DR. IRINA F.

DRAGAN (Orcid ID : 0000-0002-4045-8564)

PROF. NIKOS MATTHEOS (Orcid ID : 0000-0001-7358-7496)

Accepted Article
Article type : Original Article

A global perspective on implant education: cluster analysis of the ‘first

dental implant experience’ of dentists from 84 nationalities.

Running Title: global survey of the ‘first dental implant experience’

Key words: Dental Implants; Education, Dental; Education, Dental, Continuing; Education,
Dental, Graduate; Surveys and Questionnaires

Authors

1. Irina F. Dragan* a
2. Miha Pirc b
3. Cristina Rizea c
4. Jie Yao d
5. Aneesha Acharya* d, e
6. Nikos Mattheos d

* These authors contributed equally to the manuscript.

Authors’ affiliations:

a) Department of Periodontology, Tufts University School of Dental Medicine, Boston, USA.

b) Dental Medicine Section, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia
c) Removable Prosthodontics Department, Faculty of Dentistry, University of Medicine and
Pharmacy, Bucharest, Romania
d) Implant Dentistry, Faculty of Dentistry, The University of Hong Kong
e) Department of Periodontology, Dr. D Y Patil Dental College and Hospital, Dr. D Y Patil
Vidyapeeth, Pune, India,

This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1111/eje.12426
This article is protected by copyright. All rights reserved.
 Author for Correspondence:
Accepted Article
Nikos Mattheos

Address: Oral Rehabilitation of the Faculty of Dentistry, 3A Prince Philip Dental Hospital, 34
Hospital Road, Sai Yin Pun, Hong Kong, SAR Hong Kong.

Email: mattheos@hku.hk; nikos@mattheos.net

Phone: +852-28590526 Fax: +852-2517 0544

Abstract

Objectives: The aim of this cross-sectional study was to explore the demographic and
educational patterns related to the dentists’ first implant dental experience. Material and
methods: Participants of a Massive Open Online Course in implant dentistry who have placed
and restored implants completed a 25-item online questionnaire investigating their pathway of
education and assessing their experience with the ‘first implant placement’. Exploratory
analysis included hierarchical clustering using 9 demographic categorical factors. Results: 1015
respondents from 84 countries formed 5 distinct clusters. Age and work-experience were
dominant clustering traits, decreasing from Cluster 1 to Cluster 5. Clusters 1 and 3 represented
‘senior’ and ‘younger’ general dental practitioners, respectively, while Cluster 2 and 4,
represented postgraduate educated clinicians. Cluster 5 represented recent graduates. Asia,
South America, and Africa were over-represented in ‘younger’ clusters. Time in practice was a
significant determinant of attitudes, followed by completion of postgraduate education. There
were significant differences in reported patterns of challenges and complications depending on
dentists’ time in practice, age and postgraduate education. Challenge in implant positioning was
more frequently identified by ‘young’ postgraduate educated dentists. Recent graduates
reported having the fewest complications of all clusters. Obtaining implant education in
university settings was most frequently recommended by clusters of dentists with
postgraduate education. Conclusions: Time in practice is a parameter to be considered when
designing implant education. Absence of structured education and mentorship might lead to
inability to properly assess treatment outcomes and identify complications. Quality-assured,
practice-directed education is needed at a global level to support in particular recent graduates
who now seem to engage with implant dentistry early in their career.

Introduction:

The use of dental implants is perhaps one of the most profound changes that define modern
dentistry. Implant dentistry is not only a young scientific discipline but also a clinical practice
that is expanding exponentially. In the span of a few decades, it has grown from an experimental
and later niche therapy practiced by a few elite ‘specialists’ to a ‘mainstream’ practice with a
global reach. The diffusion of any medical technology to mainstream clinical practice is a

This article is protected by copyright. All rights reserved.

 complex phenomenon (1), fuelled in part by industry driven manufacturing and marketing (2)
alongside its cost-effectiveness (3), user perception (4, 5) and regional geosocial factors (6).
Inevitably and in parallel with the expansion of implant dentistry, the need for standardized and
structured implant dentistry education has been recognized through global consensus among
Accepted Article
clinicians, researchers, and educators (7, 8, 9, 10, 11). Pathways for competency development in
the practice of implant dentistry are diverse and vary significantly across the world. Four key
stages pertinent to implant dentistry education have been identified: pre-clinical, basic-clinical,
advanced-clinical and specialist level (12).

Typically, stakeholders such as professional associations (13) and regulatory authorities drive
the growth, formal curricular development, and accreditation of any emergent healthcare
discipline (14, 15). Whether implant dentistry merits a separate specialty or subspecialty has
been a subject of considerable debate (16,17,18). Traditionally, medical and even dental
specialities have been based upon different clinical or anatomical domains, and subspecialities
emerge either as a result of biomedical advancements within an existing speciality, or a
recognition of a societal group with unique demands (19). The practice of implant dentistry
involves competencies from oral surgery, periodontics, restorative and prosthetic dentistry,
inevitably being recognized as a multispecialty domain, while always remaining part of a
comprehensive care treatment plan. Such a composite profile makes it difficult to be ‘siloed’ as a
subspecialty of any of the existing formal dental specialties. The resultant multiplicity of
professional stakeholders, along with a perceived high commercial value, may have together
contributed to a rather unregulated surge, where actual implant practice has by far outpaced
formal university-based training (9). This gap has until recently, been largely filled by informal,
short-term, unstructured and often industry-initiated educational opportunities (20).

In addition, these developments have occurred in a broader context of the fast-changing

landscape of dentistry at the global level. Some of these changes include sharp rises in: the
dental market and workforce in developing nations (21, 22), international migration of dental
manpower (23), adoption of digital and 3-D imaging technologies in dentistry (24), industry-
professional relationships (25), e-learning (26), and social media use by professionals and
consumers (27, 28, 29). As standardization of university curricula in implant dentistry remains
a focal issue (9), an understanding of factors that underpin the educational pathways in implant
dentistry is in order. The ‘diffusion of innovation’ model, a well-studied framework that
describes the dispersion of technology (1, 4) categorises individuals as ‘innovators’, ‘early’,
‘majority’, or ‘late’ adopters. As with other technology, implant practitioners over the last
decades may be similarly categorised. Placing ‘the first dental implant’ represents a clinician’s
primordial clinical experience in the practice of implant dentistry. The correlations and details
of this experience across demographics could potentially reveal patterns, trends, and existing
shortcomings in implant-related education. Such information could be particularly valuable for
educators and policymakers. With this basis, the current study was designed to explore the ‘first
dental implant experience’ and the education that preceded it, using an online- delivered survey
offered to a pan-global sample.

This article is protected by copyright. All rights reserved.

 Materials and Methods:
Accepted Article
Study design

The study was designed as a cross-sectional, online survey based, exploratory study. Ethical
approval was obtained by the Institutional Ethical Review Board of The University of Hong Kong
and Hospital Authority West Cluster (reference number: UW 16-088).

Survey questionnaire and study sample

A 25-item original, online questionnaire was designed to assess factors and specific natures of
experiences related to clinicians’ first implant placement (supplementary file: appendix 1). The
questions were designed to address four categories: demographic data, education, first dental
implant placement experience, and professional guidance. Dentists were eligible to enroll in the
study if they had already performed at least one implant surgical placement or restoration or
both. The survey instrument was designed based on a review of pertinent literature and the
authors’ perspectives. Content and face validity of questionnaire items were assessed and
agreed upon by two authors who were experienced in the domain of implant education (I.D,
N.M). The survey document consisted of a short introduction and informed consent for
electronic approval and was hosted on Tufts University ‘Qualtrics’ online platform. It was
distributed through email and social media networks study through an electronic link. A
snowball sampling strategy was employed: voluntary data submitted by participants of an
online course in implant dentistry [the University of Hong Kong Massive Open Online Course
(MOOC) in Implant Dentistry, (www.coursera.com/learn/implant-dentistry)] was used to invite
contribution to the study by self-participation and sharing of the electronic survey link with
potentially interested dentists. No response inducements were provided. By the time the study
was launched, the MOOC had engaged more than 10,000 visitors and 6,000 active learners. The
duration for response collection was from February 2016 to February 2017.

