doi:10.1684/ejd.2021.4010
EJD, vol. 31, n◦2, March-April 2021
161
To cite this article: Kaczmarczyk R, King F, Tizek L, Biedermann T, Zink A. Visualising the past to plan the future: a network analysis of the largest European dermatology
Investigative report
Eur J Dermatol 2021; 31(2): 161-9Robert KACZMARCZYKa Felix KINGa
Linda TIZEK Tilo BIEDERMANN Alexander ZINK
Technical University of Munich, School of Medicine, Department of Dermatology and Allergy, Biedersteiner Strasse 29, 80802 Munich, Germany
Reprints:Robert Kaczmarczyk
Article accepted on 26/07/2020
Visualising the past to plan the future: a network analysis of the largest European dermatology conference
Background: The annual conference of the European Academy of Der-
matology and Venereology is one the largest dermatology conferences worldwide.
Objectives: Network analysis can be used for in-depth insightinto trending topics and underlying trends at the congress.
Materials &Methods:
Network analysis was employed to assess the entirety of the submitted abstracts to the congress in 2019. The data were processed, analysed, and visualised using easy-to-understand network graphs. Top- ics were then compared to their respective global burden (Disease Adjusted Life Years [DALYs]) and the number of respective publica- tions on PubMed in the year 2018.
Results:Overall, 1,280 lecture titles and 1,941 poster titles were included in the final analysis. The most fre- quently used terms were “patients” (n = 473), “treatment” (n = 301), and
“psoriasis” (n = 335). Relative to DALYs, “psoriasis” (+21.9%) among others, was rather over-represented, while “fungal skin diseases” (-7.6%) and “urticaria” (-6.4%) were under-represented. Compared to the rel- ative number of PubMed publications in 2018, “psoriasis” (+20.3%),
“acne” (+7.9%), and “alopecia” (+3.1%) were over-represented, while
“melanoma” (-22.5%), “dermatitis” (-4.2%) and “pruritus” (-3.4%) were rather under-represented.
Conclusion:The network analysis showed that the congress was a patient and therapy-centred event. An explanation for the particular focus on chronic inflammatory skin diseases and melanoma would be the introduction of new therapies at the congress. To delineate trends over time, a longitudinal network analysis including several con- gresses should be conducted and could be used to determine additional topics to be included in future events
Key words:
network analysis, dermatology, scientography, conference, digital medicine, psoriasis, melanoma, DALY
W
hen Europe’s dermatologists congregate annu- ally at the congress of the European Academy of Dermatology and Venereology (EADV), there is so much to see and experience that it is simply not pos- sible for everyone to do so. In 2019, the event was held in Madrid, Spain and attracted around 13,000 visitors from over 100 countries, providing a jaw-dropping number of 1,280 lectures and 1,941 posters, making it one of the largest conferences in dermatology worldwide. The latest develop- ments in scientific research and clinical practice were laid out by most prolific researchers and clinicians in dermatol- ogy. They were shared and discussed in order to influence patient care around the world[1].Because of the sheer volume of large congresses, it is hard to grasp which topics were especially focused on and which, conversely, were rather under-represented.
The focus of worldwide research on a specific topic can
aThese authors contributed equally
be somewhat interpreted by the number of publications available on PubMed, the publication database by the United States National Library of Medicine. For example, in 2018, there were around 20,000 publications focusing on dermatologic diseases such as psoriasis, atopic der- matitis, and urticaria; the highest number corresponding to melanoma-related publications [2]. Another way to interpret the significance and importance of specific diseases is to use Disability-Adjusted Life Years (DALYs), a measure of the overall global disease burden expressed as the number of years lost due to illness, disability, and early death [3, 4]. Skin conditions were responsible for 41.6 million DALYs in 2013, meaning that 1.8% of the global burden of disease was due to a skin condition.
Dermatitis, which causes 9.3 million DALYs, had the highest burden worldwide, followed by acne vulgaris (7.2 million DALYs), urticaria (4.7 million DALYs), and psoriasis (4.7 million DALYs) [3]. Considering Europe, the burden of psoriasis, for example, was higher in Western and Northern Europe compared to Eastern Europe[5].
To process, analyse, and visualise big data such as all titles of a conference with 13,000 visitors, network analysis, an upcoming computational method, could be used to make big data accessible for interpretation. Network analysis is able to map knowledge structures by the co-occurrence of specific words in any data set of any size, which makes it a valuable tool, especially in times of big data and artificial intelligence[6, 7]. This tool has been widely used in social and political sciences[8, 9]. However, in the medical field, it has only scarcely been used aside from a small number of epidemiological publications[10-12].
