- Email: [email protected]
Quantification of Urology Related Twitter Traffic Activity through a Standardized List of Social Media Communication Descriptors
Author's Accepted Manuscript Quantification of Urology-Related Twitter Traffic Activity through a Standardized List of Social Media Communication Descriptors Hendrik Borgmann, Matthew S. Katz, James Catto, Christopher Weight, Alexander Kutikov
PII: DOI: Reference:
S2352-0779(16)30212-6 10.1016/j.urpr.2016.07.011 URPR 230
To appear in: Urology Practice Accepted Date: 16 July 2016 Please cite this article as: Borgmann H, Katz MS, Catto J, Weight C, Kutikov A, Quantification of Urology-Related Twitter Traffic Activity through a Standardized List of Social Media Communication Descriptors, Urology Practice (2016), doi: 10.1016/j.urpr.2016.07.011. DISCLAIMER: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our subscribers we are providing this early version of the article. The paper will be copy edited and typeset, and proof will be reviewed before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to The Journal pertain. All press releases and the articles they feature are under strict embargo until uncorrected proof of the article becomes available online. We will provide journalists and editors with full-text copies of the articles in question prior to the embargo date so that stories can be adequately researched and written. The standard embargo time is 12:01 AM ET on that date.
1
ACCEPTED MANUSCRIPT
Quantification of Urology-Related Twitter Traffic Activity through a Standardized List of Social Media Communication Descriptors
RI PT
Hendrik Borgmann(1), Matthew S. Katz(2), James Catto(3), Christopher Weight(4), Alexander Kutikov(5)
Running title:
Quantification of Urology-related Twitter Traffic Social network, Social media, Urology, Blogging, Health Communication
AC C
EP
Keywords:
TE D
M AN U
Contact information for corresponding author: Hendrik Borgmann MD University Hospital Mainz Department of Urology Langenbeckstr. 1 55131 Mainz Germany Tel: +49 176 61502875 Fax: +49 69 6301 81137 Email: [email protected]
SC
1. University Hospital Mainz, Dept. of Urology, Mainz, Rheinland-Pfalz, Germany 2. Lowell General Hospital, Dept. of Radiation Medicine, Lowell, MA, United States 3. University of Sheffield, Academic Urology Unit, Sheffield, UK 4. University of Minnesota, Department of Urology, Minneapolis, MN, United States 5. Fox Chase Cancer Center, Division of Urologic Oncology, Philadelphia, PA, United States
2
ACCEPTED MANUSCRIPT
Abstract: Introduction:
RI PT
The social media microblogging service Twitter is gaining popularity in the field of urology as a fast and effective communication platform. We aimed to assess volume, subject matter, influencers and content of urology-related discussions on the Twitter platform using the
SC
recently proposed Urology Tag Ontology (UTO) hashtag list.
M AN U
Methods:
We queried the Symplur Signals database for tweet activity over a 1-year-period. We used the UTO, comprising 45 disease-related hashtags in 9 urologic subspecialties, to assess activity (numbers of tweets, users and impressions), users (geolocation, influencers) and content
Results:
TE D
(words, tweet enhancements) of urology-related Twitter traffic.
Twitter activity over the study period included 334,642 tweets by 104,166 users with
EP
1,397,107,305 impressions. Tweet activity varied between urological subspecialties and was largely dominated by urological oncology topics driven by #prostatecancer. Users came from
AC C
224 countries and from all continents around the globe. Healthcare organizations accounted for the largest proportion of influencers (44%) followed by other individuals (22%) and physicians (13%). The top words were “prevent” (used 20,955 times), “cancer” (19,610), “follow” (19,169), “men” (19,165) and “condom” (18,425). The median (range) number of shares was 2,200 (1,414-8,854) for the top 10 links, 2,123 (1,878-2,737) for the top 10 retweets, and 207 (12-438) for the top 10 photos. Conclusion:
3
ACCEPTED MANUSCRIPT
Twitter activity in the field of urology can be assessed using a standardized list of social media communication descriptors. The value of the Twitter communication platform is
RI PT
underscored by the large number of tweets and impressions in the urology space.
Introduction
SC
The microblogging Social Media platform Twitter has had an invigorating impact on communication between urologic healthcare professionals by enabling rapid and global 2
Indeed, Twitter has afforded users sharing of ideas across
M AN U
information exchange.1,
continents and has helped individuals stay abreast of the rapidly changing clinical and academic urologic landscape.3 An example of continuous Twitter use is the International Urological Journal Club #urojc, which is now in its 5th year.4 Moreover, the urological community has enthusiastically adopted the use of Twitter during conferences, allowing for
TE D
“augmented” experience for those in attendance and remote participation for those not able to be physically present at a particular meeting.3,
5
Interestingly, urology’s engagement with
EP
social media has outpaced adoption from other specialties with Urological Twitter traffic at academic meetings more than tripling the activity seen at all non-urological surgical
AC C
conferences combined.6 Data regarding personal benefit of Twitter use are also emerging.7 Indeed, development of professional guidelines for appropriate social media use underlines the medium’s integration into the fabric of clinical and academic exchange.8, 9 More recently, the Urology Tag Ontology (UTO) Project was launched. The UTO is a global initiative supported by key urologic social media stakeholders that seeks to standardize use of social media communication descriptors in order to facilitate communication and collaboration between health care provider and patient communities.10 The UTO hashtag list comprises 45 disease-related descriptors in 9 urologic subspecialties. Our aim in this study
4
ACCEPTED MANUSCRIPT
was to quantify and describe UTO hashtag volume, subject matter, influencers, and word
RI PT
content analysis of urology-related discussions on the Twitter platform.
Materials and methods
SC
In October 2015, we searched the Symplur Signals database for analytic insights on Twitter discussions of all hashtags from the UTO list for the time period of one year from 01 October
M AN U
2014 until 01 October 2015. Our activity analysis comprised assessment of the overall tweet activity, tweet metrics, and tweet language metrics. Overall tweet activity was recorded as number of tweets, users, and impressions (a combined measurement of the number of tweets and number of followers expressing the overall number of evoked impressions). We excluded
TE D
the hashtags #Infertility and #Testosterone from the analysis, since the tweet transcript revealed that these hashtags were largely polluted with non-medical content. We used Symplur for tweet analysis by retrieving statistics describing ratio and frequency of retweets,
EP
tweets with links, tweets with photos, tweet replies and tweets with user mentions. Tweet language analysis illustrates the language used by active participants. We identified language
AC C
type by a natural language processing algorithm provided directly by the Twitter application programming interface.
User analysis included the users’ geolocation, and an in-depth influencer analysis of the top influencers of the urology-related Twitter discussion. Geolocation of users was extracted by the Symplur Signals platform from the location information in each user profile. For the influencer analysis, we performed a Twitter profile analysis and assigned the top influencers to the healthcare categories patient, physician, non-physician healthcare professional, individual other, healthcare organization, organization other, and spam according to Symplur
5
ACCEPTED MANUSCRIPT
Signals healthcare category definitions.11 We included the top 1,000 contributors across all urology-related topics and the top 100 contributors for each urologic subspecialty as measured
RI PT
by number of tweets in the influencer analysis.