Data analysis

Data from respondents who reported having placed their first implant was analysed.
Responders were clustered on basis of 9 background and demographic factor variables, namely:
‘Age group’, ‘Gender’, ‘Continent where located’, ‘Work experience’, ‘Professional setting’,
‘Timing of first implant after graduation’, ‘Setting of first implant placement’, ‘Type of Practice’,
‘Postgraduate Training’, which together formed 37 individual categorical levels (Table 1).

All statistical analyses were performed in the R statistical environment (R version 3.1.3, URL
http://www.R-project.org/.). For exploratory analysis of how the selected factor variables
impact the first implant experience, a hierarchical cluster analysis method was employed. This
method combines two exploratory multivariate analyses: principal component analysis which
reduces multidimensionality and cluster analysis which groups datasets. Firstly, a ‘principal

This article is protected by copyright. All rights reserved.

 components ordination’ was performed on basis of the selected 9 factor variables by using
multiple correspondence analysis, performed in the R package ‘FactoMiner’
(https://cran.rproject.org/web/packages/FactoMineR/FactoMineR) (30). Multiple
correspondence analysis can be applied to a table of categorical variables and assigns
Accepted Article
coordinates to each individual along every categorical level, a process termed as ‘ordination’.
This enables two types of relationships to be visualized: that between factor categories and that
between individuals where the distance indicates the level of similarity in their response profile
of factor categories.

Next, this ordination was used to ‘cluster’ individuals by applying ‘hierarchical clustering on
principal components’ (HCPC), using Ward’s criterion in the FactoMiner package. HCPC is an
agglomerative hierarchical clustering (31), which assigns each individual to a cluster and
successively combines neighboring clusters until an appropriate number of clusters is reached.
The optimal number of clusters was selected of inter-cluster inertia gains, which measures the
degree of homogeneity between the clusters.

Each cluster was described by analyzing the relationships of each of 37 individual categories to
a particular cluster on basis of the v-test and p value. A positive v-test with a p value <0.05
implies that the particular categorical level was significantly over-represented in that cluster.
Thus each cluster can be described on the basis of these over-represented categorical levels and
their relative importance to the clustering gauged on the value of the v-test statistic.

Next descriptive statistics and pairwise Chi square tests (p-value corrected using False
Discovery Ratio (FDR), significant at p= 0.05) were applied to compare the frequency of
response levels to 15 categorical Questionnaire items (10-14, 16-25) between the defined
clusters. These items were related to the first implant experience and its relevant education.

Results:

Data Collection: A total of 1102 survey documents were collected from which 1015 respondents
reported having placed their first implants and were deemed suitable for analysis. Data from 87
participants who reported not having placed or restored their first dental implant were
included for analysis in this study.

Description of clusters:

Data collected from 1015 respondents was used for hierarchical cluster analysis. These
respondents were from 84 countries and distributed over 6 continents (Figure 1). Figure 2
presents a ‘groups representation plot’ that depicts how the factors evaluated were related to

This article is protected by copyright. All rights reserved.

 each other and contributed to the samples’ ordination in multiple correspondence analysis
along the first two axes. “Age group”, “Work experience”, and “Time of first implant” were
aligned with the first principal coordinate plot axis, which contributed to 10.4% variation
among members. “Postgraduate training” and “Practice type” were aligned along the second
Accepted Article
principal coordinate plot axis. “Professional setting”, “Continent located”, and “Setting of first
implant” were closely placed and intermediate along both of the first two axes.

The optimal number of clusters selected on basis of inter-cluster inertia gains was 5. The cluster
‘dendrogram’ or tree, which depicts the relationship between respondents, is presented in
Figure 3. The 5 defined clusters comprised of 123 respondents in Cluster-1, 246 in Cluster 2,
233 in Cluster 3, 200 in Cluster 4, and 213 in Cluster 5. The cluster analysis indicated that the
variables that predominantly characterized the clusters were “Age group”, “Work experience”
and “Time of first implant” with decreasing time-order from Cluster1-5. Cluster 4, however,
differed and was most strongly characterized by postgraduate qualification and practice type.

A description of each cluster is presented in Table 2, which shows the categories significantly
over-represented in each cluster and the percentage of responders in that cluster for that
particular category. Figure 4 depicts an infographic showing the relative importance of the
topmost contributory category levels to a particular cluster (with v.test values >3). To provide a
visualization of a ‘typical’ cluster member, the profiles of 5 individuals closest to each cluster
center are presented in Table 3.

Typical Cluster 1 individuals appeared to be older than 45 years of age, have a working
experience of more than 10 years, predominantly worked in private practice, of a G.P type
(79%) and had placed their first implant 10 years after graduation. About 80% of the
respondents in this cluster had placed their implant in non-university settings. Almost equal
numbers of postgraduate qualified and non-qualified dentists were represented in this cluster.
Similarly, location in Asia was most common (43%), followed by North America (17%), and
Europe (17.8%). About two-thirds in this cluster (74%) were males. This cluster was named as
‘senior private practitioner’ or “S-P”.

Cluster 2 members were in the age range of 30-45 (87%). Typically, this cluster also presented a
working experience of more than 10 years and 80% members were in private practice.
Postgraduate training completed was a predominant characteristic (71.1%), though a G.P
practice type was more common (60%) than specialty or postgraduate trainee (40%) type of
practice. The placement of the first implant was commonly in non-university settings (82%). In
terms of location, Asia was the most common location (54.5%) followed by Europe (40%)
which was overall significantly overrepresented in this cluster. North America was not
represented in this cluster at all. Males were again comprised two-thirds in this cluster (75%).
This cluster was considered to represent the ‘mature, postgraduate trained, private practitioner’
or “MPG-P”.

This article is protected by copyright. All rights reserved.

 In cluster 3, predominantly, the age groups of 25-30 (54.5%) and 30-35 (40.8)% years, a work
experience of 3-9 years (87.1%) and private practice (80.2%) were seen. Similar to cluster 2,
G.P type practice was more common (67.8%) but a comparatively high number of postgraduate
trainees were represented in this cluster (15.9%) and unlike cluster 2, completed postgraduate
Accepted Article
training was less common (45%). The placement of the first implant in non-university settings
was more common (70.3%) though lower than Clusters 1 and 2. Other than Asia (53.2%),
Europe (26.1%) and Africa (12%) were commonly represented. Similar to clusters 1 and 2,
males were about two-thirds in this cluster (72.9%). This cluster was deemed to represent the
‘young general–dentist private-practitioner’ or “Y-GP”.

Cluster 4, was very strongly characterized by postgraduate qualification (96.5%), specialty

practice type (85%) and working in a professional setting at a university or public hospital
(76%). University setting for the first implant placement was predominant (78%) and usually
within 0-5 years after graduation (77%). Most Cluster 4 members were aged between 25-35
years (71.5%). As compared to clusters 1 and 2, slightly more females were represented (37%)
and Asia was more commonly represented than in clusters 1 and 2 (71.5%). Thus, this cluster
mainly represented ‘young postgraduate specialist in public hospital or university’ or “YPG-U’.

The major characteristic of Cluster 5 was a work experience of 0-3 years after graduation
(96.2%) and no postgraduate training completed (81.7%). Age groups of 25-35 (57.5%) and
<25(35.5%) were typical. Professional settings included both university (35.2%) or public
hospital (18.7%) and private practice (41.7%). Corresponding frequencies for practice types of
postgraduate trainee (21.02%) or specialist (15.9%) and G.P (63.3%) were noted and setting
for the first implant was at university (35%) and private/C.E (35.5%). Interestingly, a
substantial proportion of members in this cluster did not provide information about the setting
of their first implant (28.9%). Geographically, while Asia (59.3%) was represented the highest,
Africa (14.4%) had higher relative representation in this cluster as compared to clusters 1-4.
Males again comprised about two-thirds (74.4%) of this cluster. This cluster seemed to largely
represent ‘recently graduated dentists or “RG”, working as G.P. in private practice or in
university postgraduate settings.

