Professional Documents
Culture Documents
Eje 12426
Eje 12426
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
Authors’ affiliations:
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.
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
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).
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).
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
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
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”.
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.
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.
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.
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.
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.
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.
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.
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.
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).
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.
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%).
‘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.
‘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.
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).
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%).
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.
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.
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
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
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
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
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
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
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,
ten-country analysis of six health technologies. Int J Technol Assess Health Care 2006:
22:419-28.
10. Ucer TC, Botticelli D, Stavropoulos A, Mattheos N. Current trends and status of
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
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.
15. Walsh K. Medical education, cost and policy: what are the drivers for change?
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
education. Present conditions, potential, limitations and future trends. Eur J Dent Educ
20. Vasak C, Fiederer R, Watzek G. Current state of training for implant dentistry in Europe:
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
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.
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
partitional clustering: why would we need to choose for visualizing data. Technical
32. Esfandiari S, Majdzadeh R, Feine J. Types of Canadian dentists who are more likely to
33. Addy LD, Lynch CD, Locke M, Watts A, Gilmour AS. The teaching of implant dentistry in
2008:205:609-14.
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
37. Kolte AP. Restructuring of dental implant education in India. JICDRO 2017:9:4.
40. Costa PS, Santos NC, Cunha P, Cotter J, Sousa N. The Use of Multiple Correspondence
41. Endedijk MD, Vermunt JD, Verloop N, Brekelmans M. The nature of student teachers'
42. Dubes R, Jain AK. Clustering methodologies in exploratory data analysis. Advances in
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
45. Crenshaw K, Shewchuk RM, Qu H, et al. What should we include in a cultural competence
curriculum? An
2011:86:333-41.
46. Sicotte C, Taylor L, Tamblyn R. Predicting the use of electronic prescribing among early
47. Baltar F, Brunet I. Social research 2.0: virtual snowball sampling method using
51. Saliba NA, Moimaz SA, Garbin CA, Diniz DG. Dentistry in Brazil: its history and current
52. Kandelman D, Arpin S, Baez RJ, Baehni PC, Petersen PE. Oral health care systems in
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
54. Lien AS, Jiang YD. Integration of diffusion of innovation theory into diabetes care. J
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
5 years after the Association for Dental Education in Europe consensus report. Eur J
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
62. Moore MJ, Bennett CL. The learning curve for laparoscopic cholecystectomy. The
63. Schleyer T, Eaton KA, Mock D, Barac'h V. Comparison of dental licensure, specialization
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
67. Davis D, O'Brien MA, Freemantle N, Wolf FM, Mazmanian P, Taylor-Vaisey A. Impact of
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
70. Ericsson KA. Deliberate practice and acquisition of expert performance: a general
Gender Male/Female
Continent where located Africa/ Asia/ Australasia/ Europe/ North America/ South
America/ No Information
Gender=Male 74.39
Work.Experience=3-9yrs 87.12
Gender=Female 37.00
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
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
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
Cluster 5
Gender Age.Group Continent Work.Experience Professional.Setting Practice.Type Post.Graduate.Training Setting.of.first.implant Time.of.First.implant
* Cluster members are ordered in increasing order of distance from the center of the cluster