EBioMedicine 9 (2016) 140–147

Contents lists available at ScienceDirect

EBioMedicine

journal homepage: www.ebiomedicine.com

Research Paper

Development of an E-learning System for the Endoscopic Diagnosis of

Early Gastric Cancer: An International Multicenter Randomized
Controlled Trial
K. Yao a,⁎, N. Uedo b, M. Muto c, H. Ishikawa d, H.J. Cardona e, E.C. Castro Filho f, R. Pittayanon g, C. Olano h, F. Yao i,
A. Parra-Blanco j, S.H. Ho k, A.G. Avendano l, A. Piscoya m, E. Fedorov n, A.P. Bialek o, A. Mitrakov p, L. Caro q,
C. Gonen r, S. Dolwani s, A. Farca t, L.F. Cuaresma u, J.J. Bonilla v, W. Kasetsermwiriya w, K. Ragunath x, S.E. Kim y,
M. Marini z, H. Li aa, D.G. Cimmino ab, M.M. Piskorz ac, F. Iacopini ad, J.B. So ae, K. Yamazaki af, G.H. Kim ag,
T.L. Ang ah, D.M. Milhomem-Cardoso ai, C.A. Waldbaum aj, W.A. Piedra Carvajal ak, C.M. Hayward al, R. Singh am,
R. Banerjee an, G.K. Anagnostopoulos ao, Y. Takahashi ap
a
Fukuoka University Chikushi Hospital, Chikushino, Japan
b
Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka, Japan
c
Kyoto University, Kyoto, Japan
d
Kyoto Prefectural University of Medicine, Kyoto, Japan
e
Simon Bolivar Hospital, Bogota, Colombia
f
Rio de Janeiro State University, Rio de Janeiro, Brazil
g
King Chulalongkorn Memorial Hospital, The Thai Red Cross and Chulalongkorn University, Bangkok, Thailand
h
Universidad de la República, Montevideo, Uruguay
i
Peking Union Medical College Hospital, Beijing, China
j
School of Medicine, Pontificia Universidad Catolica De Chile, Santiago, Chile
k
University of Malaya, Kuala Lumpur, Malaysia
l
Hospital Rafael Angel Calderon Guardia, CCSS, San Jose, Costa Rica
m
Universidad Peruana de Ciencias Aplicadas, Lima, Peru
n
Russia National Medical University, Moscow University Hospital, N31, Moscow, Russian Federation
o
Pomeranian Medical University, Szczecin, Poland
p
Nizhniy Novgorod Cancer Hospital, Nizhniy Novgorod, Russian Federation
q
GEDyt Gastroenterologia diagnostica y tratamiento Inst afiliafa a la UBA Buenos Aires, Argentina
r
Haydarpasa Numune Training and Research Hospital, Istanbul, Turkey
s
Division of Population Medicine, School of Medicine, Cardiff University, Cardiff, United Kingdom
t
The American British Cowdray Medical Center. Mexico City, Mexico
u
Hospital Nacional Adolfo Guevara Velasco, Cusco, Peru
v
i-gastro/Hospital Central de la Fuerza Aerea del Peru, Lima, Peru
w
Faculty of Medicine Vajira Hospital, Navamindradhiraj University, Bangkok, Thailand
x
NIHR Nottingham Digestive Disease Biomedical Research Unit, Queens Medical Centre, Nottingham University Hospital, Nottingham, United Kingdom
y
Kosin University College of Medicine, Busan, Republic of Korea
z
Gastroenterology and Operative Endoscopy Unit, Siena University Hospital, Siena, Italy
aa
Sichuan Provincial People's Hospital, Sichuan, Academy of Medical Sciences, Chengdu, China
ab
Hospital Aleman, Buenos Aires, Argentina
ac
Hospital de Clinicas Jose de San Martin, Buenos Aires, Argentina
ad
Ospedale S. Giuseppe, ASL Roma 6, Albano L, Rome, Italy
ae
National University of Singapore, Singapore, Singapore
af
University of Sao Paulo, Sao Paulo, Brazil
ag
Pusan National University School of Medicine, Busan, Republic of Korea
ah
Changi General Hospital, Singapore, Singapore
ai
General Hospital of Goiania, Goiania, Brazil
aj
Hospital de Clinicas Jose de San Martin, Buenos Aires, Argentina
ak
Hospital Mexico, San Jose, Costa Rica
al
Derriford Hospital, Plymouth, United Kingdom
am
Lyell McEwin Hospital & University of Adelaide, Adelaide, Australia
an
Asian Institute of Gastroenterology, Hyderabad, India
ao
Mitera General Hospital, Athens, Greece
ap
FAST Inc., Tokyo, Japan