We used Symplur Signals’ tools for content analysis. The 100 most frequently used words in tweets on urology-related Twitter discussions were analysed and counted. The sentiment
SC
report analysed the 1,000 most recent tweets for positive and negative sentiment by a natural language processing algorithm with two custom dictionaries for positive and negative words,
frequently shared links and photos.
M AN U
respectively. Finally, we investigated the most frequently retweeted tweets and the most
We performed statistical calculations using Statistical Package for the Social Sciences 22.0
EP
Results
TE D
software (SPSS Inc., Chicago, IL, USA) and report median and range values.
Table 1 shows the overall tweet activity for the urology-related online discussion on Twitter
AC C
for a time period of 1 year. More than 330,000 tweets were posted on urology-related diseases, which generated up to 1.4 billion impressions. On average, 1 user contributed 3.2 tweets and generated 13,412 impressions. Tweet activity varied between urological subspecialties (Figure 1) and was largely dominated by sexual medicine/ infertility and oncology topics. Supplementary Table 1 lists the numbers for tweets, impressions and users for each disease hashtag of the UTO list. #ProstateCancer had the highest tweet activity, followed by #STD, #Urology, #TesticularCancer and #Incontinence. Tweet Metric analysis showed among 334,642 total tweets 249,173 (75%) tweets with links, 185,518 (58%) tweets
6
ACCEPTED MANUSCRIPT
with mentions, 141,168 (42%) tweets with retweet, 99,731 (30%) tweets with photos, and 7,552 (2%) tweets with replies. 279,358 (84%) of the tweets were in English language.
RI PT
The geolocation of users that contributed to urology-related disease discussions on Twitter is mapped in Figure 2. Users came from 224 countries and from all continents around the globe (Figure 2). Supplementary Table 2 lists the location of users according to country and
SC
continent. North American users were most frequent (55%) ahead of European (25%) and Asian (8%) users. The top 1,000 influencers of the Twitter discussion on urologic diseases
M AN U
according to Tweet volume stratified by healthcare category are shown in Figure 3. Healthcare organizations accounted for the largest proportion of influencers (44%) followed by other individuals (22%) and physicians (13%). Non-physician healthcare professionals (6%) and patients (2%) contributed to a small portion of Twitter discussions. Supplementary
TE D
Figure 1 shows the top 100 influencers of the Twitter discussion by subspecialty. Nonhealthcare organizations were among the key influencers in the sexual medicine/ infertility subspecialty. Physicians accounted for large proportions in subspecialties with low tweet
EP
activity, such as neurourology and reconstructive urology / trauma.
AC C
The most frequently used words in UTO-related tweets were prevent (used 20,955 times), cancer (19,610), follow (19,169), men (19,165) and condom (18,425) (Figure 4). Supplementary Table 3 lists the complete content analysis containing the 100 most frequently used words in urology-related Tweets. Sentiment analysis revealed that 72% of the tweets had a positive and 28% had a negative sentiment. The top 10 retweets by all users had a median of 2,123 (1,878-2,737) retweets. The top 10 retweets by physicians are enlisted in Supplementary Table 4. The median (range) number of shares for the top 10 links was 2,200 (1,414-8,854) and 207 (12-438) for the top 10 photos.
7
ACCEPTED MANUSCRIPT
Discussion The social media microblogging service Twitter is steadily gaining popularity in the field of
RI PT
urology as a fast and effective communication platform. The novel UTO hashtag list comprising 45 disease-related hashtags in 9 urologic subspecialties covers large parts of the urologic “Twittersphere”.10 Using the Symplur Signals analytics platform, we performed a
SC
comprehensive assessment of volume, subject matter, influencers and content of urologyrelated discussions on Twitter. We identified a high volume of Twitter activity over the 1-year
M AN U
study period, including over 300,000 tweets by over 100,000 users achieving up to 1.4 billion impressions (Table 1). Tweet activity varied between urological subspecialties and was largely dominated by urological oncology topics (Figure 1) related to the topic of #ProstateCancer. Healthcare organizations accounted for the largest proportion of influencers
TE D
followed by other individuals and physicians (Figure 3). Many of the most frequently used words relate to the area of disease prevention (Figure 4; Supplementary Table 3). In this study, we combined two novel mechanisms to assess volume and content of urology-
EP
related Twitter discussions. Symplur (www.symplur.com) is a Twitter analysis website that
AC C
maintains the largest database of healthcare-related Twitter conversations, while the novel Symplur signals platform (www.symplur.com/signals) is a research analytics tool that aims to empower decision-making with real-time access to insights from over a billion healthcare social media data points. Furthermore, the UTO hashtag list is a global initiative supported by key urologic social media stakeholders aiming to standardize use of social media communication descriptors and includes 45 disease-related hashtags in 9 urologic subspecialties.10 For the first time, we harnessed both the Symplur signals platform as analytics tool and the UTO list of urologic communication descriptors to quantify urologyrelated Twitter traffic.
8
ACCEPTED MANUSCRIPT
In order to structure medicine-related Twitter communication, Tag ontologies have been previously proposed for the fields of oncology12 and radiology13. Overall Twitter activity was
RI PT
similar for the urology hashtags (334,000 Tweets; Table 1) for the time period from October 2014 until October 2015 compared to disease-specific cancer hashtags (310,000 Tweets) for the time period from July 2014 until July 2015.14 These data further support the view that the
SC
rather small specialty of urology is leading the way in Twitter use compared to other specialty communities. Accordingly, urological conferences were associated with more Twitter activity
M AN U
than non-urological surgical conferences in all parameters in the current analysis, with more than three-fold the number of tweets, users and impressions.6
Hashtags describing diseases in the field of urological oncology combined for 43% of all urology-related Tweets and thus largely dominated as subspecialty in the urologic Twittersphere (Figure 1). These findings are in line with an analysis of Tweet content
TE D
according to subspecialty at the European Association of Urology congress, in which urological oncology accounted for 21% of the Tweets.15 Subspecialisation is a trend in global 17
and these data identify opportunities for subspecialties with limited Twitter
EP
urology16,
activity to seek strategies for better engagement on a platform that already enjoys up to 1.4
AC C
billion impressions for urology-related Tweets per year. Analysis of influencer type revealed that healthcare organizations and physicians appear to be – besides individual others - the most active users of the Twitter platform in the urology space (Figure 3). As such, these professional healthcare information sources likely lend fidelity and quality to urology-related Twitter content. In contrast, patients comprised only 2% of the 1000 top influencers in the urology-related Twitter discussion, leaving significant opportunities for patient engagement on the platform. For instance, recent data suggest that breast cancer
9
ACCEPTED MANUSCRIPT
patients who participated in a Twitter chat benefited from improved disease knowledge and reduced anxiety.18
RI PT
Content analysis over the year study period showed that a large proportion of the most frequently used words focus on prevention (Figure 4; Supplementary Table 3). The words “prevent”, “awareness”, and “condom” were each used more than 16,000 times in indexed
SC
Tweets (Supplementary Table 3), demonstrating that the Twitter platform is being used to disseminate and engage the Urologic community around these relevant topics. Considering
M AN U
both increasing prevalence and costs for urological diseases 19 and cost-effectiveness of many primary prevention programs 20, utilization of Twitter for distribution of actionable prevention recommendations is notable.