The objective of this study was to use network analysis for an in-depth insight into topics that were focused on at one of the largest dermatology congresses in Europe and compared them to respective worldwide publications avail- able on PubMed and the respective global burden of disease measured with DALYs.
Materials and methods
Data sources/measurement
The 28th annual congress of the EADV took place from the 9th to the 13th of October 2019 in Madrid, Spain.
The titles from the official program were used to perform a retrospective network analysis. All congress titles were classified into poster and lecture. The data were extracted from the official program pdf file using tabula v1.2.1 open- source software[13]and then pre-processed using custom written python code. The extracted data were first split into titles, countries, cities, and authors. Redundant titles in lecture sessions were removed to prevent an artificial increase in word count. The data were then cleaned to exclude filling words such as “and”, “so”, and “or” and standardisede.g. “acne inversa” and “hidradenitis suppura- tiva”. Disease entities with names comprising more than one word were split (e.g. “atopic dermatitis” into “atopic” and
“dermatitis”).
Software
The networks were composed using open-source software Gephi v0.9.2 [14] and a ForceAtlas 2 [15] layout. The terms and their connections were sized proportionally to their counts. To keep the output manageable, filters were applied. Thus, only terms with more than 200 connections were included in the overall network graph (more than 30 connections in the lecture network graph and more than 100 connections in the poster network graph).
Comparison between conference topics and DALYs and PubMed publications
To assess the importance of each skin disease at the con- ference, the count for each skin disease was divided by the total count of all words-related to any skin disease.
Accordingly, proportions were calculated, meaning that a skin disease with a high proportion would receive a high level of attention. Using the same method, the impor- tance of a skin disease was calculated based on DALYs in 2013 [3] and the number of publications available
on PubMed between January 1st, 2018 and December 31st, 2018 [2]. Then, all relative values were compared to assess whether there was a relative over- or under- representation at the congress. A negative percentage indicates under-representation, whereas a positive percent- age reflects relative over-representation.
Statistics
Descriptive data were generated for posters and lectures.
Additionally, for each skin disease, absolute and relative numbers were calculated. The reciprocal of the network density parameter was used to assess the congress diversity.
Network diversity was conveyed through values from 1, indicating that all abstract titles were the same, and infinite, meaning that none of the words used in the abstract titles were ever used twice.
Statistical analysis was carried out using Python’s “pandas v0.24.2”[16]. Since we extracted the submissions from the pdf file of the official conference program, there was no missing data.
Moreover, we used OpenCage Geocoder library [17] to match the cities with the corresponding coordinates and visualised the locations with Google Maps API on an interactive map. To assess geographic differences, the classification of countries provided by the UN Statistics Division was used[20].
Results
Overall, we analysed a total of 30,898 words in 3,221 congress titles in the official program with an average count of 9.0 words per title.
After cleaning data to exclude filling words, the total word count of lectures was 5,715 words with an average length of 5.1 words per title and of posters 16,712 words with an average length of 8.6 words per title. The terms with the highest proportion relative to the total word count in given categories were “patients” (n =473, 3.7%), “treat- ment” (n =301, 2.8%), and “psoriasis” (n =335, 2.8%) (figure 1A). Considering lectures and posters, the diseases which received most attention at the conference were “pso- riasis” (n =335, 33.4%), “dermatitis” (n =169, 17.8%),
“acne” (n =107, 11.3%), “melanoma” (n =88, 9.3%), and
“urticaria” (n =48, 5.1%).
The 20 most common words used in posters and lectures are presented infigure 1B. Overall, 11 terms were in both top 20, namely “skin”, “treatment”, “diseases”, “patients”, “psoria- sis”, “dermatitis”, “clinical”, “atopic”, “cell”, “melanoma”, and “carcinoma”. However, in posters, the terms “patients”
(n =396, 2.4%), “case” (n =289, 1.7%), and “psoria- sis” (n =267, 1.6%) were the top three words, whereas in lectures “skin” (n =88, 1.5%), “treatment” (n =88, 1.5%), and “diseases” (n =81, 1.4%) were most common (figure 1B).