⁎ Corresponding author at: Department of Endoscopy, Fukuoka University Chikushi Hospital, 1-1-1 Zokumyouin, Chikushino City, Fukuoka, 818-8502, Japan.
E-mail address: [email protected] (K. Yao).

http://dx.doi.org/10.1016/j.ebiom.2016.05.016
2352-3964/© 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
 K. Yao et al. / EBioMedicine 9 (2016) 140–147 141

a r t i c l e i n f o a b s t r a c t

Article history: Background: In many countries, gastric cancer is not diagnosed until an advanced stage. An Internet-based e-
Received 4 February 2016 learning system to improve the ability of endoscopists to diagnose gastric cancer at an early stage was developed
Received in revised form 15 May 2016 and was evaluated for its effectiveness.
Accepted 16 May 2016 Methods: The study was designed as a randomized controlled trial. After receiving a pre-test, participants were
Available online 17 May 2016
randomly allocated to either an e-learning or non-e-learning group. Only those in the e-learning group gained
access to the e-learning system. Two months after the pre-test, both groups received a post-test. The primary
Keywords:
Endoscopic diagnosis
endpoint was the difference between the two groups regarding the rate of improvement of their test results.
Gastric cancer Findings: 515 endoscopists from 35 countries were assessed for eligibility, and 332 were enrolled in the study,
E-learning with 166 allocated to each group. Of these, 151 participants in the e-learning group and 144 in the non-e-
International multicenter randomized con- learning group were included in the analysis. The mean improvement rate (standard deviation) in the e-
trolled trial learning and non-e-learning groups was 1·24 (0·26) and 1·00 (0·16), respectively (P b 0·001).
Interpretation: This global study clearly demonstrated the efficacy of an e-learning system to expand knowledge
and provide invaluable experience regarding the endoscopic detection of early gastric cancer (R000012039).
© 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).