Although our analysis for the first time sheds light on granular details of urology-related Twitter landscape, we readily acknowledge that our study is not without limitations. The UTO
TE D
hashtag list only includes salient hashtags selected by Urological social media stakeholders and thus does not completely capture urology-related discussion on Twitter, resulting in an
EP
underestimation of activity for certain diseases or subspecialties. Although the tags in the UTO include some of the most common topics, excluding other tags may also result in loss of
AC C
fidelity regarding assessment of who and how contributes to urology-related conversations on Twitter. Similarly, although assessed for random basis, improper use or spamming of disease hashtags might have led to an overestimation of activity that represents meaningful conversation. Finally, Twitter is a rapidly changing and growing social media platform, indicating that this assessment is not generalizable beyond the study period or to other platforms. Notwithstanding, we found that the UTO hashtag list offers a structured mechanism for standardized assessment of Twitter activity in the urology community. Our work identifies
10
ACCEPTED MANUSCRIPT
significant opportunities for engagement of non-urological-oncology physician and patient communities. Future studies will need to investigate the relationship between Tweet activity
RI PT
data from UTO hashtags to academic benchmark metrics and to outcomes affecting clinical practice.
SC
Conclusion:
Twitter activity in the field of urology can be assessed using a standardized list of social
M AN U
media communication descriptors. Healthcare organizations and physicians appear to be the most active users of the Twitter platform, leaving significant opportunities for improved engagement of the patient and other healthcare provider communities. Our descriptive study indicates that many individuals and organizations already use Twitter to discuss urologyrelated health topics. More research is needed to identify how these online conversations
TE D
affect clinical care and patient outcomes. High volume of UTO-related activity suggests that communication via the Twitter platform harbours value for the urology community and its
AC C
EP
stakeholders.
11
ACCEPTED MANUSCRIPT
Acknowledgements: We thank all Twitter users who contributed to the online discussion on urology-related topics
RI PT
on Twitter. Conflict of Interest Statement None declared.
SC
Funding Sources
M AN U
None.
References
4. 5. 6. 7.
8. 9.
10. 11.
TE D
3.
EP
2.
Wilkinson, S. E., Basto, M. Y., Perovic, G. et al.: The social media revolution is changing the conference experience: analytics and trends from eight international meetings. BJU Int, 115: 839, 2015 Loeb, S.: Social media makes global urology meetings truly global. : The influence of Twitter. BJU Int, 115: 175, 2015 Loeb, S., Catto, J., Kutikov, A.: Social media offers unprecedented opportunities for vibrant exchange of professional ideas across continents. Eur Urol, 66: 118, 2014 Thangasamy, I. A., Leveridge, M., Davies, B. J. et al.: International Urology Journal Club via Twitter: 12-month experience. Eur Urol, 66: 112, 2014 Matta, R., Doiron, C., Leveridge, M. J.: The dramatic increase in social media in urology. J Urol, 192: 494, 2014 Chung, A., Woo, H.: Twitter in urology and other surgical specialties at global conferences. ANZ J Surg, 2015 Borgmann, H., DeWitt, S., Tsaur, I. et al.: Novel survey disseminated through Twitter supports its utility for networking, disseminating research, advocacy, clinical practice and other professional goals. Can Urol Assoc J, 9: E713, 2015 Murphy, D. G., Loeb, S., Basto, M. Y. et al.: Engaging responsibly with social media: the BJUI guidelines. BJU Int, 114: 9, 2014 Roupret, M., Morgan, T. M., Bostrom, P. J. et al.: European Association of Urology (@Uroweb) recommendations on the appropriate use of social media. Eur Urol, 66: 628, 2014 Kutikov, A., Woo, H. H., Catto, J. W.: Urology Tag Ontology Project: Standardizing Social Media Communication Descriptors. Eur Urol, 69: 183, 2016 [Accessed April 12, 2016] Symplur Signals: Healthcare category definitions. Avaiable at https://docs.google.com/spreadsheets/d/1HexI7X1KR0dPdFvJkd34DEtHF_18BGHzdnPNrauH NwI/edit?pref=2&pli=1#gid=0
AC C
1.
12
ACCEPTED MANUSCRIPT
18.
19.
20.
RI PT
17.
SC
16.
M AN U
15.
TE D
14.
EP
13.
[Accessed April 12, 2016] Katz, M. S.: The Oncology Tag Ontology: Professional–Centered Collaboration and Networking. Avaiable at http://www.symplur.com/blog/oncology-tagontology-professional-centered-collaboration-networking/ [Accessed April 12, 2016] Hawkins, M.: Radiology Hashtag Ontology Project. Avaiable at http://www.symplur.com/blog/radiology-hashtag-ontology-project/ Katz, M. S., Utengen, A., Anderson, P. F. et al.: Disease-Specific Hashtags for Online Communication About Cancer Care. JAMA Oncol: 1, 2015 Borgmann, H., Woelm, JH. Merseburger, A. et al.: Qualitative Twitter analysis of participants, tweet strategies, and tweet content at a major urologic conference. Can Urol Assoc J, 10: 3944, 2016 Flanigan, R. C.: Subspecialty certification in urology: a US perspective. Nat Clin Pract Urol, 3: 509, 2006 Debruyne, F. M.: Subspecialty certification in urology: a European perspective. Nat Clin Pract Urol, 3: 625, 2006 Attai, D. J., Cowher, M. S., Al-Hamadani, M. et al.: Twitter Social Media is an Effective Tool for Breast Cancer Patient Education and Support: Patient-Reported Outcomes by Survey. J Med Internet Res, 17: e188, 2015 Antonelli, J. A., Maalouf, N. M., Pearle, M. S. et al.: Use of the National Health and Nutrition Examination Survey to calculate the impact of obesity and diabetes on cost and prevalence of urolithiasis in 2030. Eur Urol, 66: 724, 2014 Lotan, Y., Buendia Jimenez, I., Lenoir-Wijnkoop, I. et al.: Primary prevention of nephrolithiasis is cost-effective for a national healthcare system. BJU Int, 110: E1060, 2012
AC C
12.
13
ACCEPTED MANUSCRIPT
Figure Legends Figure 1. Number of tweets and users of urology-related Twitter traffic stratified by urologic
RI PT
subspecialty (logarithmic scale). Figure 2. Geolocation of users contributing to urology-related Twitter traffic. The color tone
users and light tones countries with few or no users.