Network graph visualisation of conference topics
As demonstrated by the visualisation of terms infigure 2A, each with at least 200 connections, some key aspects of
Patients Diagnosis Melanoma Carcinoma Cutaneous New Report Cell Therapy Atopic Dermatology Acne Clinical Study Dermatitis Case Diseases Skin Psoriasis Treatment
Skin Treatment Diseases Patients Psoriasis Acne Dermatology Therapy Dermatitis New Clinical Management Atopic Cell Melanoma Diagnosis Carcinoma Inversa Nail Systemic
Patients Case Psoriasis Treatment Study Report Cutaneous Clinical Skin Dermatitis Diseases Atopic Cell Efficacy Cases Syndrome Carcinoma Results Safety Melanoma
175
Absolute word counts for lecture and poster titles at the EADV
0.0 0.5 1.0 1.5
Relative proportion of word counts within given category*
Lectures Posters
Relative proportion of terms in lecture (n = 5,715 words) and poster (n = 16,712 words) titles
n = 77 n = 396
n = 88 n = 125
n = 81 n = 117
n = 14 n = 289
n = 47 n = 123
n = 21 n = 191 n = 40
n = 61
n = 60 n = 38
n = 49 n = 126
n = 38 n = 97
n = 52 n = 50
n = 36 n = 92
n = 4 n = 175
n = 44 n = 53
n = 10 n = 135
n = 32 n = 70
n = 35 n = 58
n = 32 n = 50
88 88
81 77
68 61
60 52
47 44
40 39 38 36
35 32 32 30
28 27
80 60 40 20 0 0 100 200 300 400
58 59 66 70 85 87 90 92 97
117 123 125 126 135
175 191
213 267
289 396 n = 68
n =88 n =213
n = 267
2.0 2.5 3.0 3.5 4.0
Lectures* [%]
Posters* [%]
A
B
Figure 1.The most common disease terms appearing in all titles of the scientific conference program (A) and in lecture and poster titles separately (B) (disease terms in bold were present in the top 20 categories for both lecture and poster titles).
the conference were the therapy of patients with psoria- sis and its safety and efficacy (words connected in blue colour), atopic dermatitis and its treatment and outcome (green colour), and a rather broad mixture of topics around
“skin diseases”, “case reports”, “melanoma” and different types of “skin carcinoma” (brown colour) (figure 2A).
By including only words used in lecture titles with more than 30 connections, the network graph of lectures revealed several different theme clusters around “therapy”,
“psoriasis” and “patient”, “hidradenitis suppurativa” and
“skin disease” (figure 2B). The network graph of posters with more than 100 connections demonstrated that the network around “psoriasis” and “patient” along with
“treatment” was somewhat comparable between poster and lecture titles; the significant focus on case reports covering a broad field of dermatologic diseases including, for
example “basal cell carcinoma”, “lupus”, “lichen”, “lym- phoma”, “mycosis fungoides”, or “alopecia”, was specific to posters (figure 2C).
Contributing nations
Germany contributed the most lectures (n =157, 12.3%), followed by Spain (n =139; 10.9%), and the Nether- lands (n =102, 8.0%), whereas the highest number of posters came from Spain (n =274, 14.1%), followed by Tunisia (n =140, 7.2%), and Germany (n =99, 5.1%) (figure 3A). Comparing the top 20 contributing nations for lectures and posters, the data show that while only three contributing nations for lectures were non-European (US, China, and Canada), half of the poster-contributing nations were non-European (figure 3A). Visualising all
Disorders
Lasers
Acne Inversa
Allergic Based
Center Effectiveness
Presentation Pathogenesis
Light
Cohort
Cosmetic
Outcomes
Multiple
Rare
Induced Extract
Findings Oral
Lesions
Pemphigus Manage
Onychomycosis
Syndrome
Human Immunodeficiency Virus Actinic
Differential Autoimmune Novel
Plaque Psoriasis
Know
Melanoma
Acquired
Improvement Recurrent
Life
Vasculitis
Versus
Pregnancy
Type
Characteristics
Breast Controlled Cream
Drug
Mimicking Data
Squamous Cell Carcinoma
Experience
Lichen Planus reSURFACE
Early
Children
Surgical
Cutaneous
Effects Randomized