1. Introduction Standards of Reporting Trials (CONSORT) statement (Schulz et al., 2010)

and the Declaration of Helsinki (World Medical Association Declaration
Almost one million new cases of gastric cancer were estimated to of Helsinki, 2013).
have occurred throughout the world in 2012 (952,000 cases, 6.8% of The study protocol was approved by the institutional review board
total new cancer cases), making it the fifth most common malignancy of Fukuoka University Chikushi Hospital, Japan (R12-060,dated Febru-
after cancers of the lung, breast, colorectum and prostate (GLOBOCAN, ary 6, 2013), and was registered as Clinical Trial No. UMIN
2012). Gastric cancer is the third leading cause of death from cancer R000012039. Written informed consent was obtained from all partici-
among both sexes worldwide (723,000 deaths, 8.8% of total cancer pating endoscopists.
deaths). Most patients with gastric cancer are diagnosed at an advanced
stage, with an overall 5-year survival rate of approximately 28% 2.2. Participants
(American Cancer Society. Cancer facts and figures, 2014). Early detec-
tion is the key to improving the survival of gastric cancer patients We recruited endoscopists from 35 countries around the world be-
(Leung et al., 2008). Upper gastrointestinal endoscopy is a widely ac- tween March 2013 and November 2013. Inclusion criteria were 1) abil-
cepted procedure for early detection of gastric cancer. However, in ity of the web browser on a participant's computer to display and to
many countries, endoscopists have limited opportunities to acquire operate sample contents of the e-test and e-learning system; 2) suffi-
the techniques, knowledge and experience which are imperative for cient English skills to understand the materials of the e-test and e-
the endoscopic detection of early gastric cancer (EGC) when only subtle learning system; 3) provision of a fully completed pre-study question-
mucosal morphology is apparent (Veitch et al., 2015). In contrast, naire sheet; and 4) provision of a signed consent form for participation
endoscopists in Japan have more such opportunities thereby enabling in this study. Medical practitioners who did not complete the pre-test or
them to detect subtle lesions that suggest EGC. whose pre-test score was 80% or more were excluded from the study
In order to overcome these problems, we have developed an because the e-learning system was aimed at providing training for
Internet-based e-learning system which is in English, and which is avail- those who had not previously received adequate training in the endo-
able anywhere in the world, and at any time of the day, so that clinicians scopic diagnosis of EGC. A specific username and password were
worldwide can learn how to detect EGC (Yao, 2013). Recently, advanced assigned to each participant to enable e-test results to be collected via
imaging endoscopy techniques have become a topic for discussion in the Internet and to control access to the e-learning system during the
various academic meetings or publications (American Gastroenterolog- e-learning period.
ical Association (AGA), 2008). Nevertheless, white-light endoscopy is
still the most common practice throughout the world. This e-learning
3. Interventions
system is therefore dedicated to teaching diagnosis using white-light
conventional endoscopy alone (Yao, 2012).
3.1. Pre-test Evaluation
We hypothesized that if endoscopists could acquire the detailed
“knowledge, techniques and experience” essential for the early detec-
Participating endoscopists undertook a pre-test via the Internet be-
tion of gastric cancer through this e-learning system, then the detection
tween February 1st 2014 and February 28th 2014. The participants
rate of early-stage gastric cancer would increase throughout the world
viewed 40 sets of endoscopic images on their web browser. Each set
(Veitch et al., 2015). Accordingly, we investigated the feasibility of this
of endoscopic images contained 18 to 24 images that had been system-
e-learning system to improve the ability of endoscopic detection of
atically taken during screening endoscopy to record the whole gastric
EGC among endoscopists outside Japan.
mucosa in a single patient, according to systematic screening protocol
for the stomach (Veitch et al., 2015; Yao, 2013). All endoscopy images
2. Materials and Methods were acquired using a high-definition electronic endoscopy system
(EVIS Lucera Spectrum System, Olympus Co. Ltd., Tokyo, Japan) and
2.1. Study Design high-definition upper gastrointestinal endoscopes (GIF-H260; GIF-
H290, Olympus Co. Ltd.). The 40 patients consisted of 20 patients with
Global e-Endo Study Team (GEST) was organized to develop an e- EGC and 20 patients with non-cancerous findings. Each lesion was his-
learning system for improving the detection rate of EGC among topathologically confirmed as either cancer or non-cancer. The histo-
endoscopists worldwide. This study was conducted as an international, pathologic diagnosis was based on the revised Vienna classification
randomized, controlled trial to evaluate the effectiveness of the e- (Schlemper et al., 2000); C1 (negative for neoplasia), C2 (indefinite for
learning system. The study was conducted in line with the Consolidated neoplasia) and C3 (mucosal low-grade neoplasia) were diagnosed as
142 K. Yao et al. / EBioMedicine 9 (2016) 140–147

non-cancer, whereas C4 (mucosal high-grade neoplasia) and C5 (sub- limited to the submucosal layer (Japanese Gastric Cancer Association,
mucosal invasion of neoplasia) were diagnosed as cancer (Ezoe et al., 2011). The lesions are present in some, but not all, of the images. The
2011). The EGC was defined as cancer whose depth of invasion is images have sufficient quality to permit the differentiation of cancer
from non-cancer (Fig. 1). As shown in Fig. 1, the participants were
asked to analyze each image in each set for:

(1). whether a lesion was present or not in the image;

(2). if present, the location of the lesion; and
(3). endoscopic diagnosis (cancer or non-cancer).

Test scores were marked for correct answers to (1), (2) and (3), up
to a total of 100 points. To avoid potential bias, participants were not in-
formed about the number of cancerous cases among the test sets.

3.2. Allocation

Eligible participating endoscopists were randomly allocated into

two groups - an e-learning group and a non-e-learning group - based
on stratification of pre-test scores, experience of endoscopy (number
of procedures performed), whether the endoscopist was an endoscopy
nurse or medical doctor, medical institution and country. The block ran-
domization method was used for randomization using Excel software
(Japanese version, Microsoft Co. Ltd., Tokyo). The randomized allocation
was performed by the statistician (H.I.) at the data center who was not
blinded for allocation. Because the statistician has never been
acquainted with the participants' information and performed randomi-
zation automatically based on the above-mentioned stratification rule,
we did not think that this would result in any bias for randomized allo-
cation. The participants who were allocated to the e-learning group
were allowed access to the e-learning system via the Internet from
May 1st, 2014 until June 15th, 2014 (e-learning period), whereas
those in the non-e-learning group were prevented from accessing the
e-learning system.