SC
reflects the number of users per country: Dark tones represent countries with large number of
stratified by healthcare category.
M AN U
Figure 3. Top 1000 influencers of urology-related Twitter traffic according to tweet volume
Figure 4. Bubble chart visualizing the 100 most frequently used words in tweets of urology-
AC C
EP
TE D
related Twitter traffic. Bubble size correlates to word prevalence/frequency in tweets.
ACCEPTED MANUSCRIPT Total Per Month Per Week 1,397,107,305 114,830,737 26,793,839 13,412 1,102 257 334,642 3.2 104,166
27,503 0.26 8,562
6,417 0.06 1,998
Per Day 3,827,691 36.7
Per Hour 159,487 1.5
917 0.009 285
38.2 0.0004 11.9
RI PT
Metric Impressions Impressions per user Tweets Tweets per user Users who tweeted
Table 1. Overview of quantitative metrics of urology-related Twitter traffic over a time period
AC C
EP
TE D
M AN U
SC
of 1 year.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Key of Definitions for Abbreviations Urology Tag Ontology
AC C
EP
TE D
M AN U
SC
RI PT
UTO
ACCEPTED Supplementary Table 1: Tweet activity according to specificMANUSCRIPT hashtags over a 1-year time period (QIV 2014 - QIII 2015) Hashtag
Disease
Tweets Impressions Users
Tweets/User
Impressions/User
Oncology Adrenal Cancer Bladder Cancer Bladder Cancer Kidney Cancer Kidney Cancer Prostate Cancer Penile Cancer Prostate Cancer Sarcoma Testicular Cancer Testicular Cancer Urologic Oncology Upper Tract Urothelial Cancer
29 9256 538 1836 8679 6571 28 94531 1994 21576 905 73 70
72398 40231983 1872897 4753283 54104871 22489523 57578 442379110 4263258 138545202 2100073 80471 95150
24 3828 358 1061 3393 2926 17 37472 600 9782 409 47 42
1.2 2.4 1.5 1.7 2.6 2.2 1.6 2.5 3.3 2.2 2.2 1.6 1.7
3017 10510 5232 4480 15946 7686 3387 11806 7105 14163 5135 1712 2265
Endourology Kidney Stones Endourology
139 10335 61
100281 36804074 67661
72 3711 40
1.9 2.8 1.5
1393 9918 1692
Female Pelvic Medicine & Reconstructive Surgery Incontinence Interstitial Cystitis Overactive Bladder
10 16567 865 3662
26811 47526084 1631458 15435903
9 5630 260 1558
1.1 2.9 3.3 2.4
2979 8442 6275 9908
2656 2411 22
3918185 15187315 59892
1113 1213 13
2.4 2.0 1.7
3520 12520 4607
1244
6866926
549
2.3
12508
10 35 11
21398 68956 24503
9 26 8
1.1 1.3 1.4
2378 2652 3063
12 4
9293 14373
10 4
1.2 1.0
929 3593
144 10758 261 7 315 460 2454 76 88963
102006 29907616 135742 20361 575912 690402 1139096 293163 434103705
46 3028 43 6 158 144 699 58 30182
3.1 3.6 6.1 1.2 2.0 3.2 3.5 1.3 2.9
2218 9877 3157 3394 3645 4794 1630 5055 14383
5 17 2 3711 43645
16531 47299 5315 7141044 66079829
5 7 2 869 10778
1.0 2.4 1.0 4.3 4.0
3306 6757 2658 8218 6131
Calculi #EndoUrology** #KidneyStones #UroStone*
Mayer-Rokitansky-Küster-Hauser Syndrome Pediatric Cancer Pediatric Urology
Transplant #KidneyTransplant*
Renal Transplant
Neurourology #NeuroUrol* #UroBPH* #UroUTI*
Neurourology Benign Prostatic Hyperplasia Urinary Infections
Uro Recon / Trauma #UroRecon* #UroTrauma*
Reconstructive Urology Urologic Trauma
Sexual Med. / Infertility
TE D
Urologic Epidemiology Urologic Health Policy & Economics Urologic Health Services Research International Urology Journal Club Urology
AC C
Others #UroEpi* #UroHPE* #UroHSR* #urojc #Urology
Andrology Erectile Dysfunction Hypogonadism Erectile Dysfunction Male Infertility Peyronies Disease Premature Ejaculation Priapism Sexually Transmitted Disease
EP
#Andrology* #erectiledysfunction*** #Hypogonadism* #MaleED* #MaleInfertility* #Peyronies #PrematureEjaculation* #Priapism* #STD
M AN U
Pediatrics #MRKH*** #pedcsm #PedUro*
SC
Female Urology #FPMRS* #Incontinence #InterstitialCystitis* #OAB*
RI PT
#adcsm #BladderCancer #blcsm #kcsm #KidneyCancer #PCSM #PenileCancer* #ProstateCancer #scmsm #TesticularCancer #tscsm #uroonc*** #utuc**
Supplementary Table S1:Tweet activity according to specific hashtags over a 1-year time period (QIV 2014 - QIII 2015). * 3-months time-period only: Tracking of this hashtag's Tweet activity has begun with QIII 2015 ** 6-months time-period only: Tracking of this hashtag's Tweet activity has begun with QII 2015 *** 9-months time-period only: Tracking of this hashtag's Tweet activity has begun with QI 2015
ACCEPTED MANUSCRIPT Geo-Location of contributors to the Twitter discussion in urology according to continents
1 2 3 4 5 6
Continent North America Europe Asia South America Africa Australia
n (of 63,219) % 34.796 55% 15,700 25% 4.847 8% 3.058 5% 2.501 4% 2.317 4%
RI PT
Rank
Geo-Location of countries of contributors to the Twitter discussion in urology according to countri
SC
n (of 63,219) % 28.314 45% 9.211 15% 4.884 8% 1.961 3% 1,460 0% 1.447 2% 1.222 2% 984 2% 970 2% 907 1% 866 1% 755 1% 665 1% 506 1% 475 1% 430 1% 395 1% 364 1% 325 1% 275 0% 268 0% 248 0% 243 0% 242 0% 240 0% 240 0% 238 0% 211 0% 198 0% 190 0% 187 0% 183 0% 171 0% 162 0%
M AN U
TE D
EP
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 25 26 27 28 29 30 31 32 33
Country United States of America United Kingdom Canada Australia India Russia Brazil France Germany Mexico Spain Indonesia South Africa Ireland Italy Nigeria Argentina Netherlands Philippines Venezuela Kenya New Zealand Colombia Belgium Slovenia Trinidad and Tobago Finland Switzerland Turkey Iran Pakistan Sweden Saudi Arabia Japan
AC C
Rank
ACCEPTED MANUSCRIPT
SC M AN U TE D
n % (of all tweets) 20.955 6% 19,610 6% 19.169 6% 19.165 6% 18.425 6% 18,370 5% 16.584 5% 13.505 4% 12,320 4% 12.159 4% 3% 11.181 10.592 3% 10.346 3% 10.344 3% 10,340 3% 10,160 3% 9.672 3% 9,350 3% 8.625 3% 8.444 3% 8.421 3% 2% 8.305 2% 6.875 2% 6,820 2% 6.533 2% 6.433 2% 6.247
EP
word prevent cancer follow men condom appreciate awareness patients life test risk prostate free therapy research know quality shows month help check preserves study health story reuters balls
AC C
Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
RI PT
Top 100 Words used in Tweets on urology-related topics
SC M AN U