Expression Biologic
Tildrakizumab
Guselkumab
Treated
Omalizumab
History
Serum
Inhibitors Therapies
Facial
First Levels
Case Series
Vascular
Survey
Testing
Genital
Assessment
Efficacy
Tumor
Bullous Pigmented
Involvement Using
Leishmaniasis
During
Prospective
Atypical
Medical
Chronic Inflammatory
Disseminated Metastatic
Systemic
Weeks
Two
Face
Systematic Comparative
Evaluation Years
Factors
Laser
Eruption Psoriatic
Survival Follow-Up Without
Bullous Pemphigoid Epidemiological
Impact
Warts
Metabolic
Treatment
Basal Cell Carcinoma Recalcitrant Comparison
Phase 3
Psoriatic Arthritis
Case Report
Comorbidities Dermatosis
Acid
Dermoscopic
Eczema Severity Pediatric
Diagnostic Acne
Manifestations Cancer
Real Successful
Due Body
Nail
Daily
Challenging Syphilis
Mohs Keratosis
Carcinoma Successfully
Common Pruritus
Psoriasis
Health
Registry
Gene Pustular
Care
Large
Case
Disease
World
Aesthetic Manifestation
Histopathological Analysis
Combined
Congenital European
About
Erythema Leprosy
Biologics
Atopic Dermatitis
Vivo
Report
Cells
Clinical
Profile
Population
Among
Diffuse Photodynamic Superficial
Dermatological
Contact Dermatitis Tertiary
Approach Containing
High Alopecia Areata
Dermatologist
Should Adult
Chronic Spontaneous Urticaria
Review Quality
Apremilast
Caused
Treatments
Development Women
Cross-Sectional
Retrospective
Update
Hospital
Diagnosis
Hair Herpes Zoster
Week
Correlation
Hand
Role Discussion
Patch
Symptoms
Multicenter International
Lymphoma Dermatitis
Different Human
Cutis
Advanced Practice
Primary Between
Changes Trials
Old
Dermoscopy
Effective
Ulcers
Therapy
Observational
Diseases
Alopecia
Moderate-To-Severe
Patient Confocal
Society
Drugs
Genetic
Following
Results
Botulinum
Patients
Secondary
Response
Dose Nevus
Vitiligo Evidence
Dupilumab Ixekizumab
Cases
Study
Generalized
Skin
Safety
Series
Male Burden Trial
Unusual Association
Prevalence
Features Tolerance
Injection
Occupational Infections
Rosacea Real-World
Associated
Loss Family
Ustekinumab
Dermatomyositis Infection
Therapeutic
You Mycosis Fungoides
Year Index
Transplant
Linear
Surgery Adverse
Adalimumab
Germany Secukinumab
Cell Acute
Epidermal
Scalp
Severe
Granuloma
Combination
Topical Effect Lower
Morphea Palmoplantar
Dermatology
Risk
Microscopy
Dermatologists Urticaria Spectrum Phototherapy
Lupus Erythematosus
Literature Setting
Reactions Moderate
Management
Disorders Lasers
Acne Inversa
Workshop Based
Pathogenesis Light
European
About
Cosmetic
Outcomes Biologics
Atopic Dermatitis Clinical
Oral
Prevention Lesions
Photodynamic Pemphigus
Manage
Onychomycosis
Human Immunodeficiency Virus Actinic
Approach
Differential Than
Wound
Techniques
Autoimmune
Should
Novel
Ageing
Know
Melanoma
Chronic Spontaneous Urticaria
Review Quality Treatments
Important
Life
Vasculitis
Versus
Pregnancy
Update
Drug
Squamous Cell Carcinoma
Diagnosis
Hair Experience
Devices
Hand
Role Early Discussion
Paediatric
Children International
Surgical
Cutaneous Dermatitis
Advanced Practice
Biologic
Indications
Between Treated
Life-Threatening Dermatopathology
Inhibitors
Dermoscopy
Effective Therapies
Ulcers
Therapy
Facial First
Diseases
AlopeciaGuidelines
Confocal Case Series
Vascular Society
Drugs
Genetic
Results
Tips Interactive
Practical
Botulinum
Patients
Genital
Microbiome Basic
Efficacy
Bullous Where
Vitiligo Hyperhidrosis
Fillers
During
Cases
Atypical
Medical
Study
Chronic
Inflammatory
Including
Skin
Safety
No
Metastatic
Systemic
Histopathology
Trial
Experts Face
Healing
Prevalence
Scar
Yes Factors
Laser
XP
Pathophysiology
Psoriatic
Injection
Occupational
Infections Rosacea
Associated Impact
Therapeutic
Treatment
You Basal Cell Carcinoma
Communications
Place
Viral Eczema
Surgery
Scars
Adalimumab
Diagnostic
Reconstruction
Acne
Cancer
Real
Cell
Mucosal Body
Acute
Severe
Nail
Daily
Topical
Biomarkers