3.3. E-learning System

The e-learning system was composed of video lectures about basic

techniques and knowledge, and self-exercise tests to accumulate

Fig. 1. An example of the pre- and post-test. a. When the participant starts the test, an
endoscopic image appears. The first question is whether or not a localized lesion is
present. If the participant clicks “present” as shown by “A1” in the slide, the second
instruction is for the participant to click on the center of the detected lesion on the
image as shown by “Q1”. b. The third question is whether the detected lesion is
malignant or benign. c. The participant is then offered the chance to identify any Fig. 2. An example of a video clip. On the lecture page, lecture video clips can be viewed
additional lesion. online or can be downloaded.
 K. Yao et al. / EBioMedicine 9 (2016) 140–147 143

experience (Figs. 2 and 3). The content was originally constructed for
the purpose of improving the ability of a participant to endoscopically
detect EGC by conventional white light endoscopy. The e-learning con-
tent comprised:

1. Technique

– Lecture (video clips & slides) (Fig. 2)

2. Knowledge
– Test (ten questions without answers)
– Lecture (video clips & slides) (Fig. 2)
– Test (ten questions with answers and descriptions)

3. Experience: 100 cases for EGC detection training (Fig. 3)

– Mock test (ten cases with scores and no answers)

– Random version of the 100 cases
– Systematic version of the 100 cases
– Random version of the 100 cases
– Mock test (ten cases with scores and answers)

In the lecture regarding techniques, we included the following sub-

jects: (1) absolute necessity to complete ideal preparation with muco-
lytic and anti-foaming agents, (2) recommendation to use
antispasmodic agent such as hyoscine butylbromide to inhibit peristal-
sis and to improve mucosal visualization and (3) importance of avoiding
blind spots (Veitch et al., 2015; Yao, 2013). In order to avoid blind areas
during observation, we demonstrated a standardized procedure. This
includes adequate air insufflation to extend the gastric lumen in order
to separate folds, the rinsing of mucus and bubbles from the mucosal
surface, and minimally required standard practice for screening proce-
dures to enable mapping observation of the whole gastric mucosa.
That procedure was originally proposed as a systematic screening pro-
tocol for the stomach (SSS), as described previously (Veitch et al.,
2015; Yao, 2013).
In the lecture regarding knowledge, we demonstrated the following
subjects: (1) endoscopic appearance of normal gastric mucosa vs. that
of abnormal high-risk condition for gastric cancer, such as atrophic gas-
tritis or intestinal metaplasia, (2) how to detect suspicious lesions in the
stomach and (3) how to characterize a detected lesion according to
macroscopic type, i.e. gastritis-like (G), ulcerative (U) and polypoid
(P) lesions. This diagnostic system was named the GUP system and
was originally proposed for use in this e-learning system.
In order to improve the endoscopist's ability to detect subtle muco-
sal gastric cancer, experience is imperative, in addition to good tech-
niques and knowledge. After learning the techniques and acquiring
knowledge, if participants can then accumulate experience with numer-
ous endoscopic images of cases with cancerous lesions and of cases with
non-cancerous lesions, they should then be able to make a correct diag-
nosis at a glance. From such a perspective, we developed a self-exercise
test program which includes images of 100 cases for detection. The
cases comprise 50 early gastric cancers and 50 non-cancerous lesions.
We prepared 100 sets of images, each set comprising 3 images for one
Fig. 3. An example of the self-exercise tests for diagnosis of 100 cases. a. One case
case of either cancer or non-cancer, as shown in Fig. 3. The participant
comprises a set of three slides. The 1st slide showes one endoscopic photo where one
should make continuous effort to make a diagnosis of the presented im- lesion is present. First,the participant should click to choose whether the lesion is cancer
ages one after another throughout the 100 cases. The cases are arranged or non-cancer. b. Immediately after clicking on their choice, an illustration indicating
either in random order or systematic order according to the GUP sys- whether the answer is correct or incorrect appears as the 2nd slide. c. The 3rd slide
tem. Repeated practice of quick question and quick answer in 100 indicates brief instructions on how to characterize the endoscopic findings so as to make
a correct diagnosis, and shows the original endoscopic image again.
cases offers the participants substantial experience in discerning be-
tween cancer and non-cancer in their own minds.
The participants who were allocated to the non-e-learning group
were taught nothing during the e-learning period.
144 K. Yao et al. / EBioMedicine 9 (2016) 140–147