Comparison of first implant related items between clusters:

Q 10. Further training after the first implant

The lowest rate for ‘further training after first implant placement’ was noted for Cluster 5, the
young, recent graduates (RG), which was significantly lower than that for all other clusters.

This article is protected by copyright. All rights reserved.

 The highest rate for training was notable for Cluster 2 or MPG-P, which represented mature,
postgraduate trained practitioners and was significantly higher than that for Cluster 1 or S-P
and Cluster 3 or Y-GP, both of which had a lower representation of postgraduate trained
practitioners.
Accepted Article
Q 11. Type of further training after first implant placement

The dominant training mode was ‘Other’ across all clusters, described by about 60 % of
respondents in Clusters 1-3. A lower trend for this mode was noted in the ‘younger’ Clusters 4 &
5. Cluster 5 or RG had significantly lower rates than all other clusters and Cluster 4 or YPG-U
had significantly lower rates than Clusters 1(S-P), 2(MPG-P) and 3(Y-GP) (Figure 5).

An opposite trend was notable with regard to completed University-based further training.
Cluster 4 (YPG-U) (23%) and Cluster 5 (RG) (10.3%) had higher rates than the other clusters.
Cluster 4 (YPG-U) reported a significantly higher rate of undertaking 3-year University based
further training than all other clusters.

Interestingly, Cluster 2 (MPG-P) had the highest rate of 2-year University based further training
and significantly differed from Cluster 5 or RG in this regard.

Q 12: Nature of implant-related training

Didactic/lecture or theory-based training was the most common, reported by 70%-85% of all
respondents.

Training case-discussions ranged from 44-63%, with the highest rate for Cluster 2 (MPG-P),
followed by Cluster 3 (Y-GP) and significantly differed from the lowest reported by Cluster 5
(RG).

Similarly, pre-clinical lab training was highest for Cluster 2-MPG-P and generally higher for
Clusters 1-3 as compared to Cluster 4-YPG-U and Cluster 5 RG, which reported the lowest rates.
Cluster 4-YPG-U, though reporting the highest rates of completed university based-education
had a significantly lower rate of pre-clinical lab training than Cluster 2-MPG-P.

This article is protected by copyright. All rights reserved.

 Supervised surgical training ranged from 39.4% in Cluster 5-RG to 69.5% in Cluster 4 YPG-U.
Higher rates were reported by both the clusters characterized by postgraduate training: i.e.
Cluster 2-MPG-P and Cluster 4-YPG-U. A significantly lower rate was noted in Cluster 1 S-P as
compared to these two clusters. Cluster-5-RG reported significantly lower rates than all other
Accepted Article
clusters.

The rate of supervised restorative training was lower than that for surgical training ranging
from 22.5% for Cluster 5-RG to 50.4% in Cluster 2-MPG-P. Cluster-5-RG had significantly lower
rates than all other clusters. Again, the highest rates were noted for the two ‘postgraduate
trained’ clusters: Cluster 2-MPG-P and Cluster 4- YPG-U. Cluster 2-MPG-P had significantly
higher rates than Cluster 1-SP and Cluster 3 Y-GP.

Training categorized as ‘Other’ was most commonly reported by Cluster-1-SP and Cluster-5-RG.
These clusters showed rates significantly higher than that reported by Cluster-2-MPG-P and
Cluster 4-YPG-U.

Q 13. Nature of preparation for the first implant

In terms of clinical and 2D radiographic examination, which form the most basic level of
preparation for implant surgery, it was notable that the Cluster-5-RG reported significantly
lower rates than the other clusters.

Digital planning and 3-D printed cast use rates ranged 9.8% to 18.8% across the clusters. For 3-
D radiographic examination for their first implant, the ‘older’ clusters Cluster 1-SP and Cluster
2-MPG-P presented lower rates than other ‘younger’ clusters. Cluster-1-SP had significantly
lower rates than Cluster 3-Y-GP and Cluster 4-YPG-U. Concerning Digital planning, the group
that reported the highest rate was Cluster 5-RG (18.8%).

Case-discussion with instructor or mentor also varied significantly with Cluster-1-SP reporting
significantly lower rates than Clusters 2-MPG-P, Cluster 3-Y-GP and Cluster 4-YPG-U. A similar
pattern was also noted for the Cluster 5-RG with significantly lower rates than Clusters 2-4.

This article is protected by copyright. All rights reserved.

 Q 14. Type of guidance/supervision during the first implant
Accepted Article
One to one type of supervision was least reported by Cluster 1-SP (49.5%) and Cluster 5-RG
(46.9%). These were significantly lower than that for the two post-graduate trained Clusters 2-
MPG-P and 4-YPG-U.

The two ‘older’ clusters reported no supervision most commonly S-P (21.9%) and Cluster 2-
MPG-P (20.3%) and these were significantly higher than that for the other 3 Clusters 3-5.

No information on the nature of supervision was most frequently reported by Cluster 5-RG
(29.1%) and this was significantly higher than that for the Clusters 1-4.

Q 16. Duration of procedure for first implant

The shortest duration of the procedure: ‘less than one hour’ was reported by the two ‘post-
graduate trained clusters’: Cluster 2-MPG-P (50.4%) and Cluster 4-YPG-U (57.5%). Cluster 5-RG
least frequently reported placing their first implant in less than an hour (31.9%), which was
significantly lower than that for the Clusters 2-4. ‘No information’ on the duration of the
procedure was most frequently reported by Cluster 5-RG (25.4%) and was significantly higher
than that for all other Clusters.

Q 17. Edentulous span of first implant placement

The most commonly reported was ‘single implant’ across all clusters (63.4%-76.4%).

Cluster 2-MPG-P reported highest frequency of ‘multiple implants’ (22.36%) and this was
higher than that from the other ‘post-graduate trained’ cluster Cluster-4 YPG-U (16%) and
overall they differed significantly. Cluster-5-RG also differed significantly from other clusters.
25.4% of Cluster 5-RG did not provide information regarding the edentulous span whereas the
rates for other clusters were much lower (2% for Cluster -4-YPG-U to 9.8% for Cluster 3-Y-GP).

This article is protected by copyright. All rights reserved.

 Q 18. Use of surgical guide
Accepted Article
The use of a surgical guide was highest reported by Cluster 1-SP (43.9%). It was least frequently
reported by Cluster3-Y-GP ( 23.3%), which was significantly lower than that for Clusters 1-SP
and Cluster-4-YPG-U.

Q 19. Nature of challenges faced

The most frequently reported challenges across clusters were ‘identification of implant position’
(22. 8% for Cluster 1-SP to 39% for Cluster 3-Y-GP) and ‘preparation of osteotomy’ (20.2% for
Cluster 5-RG to 43% for Cluster 1-SP) (Figure 6).

Cluster 1-SP reported challenges in the identification of implant position’ at significantly lower
rates than Clusters 2-4. Similarly, Cluster-5-RG also reported significantly lower frequency in
identifying implant position than Clusters 3-Y-GP and Cluster 4-YPG-U.

Cluster 5-RG also reported a significantly lower frequency of challenges in ‘preparation of

osteotomy’ than Cluster-1-SP, Cluster 2-MPG-P and Cluster 4-YPG-U.

0% of Cluster 4-YPG-U reported ‘incision’ as a challenge, and this was significantly lower than
all other clusters. Incision was most frequently reported as a challenge by Cluster-5-RG (7.5%).

25.4% of Cluster-5-RG provided ‘No information’ on challenges as opposed to significantly

lower rates for all other clusters (3%-7.32%).

Q 20. Restoration type for the first implant

‘Cement retained crown’ was the most common type of restoration reported (61% for Cluster-
4-YPG-U to 31.46% for Cluster-5-RG). Cluster 4-YPG-U reported a significantly higher
frequency of a ‘cement retained crown’ than Cluster 3-Y-GP.

This article is protected by copyright. All rights reserved.