Mohs
Keratosis Lower
Dermatology
Non-melanoma Risk
Pruritus
Psoriasis
Registry
Microscopy
Photoageing
Care
Spectrum
Phototherapy
Chinese
Disease Aesthetic
Human Papilloma Virus Reactions
Moderate
Management
Toxin
A
B
Disorders
Acne Inversa
Allergic Based
Center Effectiveness
Presentation Light
Cohort Outcomes
Multiple
Rare
Induced Extract
Findings
Oral
Lesions
Pemphigus
Syndrome
Novel
Plaque Psoriasis
Melanoma
Acquired
Improvement
Recurrent
Life Versus
Type
Characteristics
Breast Controlled
Cream
Drug
Mimicking Data
Single
Squamous Cell Carcinoma Experience
Lichen Planus
reSURFACE
Early
Children
Cutaneous
Effects Randomized
Expression Tildrakizumab
Guselkumab
Value
Treated
Omalizumab
Serum
Facial
First Levels
Case Series
Child
Survey Assessment
Efficacy
Tumor
Bullous Pigmented
Involvement Using
Leishmaniasis During
Prospective
Atypical Chronic
Inflammatory
Disseminated
Metastatic Systemic
Weeks
Two
Tuberculosis
Face
Systematic Comparative Evaluation
Years
Factors Laser
Eruption
Follow-Up
Clinicopathological
Without
Bullous Pemphigoid Epidemiological
Impact
Warts
Metabolic
Treatment
Basal Cell Carcinoma
Recalcitrant
Comparison
Phase 3
Psoriatic Arthritis
Case Report
Dermatosis Acid
Dermoscopic
Eczema
Severity
Revealing
Pediatric
Diagnostic
Acne
Manifestations
Cancer Real
Successful
Due Body
Nail
Challenging Syphilis
Reaction Carcinoma
Successfully
Subcutaneous
Common
Pruritus
Psoriasis
Health
Registry
Gene
Pustular
Care
Large
Case
Disease
World
Manifestation
Histopathological
Analysis
Congenital About
Erythema
Leprosy
Atopic Dermatitis
Vivo
Report
Cells
Clinical
Profile
Population Among
Diffuse Photodynamic
Superficial
Dermatological
Contact Dermatitis
Tertiary
Approach Containing
High
Alopecia Areata Adult Chronic Spontaneous Urticaria
Review
Quality
Apremilast
Caused
Treatments Women
Cross-Sectional
Retrospective
Hospital
Healthy
Diagnosis
Hair
Herpes Zoster Week
Correlation
Role Lesion
Patch
Symptoms
Lymphoma
Dermatitis
Different Human
Cutis Practice
Primary Between
Changes Trials
Old
Dermoscopy Therapy
Observational
Diseases Alopecia
Moderate-To-Severe
Patient
Confocal
Following
Results
Patients
SecondaryResponse
Nevus Vitiligo
Dupilumab
Ixekizumab
Cases
Study
Generalized
Skin
Safety
Series Male
Burden Trial
Unusual Association
Female
Prevalence
Features Tolerance
Rosacea
Real-World
Associated
Loss
Family Ustekinumab
Dermatomyositis Infection
Therapeutic
Mycosis Fungoides
Year Index
Linear Surgery
Adverse Adalimumab
Germany
Secukinumab
Cell
Acute Epidermal
Scalp
Severe
Granuloma Combination
Topical
Effect
Morphea Palmoplantar
Dermatology
Risk Microscopy
Dermatologists Phototherapy
Lupus Erythematosus Literature
Setting
Reactions Moderate
Management
C
Figure 2.Network graph visualisation of disease terms in all titles of the dermatology congress program with at least 200 connections (A), in lecture titles with more than 30 connections (B), and in poster titles with more than 100 connections (C).
conference contributions on a world map, this showed several hotspots of contributors in Central and Southern Europe, for example, around Madrid. Outside Europe, hotspots included the US east coast, Southern Amer- ica, Japan, and even Australia. An interactive version of the map showing the global contribution of authors can be accessed by using the link, https://nwgraphs.web.app (figure 3B).
Comparison with DALYs and PubMed
The greatest relative discrepancies between confer- ence titles and DALYs were found for “psoriasis”
(+21.9%), “melanoma” (+5.4%), “alopecia” (+3.6%),
“pruritus” (+3.3%), and “keratinocyte carcinoma”
(+2.4%), which were over-represented. “Fungal skin diseases” (-7.6%), “urticaria” (-6.4%), “acne” (-6.3%),
“viral skin diseases” (-5.2%), and “dermatitis” (- 4.9%) were the most under-represented skin diseases (figure 4A, B).