3.4. Post-test Evaluation 3.6. Sample Size

After the e-learning period, all participants (both e-learning and In this trial, we decided to invite as large a number of participants as
non-e-learning groups) received the post-test between June 16th, possible without calculation of an actual sample size in order to expand
2014 and July 31st, 2014. The post-test evaluation used the same format the benefit of e-learning. In a usual clinical trial, we need to minimize
and methodology as the pre-test evaluation, however the participants the number of participants who receive random allocation from an eth-
did not know how the post-test questions differed from the pre-test ical viewpoint. However, in this study, there was no significant disad-
questions, in order to minimize any carry-over effect. vantage for the participants because eventually, all participants were
The e-tests and the e-learning content were prepared by the lead re- able to receive the e-learning.
searcher (K. Y.). Endoscopic images were supplied by the Department of
Endoscopy at Fukuoka University Chikushi Hospital. All images were
3.7. Statistical Analysis
taken during actual clinical practice. All patient information (including
patient ID, name, gender, age and date of examination) was removed,
The data were expressed as mean (standard deviation, SD). The dif-
and each image was allocated a new number for tracking purposes.
ference in score improvement between the two groups was examined
The e-test and the e-learning system were both originally constructed
by an independent-sample t-test, with a P value of b0.05 indicating sta-
by an information technology engineer (Y. T.)
tistical significance. All analyses were performed using SPSS software
(IBM SPSS Statistics Ver. 20, Chicago, IL).
3.5. Study Outcomes

The primary outcome was the difference in the degree of improve- 4. Results
ment in e-test results between the “e-learning” and “non-e-learning”
groups. The degree of improvement was determined as post-test result 4.1. Participant Flow and Baseline Characteristics
(score)/pre-test result (score). The secondary outcomes were the differ-
ence in the degree of improvement in the e-test according to pre-test Among the 515 endoscopists from 35 countries assessed for eligibil-
score, experience of endoscopy and geographical difference. The partic- ity, 332 participants from 27 countries who met the inclusion criteria
ipants were divided into low and high pre-test groups according to the completed the pre-test and were enrolled in the study. Of these partic-
mean pre-test score of the whole baseline group. Experience of endos- ipants, 166 were allocated to the e-learning group and 166 to the non-e-
copy was divided into less experienced (b8 years) and experienced learning group. During the e-learning period, 151 participants in the e-
(≥8 years) groups based on the median years of experience (7 years). learning group completed both e-learning and the post-test, while 144
Geographical region was divided into Asia-Pacific, Europe and Latin participants in the non-e-learning group completed only the post-test
America. (Fig. 4). The data of 151 participants in the e-learning and 144 in the

Fig. 4. Details regarding enrollment of participants, randomization and e-tests. IC: informed consent.
 K. Yao et al. / EBioMedicine 9 (2016) 140–147 145

non-e-learning groups were analyzed. The baseline characteristics were Table 2

similar in the e-learning and the non-e-learning groups (Table 1). Degree of improvement in test score between the e-learning group vs. the non-e-learning
group.