 Cluster 5 RG had the highest rate of ‘no information’ (25.35%) which was significantly higher
than that for all other clusters.
Accepted Article
The highest rates for ‘implant supported bridge was reported by Cluster 5-RG (36.6%) and
Cluster 3-Y-GP (30%). For ‘implant supported complete denture’, the highest rate was reported
for Cluster-2-MPG-P (12.2%).

Q 21. Type of complications encountered

‘No complication’ was most frequently reported by Cluster 5-RG (53%) followed by Cluster 3-Y-
GP (44.2%) and least frequently by Cluster 4-YPG-U (30.5%). Cluster-5-RG had a significantly
higher rate for reporting ‘no complication’ than Cluster 1-4. Similarly, Cluster 3 reported
significantly higher rates than Clusters 2 and 4.

‘Inappropriate implant positioning’ was overall the most frequently reported complication
ranging from 7.04% (for Cluster 5-RG) to 16.74% (for Cluster 3-Y-GP). Cluster 5-RG reported
significantly lower rates than all other clusters.

The opposite trend was noted in ‘flap design and closure’ where Cluster 5-RG and Cluster 3-Y-
GP had higher rates than other clusters. Cluster 4- YPG-U had the lowest rate (2.5%) for
problems in ‘flap design and closure’ which was significantly lower than that for Cluster 3-Y-GP
and Cluster 5-RG.

Q 22. Reflection on the level of preparation

Cluster 2-MPG-P had the highest frequency of reporting they were ‘adequately prepared’
(35.7%) while Cluster 5-RG had the lowest frequency (19.7%) which was significantly lower
than that for all other clusters. Cluster-5-RG also had the highest frequency of providing no
information (25.35%) which was significantly higher than that for all other clusters (Figure 7).

Q 23. Assignment of a formal mentor

Having a formal mentor was most frequently reported by Cluster 4-YPG-U (75%) which was
significantly higher than that for all other clusters, and least frequently reported by Cluster 5-RG
(54%).

This article is protected by copyright. All rights reserved.

 Q 24. Type of mentor assigned
Accepted Article
A mentor assigned by a University was the most frequently reported, ranging from 15.5% for
Cluster 2-MPG-P to 48% for Cluster 4-YPG-U. Cluster 4-YPG-U significantly more frequently
reported a mentor from a University than all other clusters, while Cluster 5-RG significantly
more frequently reported receiving such mentorship (27.2%) as compared to Cluster 2-MPG-P.

A mentor classified as ‘other’ was the second most frequent type, which showed a reverse trend
than that for type ‘from University’ (5% for Cluster 4-YPG-U to 22% for Cluster 1-SP). Cluster 4-
YPG-U classified mentorship as ‘other’ significantly less frequently than all other clusters.
Cluster 5-RG significantly less frequently reported of ‘other’ mentorship than Clusters-1-SP and
Cluster-2-MPG-P.

A mentor assigned from a professional association was most commonly reported by Cluster 3-
Y-GP.

Q 25. Advice for beginners

Cluster 4-YPG-U advised ‘University training’ with the highest frequency (70.8%), followed by
Cluster 2-MPG-P, while the lowest such rate was noted for Cluster 5-RG (49.8%). Cluster 4-YPG-
U advocated ‘University training’ at a significantly higher frequency than all other clusters
except the other ‘post-graduate trained’ Cluster 2-MPG-P.

Continuing education was advised significantly more frequently by Cluster 2-MPG-P (52.9%)
and Cluster 3-Y-GP (52.8%) than that by Cluster 5-RG (36.6%).

Having a mentor was advised with significantly higher frequency by Clusters 1-4 (65%-70%)
than by Cluster 5-RG (49.8%).

A similar trend was observed for the advice of ‘online education’ which ranged from 28.6%-
36.1%, where Cluster 5-RG reported lower frequency than all other clusters.

This article is protected by copyright. All rights reserved.

 Discussion:
Accepted Article
Practicing implant dentistry has been shown as related to several background factors, including
gender, location, specialty training, and professional setting (32). The demographic correlates of
implant-related education, though studied at regional level (33, 34, 35, 36, 37), are unclear at a
global perspective. In addition, as implant dentistry has been practiced widely for more than 30
years, differences in education and practice are anticipated among different generations of
dentists, reflecting the advances in technology, research, and legislation. The overall purpose of
the survey was to gather preliminary information concerning ‘the first dental implant
experience’ from a global population of practicing dentists and relate it to demographic
variables using exploratory data analysis (EDA).

EDA is valuable in discerning patterns that can lead hypothesis generation, making it
particularly suitable for empirical data (38). Multiple correspondence analysis (MCA) is a
common EDA approach used to reduce the complexity of large datasets (39) and has been
applied to health-related data (40, 41). It facilitates a visualisation of relationships between
individuals based on a composite of multiple categorical factors, thus permits clustering into
comprehensible groups (42). Importantly, such clustering allows for correlation among the
descriptive factors, as is very likely for demographic descriptors. Previous studies have
employed clustering approaches to address survey data in medical education (43, 44, 45). Past
research has demonstrated that medical technology adoption by physicians is influenced by
their characteristics and practice settings (46). In the present study, 9 demographic and
practice related variables were used to cluster the respondents, giving rise to a total of 37
individual categories, each of which served as a dimension for the agglomerative clustering.
Individuals more similar to each other along the composite of these dimensions were thus likely
to be represented within a single cluster. The optimal number of clusters was selected based on
inertia values as well as a visual inspection of the cluster dendrogram or tree, in order to
provide a comprehensive but meaningful characterization of the respondents.

The number of survey responses obtained through this virtual chain /snowball can be
considered high when compared to conventionally delivered surveys, as shown before (47). In
addition, the large participation base of the MOOC provided means to reach a high number of
potential participants. As this was a preliminary survey at the global demographic scale, no
target sample size was computed, but the objective was to reach out to as high a number of
respondents as possible. While no response inducements were provided, the inclusion of an
‘inducement question’ has been shown to increase survey response by arousing interest (48).
As no referral tracking was made, the exact response rate cannot be computed. We speculate
that inclusion of the question regarding when participants placed their first implant in relation
to graduation from dental school may have provided a basis for personal engagement for those
with this experience but could have contributed to non-response from those who did not have
any clinical experience in implant dentistry. It must be recognised that this survey dispersal is a
non-probability type entailing inherent bias and a ‘pass-along effect’ (49). These are very likely
to account for different sizes noted in clustering by respondent background. Although our

This article is protected by copyright. All rights reserved.

 analytical approach was designed to address the difference in cluster sizes, future research
should additionally include strategies for enhancing diversity in snowball sampling (50).
Accepted Article
The most important factors contributing to clustering were time-related. The clusters mainly
differed in the “Age group”, “Work experience” and “Time of first implant” which were not
surprisingly significantly inter-correlated and had highest correlations with the first principal
component (Figure 2), which describes the direction of maximum variation in the dataset.
‘Postgraduate training’ and ‘Practice type’ were similarly correlated and aligned with the second
principal component (Figure 2). Cluster 1-SP and Cluster 2-MPG-P formed ‘older’ clusters,
whereas Cluster 3-Y-GP, Cluster 4-YPG-U and Cluster 5-RG as ‘younger’ clusters. On the other
axis, Cluster 1-SP, Cluster 3-Y-GP, and Cluster 5-RG formed ‘GP’ clusters, while Cluster 2-MPG-P
and Cluster 4-YPG-U formed ‘Postgraduate educated clusters. Thus the 5 clusters differed
mainly in their age and professional career-time points and secondarily in their postgraduate
training and work setting.

