 0.06). A subgroup analysis looking at students who
had matched at least 9 eyes in the competition the previous
year revealed a mean confidence level of 3.0 (p ¼ 0.02).
Summative OSCE: Optic Disc Matching
Performance on the summative OSCE revealed that the
intervention group was significantly more accurate at
matching optic nerve photographs with 100% (15/15) of
the students identifying the correct optic nerve on first
attempt versus 53.3% (16/30) in the control group (p ¼
0.001). The mean number of attempts to make a correct
match was 1 attempt in the intervention group versus
1.7 attempts in the control group (p ¼ 0.01) (Fig. 4). The
average time to make a successful match was 265.0 seconds
(± 61.6) for the intervention group and 403.2 seconds (±
100.4) for the control group (p ¼ 0.17) (Fig. 5).
DISCUSSION
The use of the ophthalmoscope has slowly declined as
Fig. 4 — Box plot analysis comparing the number of attempts
to make a successful match for the intervention group
compared with the control group.
483
Online fundus photograph program for ophthalmoscopy proficiency—Bénard-Séguin et al.
Fig. 5 — Box plot analysis comparing the time to a successful
match in seconds (sec) for the intervention group compared
with the control group.
diminished,2 leading to a decrease in confidence in
performing fundus examination.3 The online peer fundus
photograph matching software is a novel learning tool that
provides students with an engaging and interactive learn-
ing platform, allowing them to practice independently in
order to hone their skills. Our results demonstrate that
practice with the online peer fundus photograph matching
software leads to a long-term improvement in examination
technique as measured by the modified QUOOC and a
long-term increase in self-reported confidence. Further-
more, these students demonstrated increased long-term
retention of their ophthalmoscopy skill and were signifi-
cantly more accurate at matching optic nerve photographs
than their counterparts. Students who practiced with the
online peer fundus photograph matching software took,
on average, 2 minutes less to make a successful match than
their counterparts. Although this metric was not statisti-
cally significant, reducing patient’s exposure to bright
lights is a significant clinical advantage. Despite minimal
teaching in the curriculum on assessment of retinal
pathology and cup-to-disc ratio, students who correctly
matched 9 or more eyes the preceding year were signifi-
cantly more accurate at estimating cup-to-disc ratio.
Although students who practiced with the online peer
fundus photograph matching software demonstrate supe-
rior examination technique compared with their counter-
parts, their diagnostic accuracy remains poor.
In light of these findings, we have changed the Queen’s
University undergraduate medical curriculum for ophthal-
moscopy. First-year students are now required to examine
and match a minimum of 10 classmates’ discs using the
online fundus photograph matching software program.
This allows for the teaching and practice of ophthalmo-
scopy to be removed from the second-year curriculum, so
that the focus of teaching during that time can be on
retinal and optic nerve pathologies, recognition of abnor-
malities, and the development of differential diagnoses. To
build on these concepts, we wonder whether students
484 CAN J OPHTHALMOL — VOL. 53, NO. 5, OCTOBER 2018
might benefit from logging ophthalmoscopy examinations
on patients during clerkship rotations to encourage
ophthalmoscopy in the clinical setting and to determine
whether our curricular changes result in retention of
ophthalmoscopy skills even longer term.
Our study initially sought to assess whether medical
students who had used the online peer fundus photograph
matching software would have greater clinical competency
and accuracy of diagnosis than those who had not. Despite
being in a competency-based medical education era, there
is no guidance in the literature outlining reasonable
expectations for ophthalmoscopy for the various levels of
medical trainees. Studies such as that by Chatziralli et al.
aim to outline expectations for a general practitioner.10
The study states that general practitioners should be
comfortable identifying optic disc edema, vitreous hemor-
rhage, and ocular changes related to chronic processes such
as hypertension and diabetes. However, based on our
study, it was clear that second-year medical students in our
current curriculum were not at that level of competence.