E-learning group Non-e-learning group

4.2. Study Outcomes n Mean rate SD n Mean rate SD

Overall 151 1·24 0·26 144 1·00 0·16 a

The mean pre-test score (SD) in the e-learning group was 51·4 Lower score group 86 1·34 0·29 80 1·03 0·18 a
(10·9), which improved to 62·2 (11·2) at the post-test. On the other Higher score group 65 1·19 0·14 64 1·03 0·11 a
hand, the mean pre-test score in the non-e-learning group was 52·6 Less experienced group 84 1·28 0·26 72 0·98 0·17 a
(10·3), which remained almost unchanged at 52·4 (11·4) at the post- More experienced group 67 1·19 0·26 72 1·03 0·14 a
Asia-Oceania 32 1·33 0·34 30 1·05 0·17 a
test. Accordingly, the mean rate of improvement of the test score was Europe 22 1·18 0·24 21 0·94 0·23 b
significantly better in the e-learning group than in the non-e-learning Latin America 97 1·23 0·23 93 1·00 0·13 a
group [1·24 (0·26) vs. 1·00 (0·16), P b 0·001, Table 2].
a. P b 0.001 for e-learning group vs. no e-learning group.
Subgroup analyses according to the pre-test score, experience of en- b. P = 0.002 for e-learning group vs. no e-learning group.
doscopy and geographical region showed that the mean rate of im-
provement of the test score in the e-learning group was significantly
higher than that in the non-e-learning group among all subgroups 5. Discussion
(Table 2).
There were no reports of unpleasant effects, such as the creation of 5.1. Key Results
too much overload on the living activities of the participants during
the trial. In addition, no technical trouble was encountered with the e- This is the first international randomized controlled trial to show
test and the e-learning system that caused participation to be that an e-learning system is effective for increasing the ability of
discontinued during the trial. endoscopists worldwide to expand their knowledge and gain invaluable
experience regarding the endoscopic detection of EGC, as demonstrated
Table 1 by an improvement in their test score. According to the subgroup anal-
Characteristics of the participants at baseline. yses, the e-learning system was effective irrespective of the pre-test
score, the endoscopist's experience or geographical area.
E-learning group Non-e-learning group
(n=166) (n=166)
5.2. Efficacy of E-learning System
Qualification
Medical 163 164
doctor In this study, we constructed an e-learning system based on the In-
Endoscopy 3 2 ternet. To date, conventional instruction has been conducted on a one-
nurse to-one basis by tutorial teaching. Hands-on seminars are also efficient
Pre-test score (%)
for passing on knowledge and skills hand-to-hand. Nevertheless, the ef-
Median 51.8 52.2
(range) (28.0 - 71.6) (27.5 - 74.6) ficacy of such instruction and seminars is limited to small numbers of
Experience of trainees. Lectures can provide instruction to a larger audience, but the
endoscopy (yr) impact is still limited to perhaps a few hundred attendees. Printed liter-
1-3 52 52 ature has been believed to be the most effective tool in the field of mass
4-7 38 33
8-10 30 15
education. However, in the case of endoscopy, it is difficult to provide
11- 46 66 content that can effectively teach technique as well as promote knowl-
Area Country edge. As constructed in this study, an e-learning system based on the In-
Asia-Oseania ternet offers a huge advantage over the above-mentioned conventional
Australia 0 1
teaching methods in that there is no limit on the number of learners
China 8 8
India 1 0 who can participate. In addition, we were able to upload educational
Korea 3 3 content using an originally developed system which includes an original
Malaysia 5 5 concept. The outcome in this study clearly shows that good practice
Singapore 3 3 based on good knowledge can certainly improve the ability of the par-
Thailand 9 8
Turkey 4 4
ticipants. It has been reported that a lecture from an expert does im-
Europe prove the ability of an endoscopist to make a correct diagnosis (Mabe
England 8 8 et al., 2014). However, to the best of our knowledge, this is the first re-
Italy 3 4 port to demonstrate that an e-learning system based on the Internet can
Poland 5 4
improve the diagnostic ability of gastrointestinal endoscopists
Portugal 2 2
Russia 8 9 worldwide.
Latin America
Argentina 13 12 5.3. E-learning Content
Bolivia 5 5
Brazil 28 28
Chile 7 7
The e-learning content focused on just three subjects, these being
Colombia 23 23 technique, knowledge and experience. The Internet is in fact a useful
Costa Rica 7 7 method for distributing content to an unlimited number of people, but
Ecuador 2 1 the quality of the content is obviously paramount. Among those three
El Salvador 1 1
subjects, endoscopists can acquire knowledge and technique by attend-
Guatemala 0 1
Mexico 3 2 ing conventional lectures or hands-on seminars. However, it is difficult
Paraguay 0 1 for learners to accumulate experience by a single lecture or hands-on
Peru 11 11 seminar. Therefore, we incorporated 100 cases of EGC detection training
Uruguay 6 6 into this e-learning system. We believe that simple but repetitive prac-
Venezuela 1 2
tice is useful for maintaining ability in any learning opportunity. Mabe
146 K. Yao et al. / EBioMedicine 9 (2016) 140–147