Geographical location was of less relevance to clustering than time. This can have a significant
implication for the design of implant dentistry education, as it suggests that we have already
achieved a high level of “globalisation” in the practice of implant dentistry. In this sense, “time in
practice” is a much stronger determinant of educational experiences than geographical location.
This is in particular, an important finding when it comes to CPD, as “novice” implant dentists
can be typically identified among young graduates, but also among experienced dentists of older
generations. How to best serve each of these groups has been recognized as a challenge when
designing CPD (10). Despite the large representation from Asia in the overall sample, it was
notable that North America and Australasia were overrepresented in Cluster-1-SP
predominantly comprising of ‘senior practitioners’ who had placed implants. Similarly, Europe
was significantly overrepresented in Cluster-2-MPG-P or the ‘mature postgraduate trained
private practitioner. These may simply reflect regional variations in the timescale of growth in
implant dentistry where North America, Australia, and Europe have likely experienced an
earlier popularity of implant practice as compared to Asia, South America, and Africa. The
Cluster-4-YPG-U or the ‘young postgraduate in university or public hospital settings’ showed a
relative overrepresentation of location in Asia. This trend may represent not only a younger
demographic in terms of regional dental manpower in Asia (21, 22, 51) and a somewhat later
growth in popularity of implant dentistry, but also reflect regional differences in common
‘practice models’ in terms of public versus private healthcare (52) and number of dentistry
programs (53). Most developing nations such as China, India, and Malaysia house large
populations and have seen a relative upsurge in the number of dental training programs in the
last 1-2 decades (54). Correspondingly, these data seem to suggest that at a global scale, the
younger implant practitioner demographic maybe skewed towards Asia, South America, and
Africa, which is not surprising. At the same time, it is important to recognize that the sampling
methodology followed did not aim to address differences among different geographical or
demographic categories. As the sampling was done through an online questionnaire and much
of the dissemination was based on an online course, it is possible that the respondents are
skewed towards those more familiar with and likely to use social media and online learning or
towards certain geographical locations owing to linguistic/social factors. Thus the focus of this

This article is protected by copyright. All rights reserved.

 study was to analyse the qualitative differences among identified clusters to unravel broad
patterns, rather than quantifying or comparing prevailing practices among specific
demographic groups.
Accepted Article
‘Diffusion’ is the process by which a device or technology reaches potential end users (1, 4). The
‘early adopters’ of a technology include 10-15% of a population and play an important role in
technology dissemination by serving as opinion leaders and influencers (1, 4). Cluster 1-SP or
the “Senior private practitioners” may include a significant number of such ‘early’ and ‘early
majority’ adopters. One characteristic of ‘early adopters’ is high perceived self-efficacy (1, 54)
which is ‘the belief in one’s ability to succeed in a task’(55). Perceived competence is yet
another determinant of task-motivation (56). Speculatively, higher perceived self-efficacy and
competence could account for lower rates of seeking supervision, case discussion or formal
mentorship, than Clusters 2 and 4, the ‘post-graduate’ clusters. These practitioners reported
experienced fewer challenges in identifying the implant position as compared to ‘younger’
general practitioners, while they were less likely to report“inappropriate implant positioning’ as
a complication, something that was reported the highest among young postgraduate educated
dentists (Cluster 4-YPG-U). Plausibly, while this may be attributable to clinical judgement and
pre-existing competence in restorative dentistry, it could also imply that awareness of
restorative-driven implant positioning is much higher among the more recent specialists and
general practitioners than in previous decades. Cluster-1-SP is also most likely to typify the
implant educational experience in the previous 2 decades when the practice implant dentistry
lacked much of today’s standards. They frequently categorized their implant related training as
‘other’, which may reflect the lower availability of university-based implant education in the
past and in general for private practitioners (7, 34, 57). A survey reported more than 2 decades
ago indicated that typically, implant practitioners obtained CE type of implant education (58).
The lower rate of 3-D imaging as compared to ‘younger’ clusters 3-5 and greater use of surgical
guides, probably reflects lower penetration of digital technology to this group of practitioners,
as opposed to the present-day increasing popularity of digital aids such as CBCT in implant
planning (59), especially among younger practitioners.

Contrasting Cluster 2-MPG-P with Cluster 1-SP, both experienced groups, and Cluster 4 YPG-U
with Cluster 3-Y-GP, both younger groups, postgraduate training seems to confer a major
advantage irrespective of work experience. Cluster 2-MPG-P may also reflect the implant
experience during postgraduate training in the past decade or so. As with Cluster 1-SP, this
group was more likely to report ‘no supervision’ than the ‘younger’ clusters and have had a
mentor categorized as ‘other’. It is likely that at least for some members of this cluster, implant
experience might have been obtained outside a university setting, owing to its lack in the formal
curriculum at the time (7, 34, 57). These 2 ‘Postgraduate trained’ clusters reported themselves
as having the highest level of ‘adequate preparation’, validated to an extent by their shortest
duration for implant placement and no reports of challenges in the incision and flap reflection
by the ‘younger’ specialist Cluster 4-YPG-U, and cases with higher complexity by Cluster 2-MPG-
P. This may be attributable to more procedural experience in specific surgical and/or prosthetic
domains leading to improved clinical competence (60, 61). In surgery, the error rate is
influenced by the number of completed procedures (62). However, Cluster 4-YPG-U was the

This article is protected by copyright. All rights reserved.

 most likely to identify challenge or complication related to implant positioning as compared to
similarly ‘younger’ Clusters 3 and 5, with no speciality training. Higher awareness, knowledge
base and active supervision could be contributing factors.
Accepted Article
The structure and duration of university postgraduate and continuing professional
development programs varies with location (63). Cluster 2-MPG-P reported the highest rate of
further training after the first implant, and via 2-year full-time University programs. The
dominance of Europe in this cluster may suggest a regional greater availability of the 2-year
program structure. Cluster 4-YPG-U, however, showed a dominance of 3-year postgraduate
program training and the lowest rate of ‘other’ training. Implant training has been integrated
into most implant-related specialty postgraduate programs of oral surgery, periodontics, and
prosthodontics in Europe, North America, and also in Asia, which was dominant in this cluster
(7). As there currently exist only three 3-year postgraduate programmes dedicated to implant
dentistry in the world to the authors’ knowledge, the 3 year curricula reported by the
respondents refer to “mono-specialty” (periodontics, prosthetic dentistry or oral surgery)
programmes. It is not surprising to notice that clusters of senior and younger G.Ps did not
appear to rely on university-based implant education. Most likely, this reflects the fact of
absence of universities in the field of structured practice-oriented implant education of shorter
duration. Today it is widely accepted that the practice of Implant Dentistry, at least within the
framework of general dental practice, does not require a 2 or 3-year full-time specialist training.
Nevertheless, there is still a scarcity of academic or university-lead, quality-assured curricula
catering for the needs of practitioners who are not able to attend full-time university degrees.
An interesting observation was that the rates for supervised prosthetic training for Cluster 4-
YPG-U did not seem commensurate with that of surgical supervision. While information of the
particular postgraduate specialty was not evaluated, this finding may reflect that greater
emphasis on the surgical as compared to the restorative aspects of implant dentistry. This
cluster had the highest rate of recommending formal mentorship and university education,
possibly indicating speciality training is accompanied by an awareness of complex clinical
issues.

Several interesting and somewhat concerning perspectives arose from the observations
regarding the Cluster 5-RG representing recent graduates. These are dentists who placed their
first implant on average within 0-2 years after graduation, with high rates of reporting ‘no
information’ about the setting of their first implant. They were likely to have had ‘other’, non-
university based further training, lack of a formal mentor and further training and also reported
lower rates of supervision as compared to the two ‘Postgraduate trained’ clusters 2-MPG-P and
4-YPG-U. Surprisingly, this cluster reported the highest frequency of “No complications” after
the first implant placement, which was higher than that reported by the ‘post-graduate trained’
clusters, suggesting more positive self-appraisal. As discussed above in relation to the
“improper implant positioning” this finding strengthens the argument that complications may
go unnoticed unless a structured education has preceded implant placement. Implant dentistry
education has been divided into 4 levels, level 1 and 2 refer to pre-clinical and basic clinical
training (12). Implant education may be availed as unstructured, short-duration, ‘product
training’(64), reflecting the level 1 or 2. A previous survey found most implant courses fell in

This article is protected by copyright. All rights reserved.

 this category (20) and the present study indicates such settings may be a common mode of
education for the younger ‘GP clusters’ 3-Y-GP and 5-RG. Didactic teaching was reported with
similar frequency across clusters, not surprisingly, as it is the most common implant education
aspect provided by undergraduate curricula, with few programs including clinical implant
Accepted Article
training (7, 36, 65, 66). Yet, Cluster 5-RG rarely reported procedure duration of ‘less than 1
hour’, suggesting lower surgical competence. Both ‘younger G.P’ clusters also had more
problems in ‘flap design and closure’ as compared to the older and the post-graduate trained
clusters, and higher rates of ‘inadequate preparedness’. These could reflect pre-doctoral level
didactic education as insufficient in producing clinical implant-related competencies. Moreover,
the impact of didactic continuing professional education alone in changing actual medical
practice has been inconsistent at best (67, 68). Non-academic setting based education is likely
to offer fewer opportunities for deliberate practice of surgical skills essential for competence
development (69, 70, 71), and adequate supervision to support higher level implant education.
Domain experience is required to develop cognitive skills towards clinical reasoning (60, 61).
Cluster 5-RG had lower advocation of ‘university education’, ‘CE’, and ‘formal mentorship’.
Taken together, these findings may reflect lower comprehension of implant-related
competencies among novice dentists, likely to experience their ‘first implant’ procedure in
unstructured, short-course settings. It must be emphasised that along with structured
education, one-on-one mentorship is essential for an individual to further develop their self-
assessment ability and build the competence necessary to recognize their technical limitations
or complications.