When searching on PubMed for skin diseases represented at the congress, 19,674 publications were found for the year 2018. Of these, the five most common were “melanoma”
(6,265 publications), “dermatitis” (4,323 publications),
“psoriasis” (2,581 publications), “pruritus” (1,337 publi- cations), and “urticaria” (769 publications). The greatest relative positive discrepancies between the conference and PubMed results were detected for “psoriasis”
(+20.3%), “acne” (+7.9%), “alopecia” (+3.1%), “urticaria”
(+1.2%), and “viral skin diseases” (+0.9%). In contrast,
“melanoma” (-22.5%), “dermatitis” (-4.2%), “pruritus” (- 3.4%), “cellulitis” (-1.9%), and “fungal skin diseases”
(-1.5%) were the most under-represented skin diseases (figure 4C).
Germany Spain Netherlands France Italy United Kingdom United States Switzerland Belgium Denmark Portugal Greece Austria Poland China Canada Sweden Hungary Czech Republic Serbia
Spain Tunisia Germany Turkey United Kingdom United States Italy Romania Morocco Brazil India Portugal France Rep. of South Poland Serbia Greece
Russian Federation Canada
Japan
150 100 50 0 0 50 100 150 200 250
Lectures
Top 20 contributing countries at the EADV Posters
14% 19%
10%
7%
6%
6%
5%
5%
5%
5%
4%
4%
4%
4%
3%
3%
3%
3%
2%
2%
2%
12%
9%
9%
8%
7%
7%
6%
4%
4%
3%
3%
3%
2%
2%
2%
1%
1%
1%
1%
A
B
Figure 3. A) Countries contributing the highest number of lectures and posters (the top 20 categories for both are highlighted in bold).B) Visualisation of the geographical origins of submissions in google maps. The areas are color-coded in shades of yellow and red, depending on the number of contributions. Red indicates an area with particular activity. An interactive map is available at https://nwgraphs.web.app.
Discussion
Summary
In summary, the applied network analysis of the largest European dermatology conference in 2019, the EADV, revealed several clusters of main interests regarding lec- tures and posters, as well as close collaborations from contributing European nations. In particular, the focus of the contributions was patient-centred and therapeutic approaches. It is likely that this trend was associated with novel insights into the pathomechanisms of these diseases which is underlined by the respective comparisons with the number of PubMed publications. Naturally, as new medica- tions in these fields are emerging[18, 19], conferences serve to instruct European clinicians. However, there seemed to
be a regional disparity as contributors from Western Europe provided substantially more lectures compared to their col- leagues from Eastern Europe.
Lectures and posters
The analysis has shown that the congress had a particular focus on the treatment of patients. Chronic inflammatory skin diseases, such as psoriasis and atopic dermatitis, were among the diseases which received the most attention. This showed that the congress functionned as a clinically ori- ented, patient-centred educational platform for European dermatologists. Most likely this is due to their high preva- lence in Europe as well as new emerging and promising treatment options for related diseases[20-22]. The posters had a special emphasis on the presentation of clinical case
Psoriasis
EADV vs. DALYs vs. PubMed
Dermatitis Acne Melanoma Urticaria Viral skin diseases Keratinocyte carcinoma Alopecia Pruritus Ulcer Fungal skin diseases Pyoderma Scabies Cellulitis
Fungal skin diesases Urticaria Acne Viral skin diseases Dermatitis Scabies Cellulitis Pyoderma Ulcer Keratinocyte carcinoma Pruritus Alopecia Melanoma Psoriasis
Melanoma Dermatitis Pruritus Cellulitis Fungal skin diseases Pyoderma Ulcer Scabies Keratinocyte carcinoma Viral skin diseases Urticaria Alopecia Acne Psoriasis
-20
-5 0 5 10
EADV-PubMed [%]
EADV-DALYs [%]
15 20
-10 0 10 20
0 5 10 15 20 25 30 35
PubMed [%]
DALYs [%]
EADV [%]
overrepresented underrepresented
underrepresented
overrepresented
A
B
C
Figure 4.Relative representation of disease terms in the titles of the conference program relative to the global disease burden study in 2017 and publications on PubMed in 2018 (A), DALYs (B), and publications available on PubMed (C).
reports, which was also expected and, interestingly, focused on psoriasis and atopic dermatitis as well.