et al. indicated that the learning effect may decrease if endoscopists do 9. Yao F: Participant recruitment, acquisition of data, revising the
not continue their learning practice (Mabe et al., 2014). An Internet- manuscript and final approval of the article
based e-learning system has the advantage that endoscopists can repeat 10. Parra-Blanco A: Participant recruitment, acquisition of data, revis-
the practice and maintain their ability wherever and whenever they ing the manuscript and final approval of the article
wish. 11. Ho SH: Participant recruitment, acquisition of data, revising the
manuscript and final approval of the article
5.4. Limitations 12. Avendano AG: Participant recruitment, acquisition of data, revising
the manuscript and final approval of the article
One of the limitations of this trial is that the primary outcome was 13. Piscoya A: Participant recruitment, acquisition of data, revising the
not an improvement of the detection rate of EGC in actual clinical prac- manuscript and final approval of the article
tice, but an improvement of test scores. Nevertheless, we have already 14. Fedorov E: Participant recruitment, acquisition of data, revising the
started a clinical study to investigate the improvement of EGC detection manuscript and final approval of the article
rate in real clinical practice after finalizing this trial as described in the 15. Bialek AP: Participant recruitment, acquisition of data, revising the
protocol (UMIN: R000012039). In that study, the number of detected manuscript and final approval of the article
EGCs during the post-e-learning period (one-year after this study) will 16. Mitrakov A: Participant recruitment, acquisition of data, revising
be compared with that during the pre-e-learning period (the one- the manuscript and final approval of the article
year period prior to this study). Another limitation is that this 17. Caro L: Participant recruitment, acquisition of data, revising the
Internet-based system was not designed to be interactive. We hope to manuscript and final approval of the article
improve the system thereby enabling it to accept questions or allow dis- 18. Gonen C: Participant recruitment, acquisition of data, revising the
cussion from the participants. Finally, this system has only an English manuscript and final approval of the article
version currently, however other major language versions could be pro- 19. Dolwani S: Participant recruitment, acquisition of data, revising the
vided in the future. manuscript and final approval of the article
20. Farca A: Participant recruitment, acquisition of data, revising the
5.5. Generalizations manuscript and final approval of the article
21. Cuaresma LF: Participant recruitment, acquisition of data, revising
The content of this e-learning system did not utilize advanced imag- the manuscript and final approval of the article
ing endoscopy, but conventional white-light endoscopy alone which 22. Bonilla JJ: Participant recruitment, acquisition of data, revising the
can be available in any facility in the world. If we were to upload this manuscript and final approval of the article
system onto an official website and offer participation free of charge, 23. Kasetsermwiriya W: Participant recruitment, acquisition of data, re-
then unlimited numbers of endoscopists worldwide would have the op- vising the manuscript and final approval of the article
portunity to learn how to make an endoscopic diagnosis of EGC. Fur- 24. Ragunath K: Participant recruitment, acquisition of data, revising
thermore, this e-learning system could be modified to provide the manuscript and final approval of the article
education regarding endoscopic diagnosis in other organs, such as the 25. Kim SE: Participant recruitment, acquisition of data, revising the
large intestine and the esophagus, as well as the stomach. It may con- manuscript and final approval of the article
tribute to human welfare and health by reducing the mortality from 26. Marini M: Participant recruitment, acquisition of data, revising the
gastrointestinal cancer. manuscript and final approval of the article
In conclusion, as clearly shown by the increased test scores of partic- 27. Li H: Participant recruitment, acquisition of data, revising the man-
ipants in the e-learning group, this multicenter randomized controlled uscript and final approval of the article
trial has successfully demonstrated that an Internet-based e-learning 28. Cimmino DG: Participant recruitment, acquisition of data, revising
system was effective in enabling health practitioners around the the manuscript and final approval of the article
world to improve their knowledge and experience with regard to mak- 29. Piskorz MM: Participant recruitment, acquisition of data, revising
ing an endoscopic detection of EGC. the manuscript and final approval of the article
30. Iacopini F: Participant recruitment, acquisition of data, revising the
Funding manuscript and final approval of the article
31. So JB: Participant recruitment, acquisition of data, revising the man-
The Central Research Institute of Fukuoka University (I) and JSPS uscript and final approval of the article
Core-to-Core Program (B. Asia-Africa Science Platforms). 32. Yamazaki K: Participant recruitment, acquisition of data, revising
the manuscript and final approval of the article
Author Contributions 33. Kim GH: Participant recruitment, acquisition of data, revising the
manuscript and final approval of the article
34. Ang TL: Participant recruitment, acquisition of data, revising the
1. Yao K: Conception and design, analysis and interpretation of data,
manuscript and final approval of the article
participant recruitment, and drafting the article
35. Milhomem-Cardoso DM: Participant recruitment, acquisition of
2. Uedo N: Conception and design, analysis and interpretation of data, data, revising the manuscript and final approval of the article
participant recruitment 36. Waldbaum CA: Participant recruitment, acquisition of data, revising
3. Muto M: Conception and design the manuscript and final approval of the article
4. Ishikawa H: Conception and design, analysis and interpretation of 37. Piedra Carvajal WA: Participant recruitment, acquisition of data, re-
data vising the manuscript and final approval of the article
5. Cardona HJ: Participant recruitment, acquisition of data, revising 38. Hayward CM: Participant recruitment, acquisition of data, revising
the manuscript and final approval of the article the manuscript and final approval of the article
6. Castro Fiho EC: Participant recruitment, acquisition of data, revising 39. Singh R: Participant recruitment, acquisition of data, revising the
the manuscript and final approval of the article manuscript and final approval of the article
7. Pittayanon R: Participant recruitment, acquisition of data, revising 40. Banerjee R: Participant recruitment, acquisition of data, revising the
the manuscript and final approval of the article manuscript and final approval of the article
8. Olano C: Participant recruitment, acquisition of data, revising the 41. Anagnostopoulos GK: Participant recruitment, acquisition of data,
manuscript and final approval of the article revising the manuscript and final approval of the article
 K. Yao et al. / EBioMedicine 9 (2016) 140–147 147