2% 2% 2% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1%
TE D
5.647 5.208 5,170 4.865 4,840 4.788 4.675 4,620 4.606 4,400 4.364 4.339 4,290 4,290 4,180 4,180 4.125 4,070 4.064 4.046 3,960 3.889 3.659 3,630 3.612 3.575 3,520 3,520 3,490 3.465 3.465
EP
support text screening treatment sexual diagnosis common thx diagnosed people good pads early talk self stone supporting you're cause need quoted today october kidney learn monthly raise information use sex testing
AC C
28 29 30 31 32 33 34 35 36 37 38 39 40 40 41 41 42 43 44 45 46 47 48 49 50 51 52 52 53 54 54
RI PT
ACCEPTED MANUSCRIPT
SC M AN U
1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1%
TE D
3.408 3.405 3.355 3.216 3.135 3.135 3,080 3,080 3.025 2,970 2.915 2.895 2,860 2,860 2.814 2.763 2,750 2.741 2.723 2.695 2.695 2.695 2.677 2,640 2,640 2.585 2.585 2.505 2.475 2.475 2.439
EP
great fight reminder hiv join women think age including safe healthy testicular challenge big active video avoiding abena bladder std reduce expression thanks kub read sexually right delivery young care urology
AC C
55 56 57 58 59 59 60 60 61 62 63 64 65 65 66 67 68 69 70 71 71 71 72 73 73 74 74 75 76 76 77
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1%
RI PT
2,420 2,380 2.365 2.365 2.365 2.365 2.365 2.343 2.342 2.341 2,310
SC
researchers san tested online oct better linked results nominate htt aggressive
M AN U
78 79 80 80 80 80 80 81 82 83 84
AC C
EP
TE D
Supplementary Table S3: Top 100 Words used in Tweets on urology-related topics
ACCEPTED MANUSCRIPT
Top 10 tweets by physicians having the most retweets on urology-related topics Tweet text #BladderCancer is no longer just 1 disease - Paradigm shift #GU15 @ASCO Hospital nocturnal phone calls by specialty: #urology up there nzma.org.nz/__data/assets/pdf_… #prostatecancer diagnoses down 28% 1 yr after USPSTF grade D @JUrology @urogeek @UroCancerMD jurology.com/ar5cle/S0022-534… @LoebStacy at #EAU15 on #ProstateCancer: 'new Epstein/ISUP grade groups coming soon' @AmerUrological @EUplatinum Schroder receives a touching standing ovation at #AUA15 plenary for his contributions to #prostatecancer screening Walsh's "confession" on changing paradigm: shifting prostatectomy to high risk #ProstateCancer #pcwc15 Unusual pattern of #bladdercancer - "Chordoid"-like pattern. #gupath Moving toward cytoreductive prostatectomy for M1 #prostatecancer? Pros/cons by @DrDanielMoon-need more data #pcwc15 Clear cell RCC hemangioma-like areas, potential pitfall for renal mass biopsy ncbi.nlm.nih.gov/m/pubmed/2336… #kidneycancer .@theNCI Annual Report on Cancer: #prostatecancer had largest mortality decline of all cancers 1.usa.gov/1BLNNzg
RI PT
User name @mehrazinmd @loebstacy @danbarocas @karitikkinen @loebstacy @loebstacy @gleason4plus5 @loebstacy @williamson_sr @loebstacy
SC
number of retweets 28 28 26 26 23 22 21 17 16 16
M AN U
Rank 1 1 2 2 3 4 5 6 7 8
AC C
EP
TE D
Supplementary Table S4:Top 10 tweets by physicians having the most retweets on urology-related topics
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
PII: DOI: Reference:
S2352-0779(16)30212-6 10.1016/j.urpr.2016.07.011 URPR 230
To appear in: Urology Practice Accepted Date: 16 July 2016 Please cite this article as: Borgmann H, Katz MS, Catto J, Weight C, Kutikov A, Quantification of Urology-Related Twitter Traffic Activity through a Standardized List of Social Media Communication Descriptors, Urology Practice (2016), doi: 10.1016/j.urpr.2016.07.011. DISCLAIMER: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our subscribers we are providing this early version of the article. The paper will be copy edited and typeset, and proof will be reviewed before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to The Journal pertain. All press releases and the articles they feature are under strict embargo until uncorrected proof of the article becomes available online. We will provide journalists and editors with full-text copies of the articles in question prior to the embargo date so that stories can be adequately researched and written. The standard embargo time is 12:01 AM ET on that date.
1
ACCEPTED MANUSCRIPT
Quantification of Urology-Related Twitter Traffic Activity through a Standardized List of Social Media Communication Descriptors
RI PT
Hendrik Borgmann(1), Matthew S. Katz(2), James Catto(3), Christopher Weight(4), Alexander Kutikov(5)
Running title:
Quantification of Urology-related Twitter Traffic Social network, Social media, Urology, Blogging, Health Communication
AC C
EP
Keywords:
TE D
M AN U
Contact information for corresponding author: Hendrik Borgmann MD University Hospital Mainz Department of Urology Langenbeckstr. 1 55131 Mainz Germany Tel: +49 176 61502875 Fax: +49 69 6301 81137 Email: [email protected]
SC
1. University Hospital Mainz, Dept. of Urology, Mainz, Rheinland-Pfalz, Germany 2. Lowell General Hospital, Dept. of Radiation Medicine, Lowell, MA, United States 3. University of Sheffield, Academic Urology Unit, Sheffield, UK 4. University of Minnesota, Department of Urology, Minneapolis, MN, United States 5. Fox Chase Cancer Center, Division of Urologic Oncology, Philadelphia, PA, United States
2
ACCEPTED MANUSCRIPT
Abstract: Introduction:
RI PT
The social media microblogging service Twitter is gaining popularity in the field of urology as a fast and effective communication platform. We aimed to assess volume, subject matter, influencers and content of urology-related discussions on the Twitter platform using the
SC
recently proposed Urology Tag Ontology (UTO) hashtag list.
M AN U
Methods:
We queried the Symplur Signals database for tweet activity over a 1-year-period. We used the UTO, comprising 45 disease-related hashtags in 9 urologic subspecialties, to assess activity (numbers of tweets, users and impressions), users (geolocation, influencers) and content
Results:
TE D
(words, tweet enhancements) of urology-related Twitter traffic.