Contributors
Lectures were dominated by Western European nations with Germany at the top of the list, followed by the host country, Spain, and the Netherlands. Eastern European
countries were rather under-represented. For future confer- ences, network analysis could be used to include additional countries, for example, by special invitation and preferred ranking of topics tailored to the needs of different countries with different structures in dermatologic healthcare. As the situation was different for poster submissions, in which top contributors came from around the world, one option might be to switch posters to lectures for selected topics,
considering that a remarkable number of posters but only few lectures were provided by countries such as Romania.
Disease-adjusted life years
We detected a discrepancy between the analysed network and the DALYs of skin diseases worldwide, as published in 2017 by Karimkhaniet al.[3]. This analysis could sug- gest that the diseases, psoriasis and melanoma, received disproportionately more attention in the 2019 conference.
However, recent advances in dermatology and healthcare for both diseases combined regarding the overall disease burden might be a rather simple explanation [22, 23].
Instead, atopic dermatitis was rather under-represented at the congress when compared to the global disease bur- den, which is even more surprising considering the recent advances and new treatment options on the rise [24-26]
and the fact that many atopic dermatitis patients are still substantially impaired by the disease[21]. Based on their global burden measured by DALYs [3], viral and fun- gal skin diseases were also rather under-represented. It seems reasonable to tie this disparity with recent advances and new insights into the pathomechanism of, for exam- ple, psoriasis and melanoma, for which there are several new treatment options available, whereas new develop- ments and treatments for fungal and viral skin diseases are scarce[27].
PubMed
The distribution of the contributions to the largest European dermatology congress in 2019 were mostly in line with the focus of international dermatological research published on PubMed[2]. The only relevant discordances that we were able to detect were between congress titles and PubMed results of 2018 regarding psoriasis and melanoma. Pso- riasis therefore received a relatively excessive amount of attention, and melanoma not enough. The disproportion- ate attention on psoriasis, again, might be due to the fact that Europe, as a wealthy part of the world with good pub- lic health care systems, is a hotspot for psoriasis research and healthcare, and is promoted by the WHO, as doc- umented in their global psoriasis report [5]. The reason why melanoma was under-represented could be linked to the fact that despite a constant increase in incidence and mortality in Europe, it is still a disease that is not so prevalent with research limited to large and specialized centres[28].
Limitations
There are some study limitations. A downfall of the method is that the titles of contributions may not always reflect the true content. If an author chooses not to include the name of the disease corresponding to the focus of a contribution, but instead uses a rather vague title, qualitative categorization is not possible using network analysis.
Fungal and viral skin diseases are harder to standardize than terms such as “psoriasis” and “melanoma”, because sometimes only the pathogen of a certain disease is included in the title. Also, fungal and viral skin diseases cannot be further specified, similar to titles such as “non- melanoma skin cancer”. In order to exclude unnecessary
words (such as “and” or “between”), meticulous selection is required to reduce information loss while keeping the dataset manageable.
Obviously, the generalisability of the results is limited.
Firstly, not all relevant European scientists and clinicians in the dermatologic field contribute to annual conferences such as the EADV 2019. Therefore, some actors might not have been part of the analyses. Secondly, the quali- tative analyses only signalled the volume of contributions.
A researcher, for example, who made a single contribution corresponding to an important study, was treated by net- work analysis like any other contributor. Thirdly, all authors of a contribution were treated equally. This means that the hierarchy among the authors was not reflected by the results of the analysis, thus the leading author, who may have con- tributed the idea and composed most of the text, was not distinguished from others who may have only contributed a minor part,e.g.expert advice.
Conclusion
Network analysis of conferences may provide an overview of the focus of the targeted community. It is a modern and useful tool to access large sets of data and make them easy to evaluate. In the future, longitudinal network analysis of several congresses may provide more detailed information, including the delineation of topic trends, and may be used to determine additional regions and topics to be included in future events.
Disclosure. Conflicts of interest: none.
References
1.EADV 2019 Congress in Dermatology and Venereology Offi- cial Website; 2020. Available at: https://eadvmadrid2019.org/
(accessed 16 April 2020).
2.United States National Library of Medicine. Dermatology;
2020. Available at: https://www.ncbi.nlm.nih.gov/pubmed/?term
=%22dermatology%22 (accessed 16 April 2020).
3.Karimkhani C, Dellavalle RP, Coffeng LE,et al. Global skin disease morbidity and mortality: an update from the Global Burden of Disease Study 2013. JAMA Dermatol2017; 153: 406-12.
4.World Health Organization. WHO methods and data sources for global burden of disease estimates: 2000-2011. 2013 (accessed 16 April 2020).
5.Michalek IM, Loring B, John SM. Global Report on Psoriasis.