42. Takahashi Y: Engineering, acquisition of data, revising the manu- GLOBOCAN, 2012. Estimated Incidence, Mortality and Prevalence in 2012. Available from
http://globocan.iarc.fr/Default.aspx.
script and final approval of the article Japanese Gastric Cancer Association, 2011. Japanese classification of gastric carcinoma:
3rd English edition. Gastric Cancer 14, 101–112.
Acknowledgements Leung, W.K1., Wu, M.S., Kakugawa, Y., et al., 2008. Screening for gastric cancer in Asia:
current evidence and practice. Lancet Oncol. 9, 279–287.
Mabe, K., Yao, K., Nojima, M., et al., 2014. An educational intervention to improve the
This trial was supported by the Central Research Institute of Fukuoka endoscopist's ability to correctly diagnose small gastric lesions using magnifying en-
University (I) (111001) and JSPS Core-to-Core Program (B. Asia-Africa doscopy with narrow-band imaging. Ann. Gastroenterol. 27 (149-5).
Schlemper, R.J., Kato, Y., Stolte, M., 2000. Diagnostic criteria for gastrointestinal carcinoma
Science Platforms). We would like to express our sincere appreciation in Japan and western countries: proposal for a new classification system of gastroin-
to Prof. Fatima Aparecida Ferreira Figueiredo (Head of the Digestive En- testinal epithelial neoplasia. J. Gastroenterol. Hepatol. 15 (Suppl), C52–C60.
doscopy Department of the Copa D'Or Hospital and Quinta D'Or Hospi- Schulz KF, Altman DG, Moher D; CONSORT Group. CONSORT 2010 Statement: updated
guidelines for reporting parallel group randomised trials. BMC Med. 2010; 24:18.
tal, Rio de Janeiro, Brazil) who participated as one of the team leaders of
Veitch, A.M., Uedo, N., Yao, K., East, J.E., 2015. Upper GI endoscopy quality and quality as-
GEST and who made a substantial effort to bring this project to fruition surance: optimizing early upper GI cancer detection and reducing miss rates. Nat.
before she passed away in 2014. We wish to thank Miss Katherine Miller Rev. Gastroenterol. Hepatol. 12, 660–667.
(Royal English Language Centre, Fukuoka, Japan) for correcting the En- World Medical Association Declaration of Helsinki, 2013. Ethical principles for medical re-
search involving human subjects. JAMA 310, 2191–2194.
glish used in this article. Yao, K., 2012. Development of e-Learning System for Endoscopic Diagnosis of Gastric Can-
cer: An International Multicenter Trial. Global e-Endo Study Team (GEST) (2012;
References Available from: http://www.temdec.med.kyushu-u.ac.jp/html/katsudo/ATS/
2012ATS/presentation/SIG-5_Kenshi_Yao.pdf).
American Cancer Society. Cancer facts & figures, 2014. Available from http://www. Yao, K., 2013. The endoscopic diagnosis of early gastric cancer. Ann. Gastroenterol. 26,
cancer.org/research/cancerfactsstatistics/cancerfactsfigures2014/. 11–22.
American Gastroenterological Association (AGA), 2008. Institute technology assessment
on image-enhanced endoscopy. Gastroenteology 134, 327–340.
Ezoe, Y., Muto, M., Uedo, N., et al., 2011. Magnifying narrowband imaging is more accurate
than conventional white-light imaging in diagnosis of gastric mucosal cancer. Gastro-
enterology 141, 2017–2025.