Twitter activity over the study period included 334,642 tweets by 104,166 users with
EP
1,397,107,305 impressions. Tweet activity varied between urological subspecialties and was largely dominated by urological oncology topics driven by #prostatecancer. Users came from
AC C
224 countries and from all continents around the globe. Healthcare organizations accounted for the largest proportion of influencers (44%) followed by other individuals (22%) and physicians (13%). The top words were “prevent” (used 20,955 times), “cancer” (19,610), “follow” (19,169), “men” (19,165) and “condom” (18,425). The median (range) number of shares was 2,200 (1,414-8,854) for the top 10 links, 2,123 (1,878-2,737) for the top 10 retweets, and 207 (12-438) for the top 10 photos. Conclusion:
3
ACCEPTED MANUSCRIPT
Twitter activity in the field of urology can be assessed using a standardized list of social media communication descriptors. The value of the Twitter communication platform is
RI PT
underscored by the large number of tweets and impressions in the urology space.
Introduction
SC
The microblogging Social Media platform Twitter has had an invigorating impact on communication between urologic healthcare professionals by enabling rapid and global 2
Indeed, Twitter has afforded users sharing of ideas across
M AN U
information exchange.1,
continents and has helped individuals stay abreast of the rapidly changing clinical and academic urologic landscape.3 An example of continuous Twitter use is the International Urological Journal Club #urojc, which is now in its 5th year.4 Moreover, the urological community has enthusiastically adopted the use of Twitter during conferences, allowing for
TE D
“augmented” experience for those in attendance and remote participation for those not able to be physically present at a particular meeting.3,
5
Interestingly, urology’s engagement with
EP
social media has outpaced adoption from other specialties with Urological Twitter traffic at academic meetings more than tripling the activity seen at all non-urological surgical
AC C
conferences combined.6 Data regarding personal benefit of Twitter use are also emerging.7 Indeed, development of professional guidelines for appropriate social media use underlines the medium’s integration into the fabric of clinical and academic exchange.8, 9 More recently, the Urology Tag Ontology (UTO) Project was launched. The UTO is a global initiative supported by key urologic social media stakeholders that seeks to standardize use of social media communication descriptors in order to facilitate communication and collaboration between health care provider and patient communities.10 The UTO hashtag list comprises 45 disease-related descriptors in 9 urologic subspecialties. Our aim in this study
4
ACCEPTED MANUSCRIPT
was to quantify and describe UTO hashtag volume, subject matter, influencers, and word
RI PT
content analysis of urology-related discussions on the Twitter platform.
Materials and methods
SC
In October 2015, we searched the Symplur Signals database for analytic insights on Twitter discussions of all hashtags from the UTO list for the time period of one year from 01 October
M AN U
2014 until 01 October 2015. Our activity analysis comprised assessment of the overall tweet activity, tweet metrics, and tweet language metrics. Overall tweet activity was recorded as number of tweets, users, and impressions (a combined measurement of the number of tweets and number of followers expressing the overall number of evoked impressions). We excluded
TE D
the hashtags #Infertility and #Testosterone from the analysis, since the tweet transcript revealed that these hashtags were largely polluted with non-medical content. We used Symplur for tweet analysis by retrieving statistics describing ratio and frequency of retweets,
EP
tweets with links, tweets with photos, tweet replies and tweets with user mentions. Tweet language analysis illustrates the language used by active participants. We identified language
AC C
type by a natural language processing algorithm provided directly by the Twitter application programming interface.
User analysis included the users’ geolocation, and an in-depth influencer analysis of the top influencers of the urology-related Twitter discussion. Geolocation of users was extracted by the Symplur Signals platform from the location information in each user profile. For the influencer analysis, we performed a Twitter profile analysis and assigned the top influencers to the healthcare categories patient, physician, non-physician healthcare professional, individual other, healthcare organization, organization other, and spam according to Symplur
5
ACCEPTED MANUSCRIPT
Signals healthcare category definitions.11 We included the top 1,000 contributors across all urology-related topics and the top 100 contributors for each urologic subspecialty as measured
RI PT
by number of tweets in the influencer analysis.
We used Symplur Signals’ tools for content analysis. The 100 most frequently used words in tweets on urology-related Twitter discussions were analysed and counted. The sentiment
SC
report analysed the 1,000 most recent tweets for positive and negative sentiment by a natural language processing algorithm with two custom dictionaries for positive and negative words,
frequently shared links and photos.
M AN U
respectively. Finally, we investigated the most frequently retweeted tweets and the most
We performed statistical calculations using Statistical Package for the Social Sciences 22.0
EP
Results
TE D
software (SPSS Inc., Chicago, IL, USA) and report median and range values.
Table 1 shows the overall tweet activity for the urology-related online discussion on Twitter
AC C
for a time period of 1 year. More than 330,000 tweets were posted on urology-related diseases, which generated up to 1.4 billion impressions. On average, 1 user contributed 3.2 tweets and generated 13,412 impressions. Tweet activity varied between urological subspecialties (Figure 1) and was largely dominated by sexual medicine/ infertility and oncology topics. Supplementary Table 1 lists the numbers for tweets, impressions and users for each disease hashtag of the UTO list. #ProstateCancer had the highest tweet activity, followed by #STD, #Urology, #TesticularCancer and #Incontinence. Tweet Metric analysis showed among 334,642 total tweets 249,173 (75%) tweets with links, 185,518 (58%) tweets
6
ACCEPTED MANUSCRIPT
with mentions, 141,168 (42%) tweets with retweet, 99,731 (30%) tweets with photos, and 7,552 (2%) tweets with replies. 279,358 (84%) of the tweets were in English language.
RI PT
The geolocation of users that contributed to urology-related disease discussions on Twitter is mapped in Figure 2. Users came from 224 countries and from all continents around the globe (Figure 2). Supplementary Table 2 lists the location of users according to country and
SC
continent. North American users were most frequent (55%) ahead of European (25%) and Asian (8%) users. The top 1,000 influencers of the Twitter discussion on urologic diseases
M AN U
according to Tweet volume stratified by healthcare category are shown in Figure 3. Healthcare organizations accounted for the largest proportion of influencers (44%) followed by other individuals (22%) and physicians (13%). Non-physician healthcare professionals (6%) and patients (2%) contributed to a small portion of Twitter discussions. Supplementary
TE D
Figure 1 shows the top 100 influencers of the Twitter discussion by subspecialty. Nonhealthcare organizations were among the key influencers in the sexual medicine/ infertility subspecialty. Physicians accounted for large proportions in subspecialties with low tweet
EP
activity, such as neurourology and reconstructive urology / trauma.
AC C
The most frequently used words in UTO-related tweets were prevent (used 20,955 times), cancer (19,610), follow (19,169), men (19,165) and condom (18,425) (Figure 4). Supplementary Table 3 lists the complete content analysis containing the 100 most frequently used words in urology-related Tweets. Sentiment analysis revealed that 72% of the tweets had a positive and 28% had a negative sentiment. The top 10 retweets by all users had a median of 2,123 (1,878-2,737) retweets. The top 10 retweets by physicians are enlisted in Supplementary Table 4. The median (range) number of shares for the top 10 links was 2,200 (1,414-8,854) and 207 (12-438) for the top 10 photos.