Currently, general dental practitioners place at least as many implants as specialists (20) and
absolute numbers of implant practitioners is growing exponentially, especially in emerging
markets, which were correspondingly over-represented in the 3 ‘younger’ clusters. Despite the
inherent limitations of a non-random and non-uniform demographic sampling, these
exploratory findings may be viewed as a birds-eye view of real-world scenarios. The patterns
unraveled suggest a great need exists for the implementation of competency-based, quality-
assured, practice-directed, comprehensive, and standardized educational opportunities at the
global scale. In addition, frameworks to regulate industry involvement in continuing
professional development are as essential in dentistry as in medicine.

Conclusion: The present exploratory analysis of a global survey indicated that educational
experience relevant to the first implant placement is variable and can be related to user
demographic factors of which time in practice, age, type of practice and postgraduate training
are of major importance.

● Clusters characterised by general practice and older age relied frequently on non-
university implant education but started the practice of implant dentistry after many
years of clinical experience.
● Unlike older practitioners, however, a cluster of recently graduated dentists exists who
initiated implant practice within 0-2 years of graduation, was linked to less frequent
supervision and lower competence-indicators, while acknowledging they were

This article is protected by copyright. All rights reserved.

 inadequately prepared for implant practice. Globally, this demographic appears to be
more prevalent in Asia and Africa and is the one most likely to benefit from structured,
academic, and quality assured implant education pathways.
● The least experienced and least educated dentists being surprisingly the ones who
Accepted Article
reported fewer complications. Furthermore, among young dentists, proper implant
positioning was a challenge most frequently identified by postgraduate educated
dentists, but significantly less frequently by non-specialists.
● Time in practice might be a significant determinant to differentiate educational
approaches when targeting dentists who have not practiced implant dentistry.
● Dentists who were educated in implant dentistry in university settings are either
specialists or graduates of 2-year degree programmes. There is an evident lack of
university involvement in any other education pathways.
● Structured education involving university settings appears to be strongly recommended
by those who have acquired their implant skills in this pathway.

Acknowledgements: The authors have nothing to disclose. The authors would like to
acknowledge, the Faculty of Dentistry, The University of Hong Kong, the ‘Technology-Enriched
Learning Initiative’, The University of Hong Kong as well as Tufts University School of Dental
Medicine, USA for their support in this study. The authors would also like to acknowledge Dr.
Juliane v. Hoyningen-Huene, Dr. Magdalena Wilczak and the Young Dentists Worldwide for the
significant help and support.

References:

1. Rogers EM. A Prospective and Retrospective Look at the Diffusion Model. J Health

Commun 2004: 9: suppl (1):13–9.

2. Wilson CB. Adoption of new surgical technology. BMJ 2006: 332: 112–114.

3. Kim Y, Park JY, Park, SY, et al. Economic Evaluation of Single-Tooth Replacement: Dental

Implant Versus Fixed Partial Denture. Int J Oral Maxillofac Implants 2014: 29: 600–607.

4. Rogers EM. Diffusion of Innovations (4th Edition). New York, NY: THE FREE PRESS,

Simon and Schuster, 2010.

This article is protected by copyright. All rights reserved.

 5. Holden RJ, Karsh BT. The technology acceptance model: its past and its future in health

care. J Biomed Inform 2010: 43:159-72.

Accepted Article
6. Packer C, Simpson S, Stevens A; EuroScan: the European Information Network on New

and Changing Health Technologies. International diffusion of new health technologies: a

ten-country analysis of six health technologies. Int J Technol Assess Health Care 2006:

22:419-28.

7. Mattheos N, Albrektsson T, Buser D, et al. 1st European Consensus Workshop in Implant

Dentistry University Education. Teaching and assessment of implant dentistry in

undergraduate and postgraduate education: a European consensus. Eur J Dent Educ

2009: 13: Suppl (1):11-7.

8. De Bruyn H, Koole S, Mattheos N, Lang NP. A survey on undergraduate implant dentistry

education in Europe. Eur J Dent Educ 2009: 13: Suppl (1):3-9.

9. Mattheos N, de Bruyn H, Hultin M, et al. Developing implant dentistry education in

Europe: the continuum from

undergraduate to postgraduate education and continuing professional development. Eur

J Dent Educ 2014:18: Suppl (1):3-10.

10. Ucer TC, Botticelli D, Stavropoulos A, Mattheos N. Current trends and status of

continuing professional development in implant dentistry in Europe. Eur J Dent Educ

2014: 18: Suppl (1):52-9.

11. Jahangiri L. Implant dentistry in predoctoral education: where are we? J Dent Educ 2015:

79:239-40.

12. Donos N, Mardas N, Buser D; 1st European Consensus Workshop in Implant Dentistry

University Education. An outline of competencies and the appropriate postgraduate

educational pathways in implant dentistry. Eur J Dent Educ 2009:13: Suppl (1):45-54.

13. Swan JA, Newell S. The role of professional associations in technology diffusion. Organ

Stud 1995:16:847-74.

This article is protected by copyright. All rights reserved.

 14. Frenk J, Chen L, Bhutta ZA, et al. Health professionals for a new century: transforming

education to strengthen health systems in an interdependent world. Lancet 2010:

Accepted Article
376:1923-58.

15. Walsh K. Medical education, cost and policy: what are the drivers for change?

Commentary. Ann Ist Super Sanita 2014: 50:205-6.

16. McCann D. Implantology. Does it merit specialty status? J Am Dent Assoc 1990: 121:322-

3.

17. Bhagania M. Implantology: is it the end of the road for dental specialties? J Oral

Maxillofac Surg 2009:67:1575.

18. Mattheos N, Wismeijer D, Shapira L. Implant dentistry in postgraduate university

education. Present conditions, potential, limitations and future trends. Eur J Dent Educ

2014:18: Suppl (1):24-32.

19. Cassel CK, Reuben DB. Specialization, subspecialization, and subsubspecialization in

internal medicine. N Engl J Med 2011:364:1169-73.

20. Vasak C, Fiederer R, Watzek G. Current state of training for implant dentistry in Europe:

a questionnaire-based survey. Clin Oral Implants Res 2007:18:668.

21. Fu Y, Ling J, Jang B, Yin H. Perspectives on dental education in mainland China. Int Dent J

2006:56:265-71.

22. Vundavalli S. Dental manpower planning in India: current scenario and future

projections for the year 2020. Int Dent J. 2014:64:62-7.

23. Balasubramanian M, Brennan DS, Spencer AJ, Short SD. The international migration of

dentists: directions for research and policy. Community Dent Oral Epidemiol 2016:

44:301-12.

24. van Noort R. The future of dental devices is digital. Dent Mater 2012:28:3-12.

25. Rothman DJ, McDonald WJ, Berkowitz CD, et al. Professional medical associations and

their relationships with industry: a proposal for controlling conflict of interest. JAMA

2009:301:1367-72.

This article is protected by copyright. All rights reserved.