Geneva, Switzerland: World Health Organization, 2016.
6.Choi J, Yi S, Lee KC. Analysis of keyword networks in MIS research and implications for predicting knowledge evolution. Inform Manag 2011; 48: 371-81.
7.Zhang W, Zhang Q, Yu B, Zhao L. Knowledge map of creativity research based on keywords network and co-word analysis, 1992- 2011. Qual Quant2015; 49: 1023-38.
8.Borgatti SP, Mehra A, Brass DJ, Labianca G. Network analysis in the social sciences. Science2009; 323: 892-5.
9.Ward MD, Stovel K, Sacks A. Network Analysis and political sci- ence. Annu Rev Polit Sci2011; 14: 245-64.
10.Peter RS, Brehme T, Völzke H,et al. Epidemiologic research topics in Germany: a keyword network analysis of 2014 DGEpi conference presentations. Eur J Epidemiol2016; 31: 635-8.
11.Kaczmarczyk R, Bauerdorf F, Zink A. Whatˇıs driving dermatol- ogy? Contribution title analysis of the largest German Dermatology Congress 2019 2020 [under review].
12.Bauerdorf F, Kaczmarczyk R, Florestan T, Biedermann T. Trends and Perspectives for Dermatologic Research in Europe: An abstract title analysis of ESDR and IID congresses 2010-2019. J Invest Dermatol 2020; 140: S197-200.
13.Aristarán M, Swicegoog T, Tigas M, et al.Tabula 1.2.1; 2018.
Available at: https://tabula.technology/ (accessed 17 April 2020).
14.Bastian M, Heymann S, Jacomy M. Gephi: an open source soft- ware for exploring and manipulating networks. Third International AAAI Conference on Weblogs and Social Media 2009 (accessed 17 April 2020).
15.Jacomy M, Venturini T, Heymann S, Bastian M. ForceAtlas2, a continuous graph layout algorithm for handy network visualization designed for the Gephi software. PloS One2014; 9: e98679.
16.McKinney W. Data Structures for Statistical Computing in Python. Available at: https://conference.scipy.org/proceedings/
scipy2010/pdfs/mckinney.pdf (accessed 17 April 2020).
17.OpenCage GmbH. Easy, Open, Worldwide, Affordable Geocoding; 2020. Available at: https://opencagedata.com/
(accessed 17 April 2020).
18.Egeberg A, Ottosen MB, Gniadecki R, et al. Safety, efficacy and drug survival of biologics and biosimilars for moderate-to-severe plaque psoriasis. Br J Dermatol2018; 178: 509-19.
19.Foulkes AC, Ferguson F, Grindlay DJC,et al. What’s new in pso- riasis treatment? An analysis of systematic reviews published in 2015.
Clin Exp Dermatol2018; 43: 759-65.
20.Costello C, Maarouf M, Shi V. Navigating targeted therapeu- tics in dermatology: biologics and small molecules. J Drugs Dermatol 2018; 17: 1330-2.
21.Ring J, Zink A, Arents BWM,et al. Atopic eczema: burden of disease and individual suffering - results from a large EU study in adults.
J Eur Acad Dermatol Venereol2019; 33: 1331-40.
22.Tauber M, Apoil PA, Richet C,et al. Effect of dupilumab on atopic manifestations in patients treated for atopic dermatitis in real-life prac- tice. Br J Dermatol2019; 180: 1551-2.
23.Karimkhani C, Green AC, Nijsten T,et al. The global burden of melanoma: results from the Global Burden of Disease Study 2015. Br J Dermatol2017; 177: 134-40.
24.Guttman-Yassky E, Bissonnette R, Ungar B, et al. Dupilumab progressively improves systemic and cutaneous abnormalities in patients with atopic dermatitis. J Allergy Clin Immunol2019; 143:
155-72.
25.Hamann CR, Thyssen JP. Monoclonal antibodies against inter- leukin 13 and interleukin 31RA in development for atopic dermatitis. J Am Acad Dermatol2018; 78: S37-42.
26.He H, Guttman-Yassky E. JAK inhibitors for atopic dermatitis: an update. Am J Clin Dermatol2019; 20: 181-92.
27.Ramdass P, Mullick S, Farber HF. Viral skin diseases. Primary Care2015; 42: 517-67.
28.Garbe C, Keim U, Eigentler TK,et al. Time trends in incidence and mortality of cutaneous melanoma in Germany. J Eur Acad Dermatol Venereol2019; 33: 1272-80.