7
ACCEPTED MANUSCRIPT
Discussion The social media microblogging service Twitter is steadily gaining popularity in the field of
RI PT
urology as a fast and effective communication platform. The novel UTO hashtag list comprising 45 disease-related hashtags in 9 urologic subspecialties covers large parts of the urologic “Twittersphere”.10 Using the Symplur Signals analytics platform, we performed a
SC
comprehensive assessment of volume, subject matter, influencers and content of urologyrelated discussions on Twitter. We identified a high volume of Twitter activity over the 1-year
M AN U
study period, including over 300,000 tweets by over 100,000 users achieving up to 1.4 billion impressions (Table 1). Tweet activity varied between urological subspecialties and was largely dominated by urological oncology topics (Figure 1) related to the topic of #ProstateCancer. Healthcare organizations accounted for the largest proportion of influencers
TE D
followed by other individuals and physicians (Figure 3). Many of the most frequently used words relate to the area of disease prevention (Figure 4; Supplementary Table 3). In this study, we combined two novel mechanisms to assess volume and content of urology-
EP
related Twitter discussions. Symplur (www.symplur.com) is a Twitter analysis website that
AC C
maintains the largest database of healthcare-related Twitter conversations, while the novel Symplur signals platform (www.symplur.com/signals) is a research analytics tool that aims to empower decision-making with real-time access to insights from over a billion healthcare social media data points. Furthermore, the UTO hashtag list is a global initiative supported by key urologic social media stakeholders aiming to standardize use of social media communication descriptors and includes 45 disease-related hashtags in 9 urologic subspecialties.10 For the first time, we harnessed both the Symplur signals platform as analytics tool and the UTO list of urologic communication descriptors to quantify urologyrelated Twitter traffic.
8
ACCEPTED MANUSCRIPT
In order to structure medicine-related Twitter communication, Tag ontologies have been previously proposed for the fields of oncology12 and radiology13. Overall Twitter activity was
RI PT
similar for the urology hashtags (334,000 Tweets; Table 1) for the time period from October 2014 until October 2015 compared to disease-specific cancer hashtags (310,000 Tweets) for the time period from July 2014 until July 2015.14 These data further support the view that the
SC
rather small specialty of urology is leading the way in Twitter use compared to other specialty communities. Accordingly, urological conferences were associated with more Twitter activity
M AN U
than non-urological surgical conferences in all parameters in the current analysis, with more than three-fold the number of tweets, users and impressions.6
Hashtags describing diseases in the field of urological oncology combined for 43% of all urology-related Tweets and thus largely dominated as subspecialty in the urologic Twittersphere (Figure 1). These findings are in line with an analysis of Tweet content
TE D
according to subspecialty at the European Association of Urology congress, in which urological oncology accounted for 21% of the Tweets.15 Subspecialisation is a trend in global 17
and these data identify opportunities for subspecialties with limited Twitter
EP
urology16,
activity to seek strategies for better engagement on a platform that already enjoys up to 1.4
AC C
billion impressions for urology-related Tweets per year. Analysis of influencer type revealed that healthcare organizations and physicians appear to be – besides individual others - the most active users of the Twitter platform in the urology space (Figure 3). As such, these professional healthcare information sources likely lend fidelity and quality to urology-related Twitter content. In contrast, patients comprised only 2% of the 1000 top influencers in the urology-related Twitter discussion, leaving significant opportunities for patient engagement on the platform. For instance, recent data suggest that breast cancer
9
ACCEPTED MANUSCRIPT
patients who participated in a Twitter chat benefited from improved disease knowledge and reduced anxiety.18
RI PT
Content analysis over the year study period showed that a large proportion of the most frequently used words focus on prevention (Figure 4; Supplementary Table 3). The words “prevent”, “awareness”, and “condom” were each used more than 16,000 times in indexed
SC
Tweets (Supplementary Table 3), demonstrating that the Twitter platform is being used to disseminate and engage the Urologic community around these relevant topics. Considering
M AN U
both increasing prevalence and costs for urological diseases 19 and cost-effectiveness of many primary prevention programs 20, utilization of Twitter for distribution of actionable prevention recommendations is notable.
Although our analysis for the first time sheds light on granular details of urology-related Twitter landscape, we readily acknowledge that our study is not without limitations. The UTO
TE D
hashtag list only includes salient hashtags selected by Urological social media stakeholders and thus does not completely capture urology-related discussion on Twitter, resulting in an
EP
underestimation of activity for certain diseases or subspecialties. Although the tags in the UTO include some of the most common topics, excluding other tags may also result in loss of
AC C
fidelity regarding assessment of who and how contributes to urology-related conversations on Twitter. Similarly, although assessed for random basis, improper use or spamming of disease hashtags might have led to an overestimation of activity that represents meaningful conversation. Finally, Twitter is a rapidly changing and growing social media platform, indicating that this assessment is not generalizable beyond the study period or to other platforms. Notwithstanding, we found that the UTO hashtag list offers a structured mechanism for standardized assessment of Twitter activity in the urology community. Our work identifies
10
ACCEPTED MANUSCRIPT
significant opportunities for engagement of non-urological-oncology physician and patient communities. Future studies will need to investigate the relationship between Tweet activity
RI PT
data from UTO hashtags to academic benchmark metrics and to outcomes affecting clinical practice.
SC
Conclusion:
Twitter activity in the field of urology can be assessed using a standardized list of social
M AN U
media communication descriptors. Healthcare organizations and physicians appear to be the most active users of the Twitter platform, leaving significant opportunities for improved engagement of the patient and other healthcare provider communities. Our descriptive study indicates that many individuals and organizations already use Twitter to discuss urologyrelated health topics. More research is needed to identify how these online conversations
TE D
affect clinical care and patient outcomes. High volume of UTO-related activity suggests that communication via the Twitter platform harbours value for the urology community and its
AC C
EP
stakeholders.
11
ACCEPTED MANUSCRIPT
Acknowledgements: We thank all Twitter users who contributed to the online discussion on urology-related topics
RI PT
on Twitter. Conflict of Interest Statement None declared.
SC
Funding Sources
M AN U
None.
References
4. 5. 6. 7.
8. 9.
10. 11.
TE D
3.
EP
2.