 26. Santos GN, Leite AF, Figueiredo PT, et al. Effectiveness of E-Learning in Oral Radiology

Education: A Systematic Review. J Dent Educ 2016: 80:1126-39.

Accepted Article
27. Henry RK, Molnar A, Henry JC. A survey of US dental practices' use of social media. J

Contemp Dent Pract 2012:13:137-41.

28. Ho A, McGrath C, Mattheos N. Social media patient testimonials in implant dentistry:

information or misinformation? Clin Oral Implants Res 2017: 28:791-800.

29. Maloney S, Tunnecliff J, Morgan P, et al. Continuing Professional Development via Social

Media or Conference Attendance: A Cost Analysis. JMIR Med Educ. 2017: 3:e5.

30. Lê S, Josse J, Husson F. FactoMineR: An R Package for Multivariate Analysis. J Stat Softw

2008: 25: 1-18.

31. Husson F, Josse J, Pages J. Principal component methods-hierarchical clustering-

partitional clustering: why would we need to choose for visualizing data. Technical

Report-Agrocampus. A. C. ouest, Applied mathematics department, Rennes, 2010: 1-17.

32. Esfandiari S, Majdzadeh R, Feine J. Types of Canadian dentists who are more likely to

provide dental implant treatment. Implant Dent 2011:20:76-84.

33. Addy LD, Lynch CD, Locke M, Watts A, Gilmour AS. The teaching of implant dentistry in

undergraduate dental schools in the United Kingdom and Ireland. Br Dent J

2008:205:609-14.

34. Mattheos N, Ivanovski S, Heitz-Mayfield L, Klineberg I, Sambrook P, Scholz S. University

teaching of implant dentistry: guidelines for education of dental undergraduate students

and general dental practitioners. An Australian consensus document. Aust Dent J 2010:

55:329-32.

35. Guo YN, Dudley JE, Logan RM, Richards LC. Implant dentistry in Australia: the present

and future. A survey of Australian dentists and specialists. Aust Dent J 2017:62:500-509.

36. Kihara H, Sun J, Sakai M, Nagai S, Da Silva J. A Survey of Dental Implant Instruction in

Predoctoral Dental Curricula in North America. J Dent Educ 2017:81:1085-1090.

37. Kolte AP. Restructuring of dental implant education in India. JICDRO 2017:9:4.

This article is protected by copyright. All rights reserved.

 38. Romero C, Ventura S. Educational data mining: a review of the state of the art. IEEE

Trans Syst Man Cybern C Appl Rev 2010:40:601-18.

Accepted Article
39. Sourial N, Wolfson C, Zhu B,et al. Correspondence analysis is a useful tool to uncover the

relationships among categorical variables. J Clin Epidemiol. 2010:63:638-46.

40. Costa PS, Santos NC, Cunha P, Cotter J, Sousa N. The Use of Multiple Correspondence

Analysis to Explore Associations between Categories of Qualitative Variables in Healthy

Ageing. J Aging Res 2013:2013:302163.

41. Endedijk MD, Vermunt JD, Verloop N, Brekelmans M. The nature of student teachers'

regulation of learning in teacher education. Br J Educ Psychol. 2012: 82:469-91.

42. Dubes R, Jain AK. Clustering methodologies in exploratory data analysis. Advances in

computers. Elsevier, 1980: 19:113-228.

43. Bragg D, Treat R, Simpson DE. Have clinical teaching effectiveness ratings changed with

the Medical College of Wisconsin's entry into the health care marketplace? Acad Med

2000:75:(10 Suppl):S59-61.

44. Hynes H, Stoyanov S, Drachsler H, et al. Designing Learning Outcomes for Handoff

Teaching of

Medical Students Using Group Concept Mapping: Findings From a Multicountry

European Study. Acad Med 2015:90:988-94.

45. Crenshaw K, Shewchuk RM, Qu H, et al. What should we include in a cultural competence

curriculum? An

emerging formative evaluation process to foster curriculum development. Acad Med

2011:86:333-41.

46. Sicotte C, Taylor L, Tamblyn R. Predicting the use of electronic prescribing among early

adopters in primary care. Can Fam Physician 2013:59:e312-21.

47. Baltar F, Brunet I. Social research 2.0: virtual snowball sampling method using

Facebook. Internet Res 2012:22:57-74.

This article is protected by copyright. All rights reserved.

 48. Geurts M, O'Neill MF, Albaum G, Lawrence K. Increasing response rates using an

inducement question in mail surveys. J Direct Marketing 1988:2:35-42.

Accepted Article
49. Norman AT, Russell CA. The pass‐along effect: Investigating word‐of‐mouth effects on

online survey procedures. J Comput Mediat Commun 2006:11:1085-103.

50. Kirchherr J, Charles K. Enhancing the sample diversity of snowball samples:

Recommendations from a research project on anti-dam movements in Southeast Asia.

PloS one. 2018: 13:e0201710.

51. Saliba NA, Moimaz SA, Garbin CA, Diniz DG. Dentistry in Brazil: its history and current

trends. J Dent Educ 2009:73:225-31.

52. Kandelman D, Arpin S, Baez RJ, Baehni PC, Petersen PE. Oral health care systems in

developing and developed countries. Periodontol 2000 2012:60:98-109.

53. Knevel R, Gussy MG, Farmer J. Exploratory scoping of the literature on factors that

influence oral health workforce planning and management in developing countries. Int J

Dent Hyg 2017:15:95-105.

54. Lien AS, Jiang YD. Integration of diffusion of innovation theory into diabetes care. J

Diabetes Investig 2017:8:259-260.

55. Bandura A. Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev

1977:84:191-215.

56. Arnold HJ. Task performance, perceived competence, and attributed causes of

performance as determinants of intrinsic motivation. Acad Manage J 1985:28:876-88.

57. Koole S, Vandeweghe S, Mattheos N, De Bruyn H. Implant dentistry education in Europe:

5 years after the Association for Dental Education in Europe consensus report. Eur J

Dent Educ 2014:18: Suppl (1):43-51.

58. Payant L, Williams JE, Zwemer JD. Survey of dental implant practice. J Oral Implantol

1994:20: 50-58.

59. Carter JB, Stone JD, Clark RS, Mercer JE. Applications of Cone-Beam Computed

Tomography in Oral and Maxillofacial Surgery: An Overview of Published Indications

This article is protected by copyright. All rights reserved.

 and Clinical Usage in United States Academic Centers and Oral and Maxillofacial Surgery

Practices. J Oral Maxillofac Surg 2016:74:668-79.

Accepted Article
60. Crespo KE, Torres JE, Recio ME. Reasoning process characteristics in the diagnostic

skills of beginner, competent, and expert dentists. J Dent Educ 2004:68:1235-44.

61. Maupomé G, Schrader S, Mannan S, Garetto L, Eggertsson H. Diagnostic thinking and

information used in clinical decision-making: a qualitative study of expert and student

dental clinicians. BMC Oral Health 2010:13:10:11.

62. Moore MJ, Bennett CL. The learning curve for laparoscopic cholecystectomy. The

Southern Surgeons Club. Am J Surg. 1995:170:55-9.

63. Schleyer T, Eaton KA, Mock D, Barac'h V. Comparison of dental licensure, specialization

and continuing education in five countries. Eur J Dent Educ 2002:6:153-61.

64. Ivanovski S, Mattheos N, Scholz S, Heitz-Mayfield L. University postgraduate training in

implant dentistry for the general dental practitioner. Aust Dent J 2010:55:339-45.

65. Atashrazm P, Vallaie N, Rahnema R, Ansari H, Shahab MP. Worldwide Predoctoral Dental

Implant Curriculum Survey. J Dent (Tehran) 2011:8:12-18.

66. Fijnheer C, Langhorst FR, Wismeijer D. Evaluation of the implant-related restorative

undergraduate programme at ACTA, the Netherlands. Part I: students' satisfaction. Eur J

Dent Educ 2016:20:84-93.