They were, at most, able to identify the presence of some
pathologic processes but were unable to provide a differ-
ential diagnosis for these findings, let alone a final
diagnosis. It is with this expectation in mind that we call
for the defining of stage-appropriate competence for
ophthalmoscopy in medical training, which will guide
leaders in their development of undergraduate and post-
graduate curricula and assist medical trainees in their
learning.
This study had certain limitations. This was a voluntary
study, and a sample of convenience was used. This may
have resulted in the study being underpowered to detect
certain differences between groups due to a small sample
size. Furthermore, this study was not randomised or
blinded. The voluntary nature of this study may have
introduced a selection bias as students who practiced with
the online fundus photograph program may have been
inherently more interested in ophthalmoscopy and could
have been motivated to use the software as they had a
greater natural ability to pick up the skill compared
with their counterparts. We accounted for differences
between groups to the best of our ability; however, it is
possible that certain underlying confounding factors
remain.
In summary, this study demonstrates that practice using
an online peer fundus photograph matching software
program leads to long-term improvement in student’s
abilities with the ophthalmoscope, as well as an increase in
self-reported confidence. Future studies should focus on
the development of stage-of-training–appropriate compe-
tency-based guidelines for ophthalmoscopy that will help
guide the development of undergraduate curricula and
novel teaching methods to increase students’ and practi-
tioners’ abilities to accurately identify retinal and optic disc
pathology so that they can use this information for clinical
decision making.
 Online fundus photograph program for ophthalmoscopy proficiency—Bénard-Séguin et al.
APPENDIX
Supplementary data
Supplementary data associated with this article can be
found in the online version at https://doi.org/10.1016/j.
jcjo.2017.11.022.
REFERENCES
1. Keeler CR. The ophthalmoscope in the lifetime of Hermann von
Helmholtz. Arch Ophthalmol. 2002;120:194-201.
2. Bellan L. Ophthalmology undergraduate education in Canada. Can J
Ophthalmol. 1998;33:3-7.
3. Schulz C, Hodgkins P. Factors associated with confidence in
fundoscopy. Clin Teach. 2014;11:431-5.
4. Roberts E, Morgan R, King D, Clerkin L. Funduscopy: a forgotten
art? Postgrad Med J. 1999;75:282-4.
5. Dalay S, Umar F, Saeed S. Fundoscopy: a reflection upon medical
training? Clin Teach. 2013;10:103-6.
6. Miller GE. The assessment of clinical skills/competence/performance
Acad Med. 1990;65:S63-7.
7. Adamo G. Simulated and standardized patients in OSCEs: achieve-
ments and challenges 1992–2003. Med Teach. 2003;25:262-70.
8. Haque R, Abouammoh MA, Sharma S. Validation of the Queen’s
University Ophthalmoscopy Objective Structured Clinical Exami-
nation Checklist to predict direct ophthalmoscopy proficiency. Can J
Ophthalmol. 2012;47:484-8.
9. Kwok J, Liao W, Baxter S. Evaluation of an online peer fundus
photograph matching program in teaching direct ophthalmoscopy to
medical students. Can J Ophthalmol. 2017;52:441-6.
CAN J OPHTHALMOL — VOL. 53, NO. 5, OCTOBER 2018
10. Chatziralli IP, Kanonidou ED, Keryttopoulos P, Dimitriadis P,
Papazisis LE. The value of fundoscopy in general practice. Open
Ophthalmol J. 2012;6:4-5.
Footnotes and Disclosure:
The authors have no proprietary or commercial interest in any
materials discussed in this article.
We would like to acknowledge Vic Sahai for his help with the
statistical analysis.
From Queen’s University, Kingston, Ont
1
Current/Present address: University of Toronto, Ontario,
Canada.
Presented at the Canadian Conference on Medical Education,
Winnipeg, Manitoba, Canada, 2017/05/01.
Originally received Aug. 22, 2017. Final revision Nov. 15, 2017.
Accepted Nov. 27, 2017.
Correspondence to Stephanie Baxter, MD, 166 Brock St.,
Kingston, Ont. K7L 5G2; baxters@queensu.ca
485