67. Davis D, O'Brien MA, Freemantle N, Wolf FM, Mazmanian P, Taylor-Vaisey A. Impact of

formal continuing medical education: do conferences, workshops, rounds, and other

traditional continuing education activities change physician behavior or health care

outcomes? JAMA 1999:282:867-74.

68. Zeiger RF. Toward continuous medical education. J Gen Intern Med 2005: 20:91-4.

69. Kirkman MA. Deliberate practice, domain-specific expertise, and implications for

surgical education in current climes. J Surg Educ 2013: 70:309-17.

70. Ericsson KA. Deliberate practice and acquisition of expert performance: a general

overview. Acad Emerg Med 2008:15:988-94.

This article is protected by copyright. All rights reserved.

 71. Ericsson KA. Acquisition and maintenance of medical expertise: a perspective from the

expert-performance approach with deliberate practice. Acad Med 2015:90:1471-86.

Accepted Article

This article is protected by copyright. All rights reserved.

 Table 1: Explanatory factor variables and their categorical levels
Accepted Article
Factor Variable Categorical levels
Age group <25/ 25-30/ 30-35/ 36-45/ >45 (years)

Gender Male/Female

Continent where located Africa/ Asia/ Australasia/ Europe/ North America/ South
America/ No Information

Work Experience 0-3/ 4-9/ >10 (years)

Professional Setting Private/ Public/ University

Time of first implant 0-2/3-5/>10 (years after graduation)

Setting of First implant At C.E course/ At Private practice/ At University/ Other/

No information

Practice Type G.P./ Specialty/ Post-Graduate Student

Post-Graduate Training No PG Training obtained / PG Training obtained

This article is protected by copyright. All rights reserved.

 Table 2: Percentage of responders within each cluster for category levels significantly
over-represented in that cluster*
Cluster 1 (n=123) % of Cluster 1 Cluster 2 (n=246) % of Cluster 2
Accepted Article
Age>45yrs 73.17

Time.of.First.implant=>10yrs 78.05 Age.Group=36-45yrs 87.40

Work.Experience=>10yrs 93.50 Work.Experience=>10yrs 93.50

Continent=North.America 17.07 Time.of.First.implant=6-10yrs 37.80

Practice.Type=G.P. 78.86 Setting.of.first.implant=Private 62.60

Professional.Setting=Private 81.30 Continent=Europe 39.84

Continent=Australasia 8.94 Professional.Setting=Private 80.08

Age.Group=No.information 2.44 Post.Graduate.Training=Yes 70.73

Setting.of.first.implant=Private 52.03 Setting.of.first.implant=Other 8.54

Setting.of.first.implant=at.CE 17.07 Practice.Type=G.P. 60.57

Gender=Male 74.39

Cluster 3 (n=233) % of Cluster 3

Cluster 4 (n=200) % of Cluster 4

Work.Experience=3-9yrs 87.12

Time.of.First.implant=3-5yrs 52.79 Practice.Type=Specialty 85.00

Age.Group=25-30yrs 54.51 Setting.of.first.implant=University 78.50

Age.Group=30-35yrs 40.77 Post.Graduate.Training=Yes 96.50

Professional.Setting=Private 80.26 Professional.Setting=University 49.00

Practice.Type=G.P. 67.81 Age.Group=30-35yrs 44.00

Setting.of.first.implant=Private 52.79 Work.Experience=3-9yrs 54.50

Post.Graduate.Training=No 54.94 Continent=Asia 71.50

Setting.of.first.implant=CE 17.60 Professional.Setting=Public 27.00

Practice.Type=Post-Graduate.Student 15.88 Continent=South.America 8.50

Gender=Female 37.00

Cluster 5 (n=213) % of Cluster 5 Continent=No.information 1.00

Work.Experience=0-3yrs 96.71

Time.of.First.implant=0-2yrs 93.43

Age.Group<25yrs 35.68

Post.Graduate.Training=No 82.16

Age.Group=25-30yrs 57.75

Setting.of.first.implant=No.Information 25.35

Professional.Setting=University 35.21

This article is protected by copyright. All rights reserved.

 Practice.Type=Post.Graduate.Student 20.66

Professional.Setting=No.information 4.23

Continent=Africa 14.55
Accepted Article
Practice.Type=G.P. 63.38

Gender=Female 38.50

Continent=No.information 0.94

* p-value <0.05 and positive v test value, category levels are ranked according to v test values which represent their
importance in clustering

This article is protected by copyright. All rights reserved.

ccepted Articl
Table 3: Profiles of the top 5 representative cluster members for Clusters 1-5 based on distance from the center of each cluster*

Cluster 1
Gender Age.Group Continent Work.Experience Professional.Setting Practice.Type Post.Graduate.Training Setting.of.first.implant Time.of.First.implant
Male >45 Asia >10yrs Private Post-Graduate Student No PG training at.Private >10
Male >45 Asia >10yrs Private G.P. PG training at.Private >10
Male >45 Asia >10yrs Private G.P. PG training at.Private >10
Male >45 Asia >10yrs Private G.P. PG training at.Private >10
Male >45 Asia >10yrs Private G.P. PG training at.Private >10

Cluster 2
Gender Age.Group Continent Work.Experience Professional.Setting Practice.Type Post.Graduate.Training Setting.of.first.implant Time.of.First.implant
Male 36-45 Asia >10yrs Private Specialty No PG training at.Private 3-5yrs
Male 36-45 Asia >10yrs Private G.P. PG training at.University 6-10yrs
Male 36-45 Asia >10yrs Private Specialty No PG training at.Private 6-10yrs
Male 36-45 Asia >10yrs Private G.P. PG training at.Private 3-5yrs
Male 36-45 Asia >10yrs Private G.P. PG training at.Private 3-5yrs

Cluster 3
Gender Age.Group Continent Work.Experience Professional.Setting Practice.Type Post.Graduate.Training Setting.of.first.implant Time.of.First.implant
Female 25-30 Europe 3-9yrs Private G.P. PG training at.CE 0-2yrs
Female 25-30 Asia 3-9yrs Private G.P. PG training at.Private 3-5yrs
Female 25-30 Asia 3-9yrs Private G.P. PG training at.Private 3-5yrs
Female 25-30 Asia 3-9yrs Private G.P. PG training at.Private 3-5yrs
Male 30-35 Africa 3-9yrs Private G.P. No PG training at.University 6-10yrs

This article is protected by copyright. All rights reserved.

ccepted Articl
Cluster 4
Gender Age.Group Continent Work.Experience Professional.Setting Practice.Type Post.Graduate.Training Setting.of.first.implant Time.of.First.implant

Male 36-45 Africa 3-9yrs Public Specialty PG training at.University 0-2yrs

Male 36-45 Africa 3-9yrs Public Specialty PG training at.University 0-2yrs

Male 25-30 Asia 3-9yrs University Specialty PG training at.University 6-10yrs

Male 30-35 Africa >10yrs University G.P. PG training at.University 6-10yrs

Female 30-35 Asia 3-9yrs Private Specialty PG training at.University 0-2yrs

Cluster 5
Gender Age.Group Continent Work.Experience Professional.Setting Practice.Type Post.Graduate.Training Setting.of.first.implant Time.of.First.implant

Female 25-30 Asia 0-3yrs University G.P. No PG training No.Information 0-2yrs

Female 25-30 Asia 0-3yrs University G.P. No PG training at.University 0-2yrs

Male 25-30 Asia 0-3yrs University G.P. No PG training at.University 0-2yrs

Male 25-30 Asia 0-3yrs University G.P. No PG training at.University 0-2yrs

Male 25-30 Europe 0-3yrs Public G.P. No PG training at.University 0-2yrs

* Cluster members are ordered in increasing order of distance from the center of the cluster

This article is protected by copyright. All rights reserved.

Accepted Article

This article is protected by copyright. All rights reserved.

Accepted Article

This article is protected by copyright. All rights reserved.

Accepted Article

This article is protected by copyright. All rights reserved.

Accepted Article

This article is protected by copyright. All rights reserved.

Accepted Article

This article is protected by copyright. All rights reserved.

Accepted Article

This article is protected by copyright. All rights reserved.

Accepted Article

This article is protected by copyright. All rights reserved.