Wilkinson, S. E., Basto, M. Y., Perovic, G. et al.: The social media revolution is changing the conference experience: analytics and trends from eight international meetings. BJU Int, 115: 839, 2015 Loeb, S.: Social media makes global urology meetings truly global. : The influence of Twitter. BJU Int, 115: 175, 2015 Loeb, S., Catto, J., Kutikov, A.: Social media offers unprecedented opportunities for vibrant exchange of professional ideas across continents. Eur Urol, 66: 118, 2014 Thangasamy, I. A., Leveridge, M., Davies, B. J. et al.: International Urology Journal Club via Twitter: 12-month experience. Eur Urol, 66: 112, 2014 Matta, R., Doiron, C., Leveridge, M. J.: The dramatic increase in social media in urology. J Urol, 192: 494, 2014 Chung, A., Woo, H.: Twitter in urology and other surgical specialties at global conferences. ANZ J Surg, 2015 Borgmann, H., DeWitt, S., Tsaur, I. et al.: Novel survey disseminated through Twitter supports its utility for networking, disseminating research, advocacy, clinical practice and other professional goals. Can Urol Assoc J, 9: E713, 2015 Murphy, D. G., Loeb, S., Basto, M. Y. et al.: Engaging responsibly with social media: the BJUI guidelines. BJU Int, 114: 9, 2014 Roupret, M., Morgan, T. M., Bostrom, P. J. et al.: European Association of Urology (@Uroweb) recommendations on the appropriate use of social media. Eur Urol, 66: 628, 2014 Kutikov, A., Woo, H. H., Catto, J. W.: Urology Tag Ontology Project: Standardizing Social Media Communication Descriptors. Eur Urol, 69: 183, 2016 [Accessed April 12, 2016] Symplur Signals: Healthcare category definitions. Avaiable at https://docs.google.com/spreadsheets/d/1HexI7X1KR0dPdFvJkd34DEtHF_18BGHzdnPNrauH NwI/edit?pref=2&pli=1#gid=0
AC C
1.
12
ACCEPTED MANUSCRIPT
18.
19.
20.
RI PT
17.
SC
16.
M AN U
15.
TE D
14.
EP
13.
[Accessed April 12, 2016] Katz, M. S.: The Oncology Tag Ontology: Professional–Centered Collaboration and Networking. Avaiable at http://www.symplur.com/blog/oncology-tagontology-professional-centered-collaboration-networking/ [Accessed April 12, 2016] Hawkins, M.: Radiology Hashtag Ontology Project. Avaiable at http://www.symplur.com/blog/radiology-hashtag-ontology-project/ Katz, M. S., Utengen, A., Anderson, P. F. et al.: Disease-Specific Hashtags for Online Communication About Cancer Care. JAMA Oncol: 1, 2015 Borgmann, H., Woelm, JH. Merseburger, A. et al.: Qualitative Twitter analysis of participants, tweet strategies, and tweet content at a major urologic conference. Can Urol Assoc J, 10: 3944, 2016 Flanigan, R. C.: Subspecialty certification in urology: a US perspective. Nat Clin Pract Urol, 3: 509, 2006 Debruyne, F. M.: Subspecialty certification in urology: a European perspective. Nat Clin Pract Urol, 3: 625, 2006 Attai, D. J., Cowher, M. S., Al-Hamadani, M. et al.: Twitter Social Media is an Effective Tool for Breast Cancer Patient Education and Support: Patient-Reported Outcomes by Survey. J Med Internet Res, 17: e188, 2015 Antonelli, J. A., Maalouf, N. M., Pearle, M. S. et al.: Use of the National Health and Nutrition Examination Survey to calculate the impact of obesity and diabetes on cost and prevalence of urolithiasis in 2030. Eur Urol, 66: 724, 2014 Lotan, Y., Buendia Jimenez, I., Lenoir-Wijnkoop, I. et al.: Primary prevention of nephrolithiasis is cost-effective for a national healthcare system. BJU Int, 110: E1060, 2012
AC C
12.
13
ACCEPTED MANUSCRIPT
Figure Legends Figure 1. Number of tweets and users of urology-related Twitter traffic stratified by urologic
RI PT
subspecialty (logarithmic scale). Figure 2. Geolocation of users contributing to urology-related Twitter traffic. The color tone
users and light tones countries with few or no users.
SC
reflects the number of users per country: Dark tones represent countries with large number of
stratified by healthcare category.
M AN U
Figure 3. Top 1000 influencers of urology-related Twitter traffic according to tweet volume
Figure 4. Bubble chart visualizing the 100 most frequently used words in tweets of urology-
AC C
EP
TE D
related Twitter traffic. Bubble size correlates to word prevalence/frequency in tweets.
ACCEPTED MANUSCRIPT Total Per Month Per Week 1,397,107,305 114,830,737 26,793,839 13,412 1,102 257 334,642 3.2 104,166
27,503 0.26 8,562
6,417 0.06 1,998
Per Day 3,827,691 36.7
Per Hour 159,487 1.5
917 0.009 285
38.2 0.0004 11.9
RI PT
Metric Impressions Impressions per user Tweets Tweets per user Users who tweeted
Table 1. Overview of quantitative metrics of urology-related Twitter traffic over a time period
AC C
EP
TE D
M AN U
SC
of 1 year.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Key of Definitions for Abbreviations Urology Tag Ontology
AC C
EP
TE D
M AN U
SC
RI PT
UTO
ACCEPTED Supplementary Table 1: Tweet activity according to specificMANUSCRIPT hashtags over a 1-year time period (QIV 2014 - QIII 2015) Hashtag
Disease
Tweets Impressions Users
Tweets/User
Impressions/User
Oncology Adrenal Cancer Bladder Cancer Bladder Cancer Kidney Cancer Kidney Cancer Prostate Cancer Penile Cancer Prostate Cancer Sarcoma Testicular Cancer Testicular Cancer Urologic Oncology Upper Tract Urothelial Cancer
29 9256 538 1836 8679 6571 28 94531 1994 21576 905 73 70
72398 40231983 1872897 4753283 54104871 22489523 57578 442379110 4263258 138545202 2100073 80471 95150
24 3828 358 1061 3393 2926 17 37472 600 9782 409 47 42
1.2 2.4 1.5 1.7 2.6 2.2 1.6 2.5 3.3 2.2 2.2 1.6 1.7
3017 10510 5232 4480 15946 7686 3387 11806 7105 14163 5135 1712 2265
Endourology Kidney Stones Endourology
139 10335 61
100281 36804074 67661
72 3711 40
1.9 2.8 1.5
1393 9918 1692
Female Pelvic Medicine & Reconstructive Surgery Incontinence Interstitial Cystitis Overactive Bladder
10 16567 865 3662
26811 47526084 1631458 15435903
9 5630 260 1558
1.1 2.9 3.3 2.4
2979 8442 6275 9908
2656 2411 22
3918185 15187315 59892
1113 1213 13
2.4 2.0 1.7
3520 12520 4607
1244
6866926
549
2.3
12508
10 35 11
21398 68956 24503
9 26 8
1.1 1.3 1.4
2378 2652 3063
12 4
9293 14373
10 4
1.2 1.0
929 3593
144 10758 261 7 315 460 2454 76 88963
102006 29907616 135742 20361 575912 690402 1139096 293163 434103705
46 3028 43 6 158 144 699 58 30182
3.1 3.6 6.1 1.2 2.0 3.2 3.5 1.3 2.9
2218 9877 3157 3394 3645 4794 1630 5055 14383
5 17 2 3711 43645
16531 47299 5315 7141044 66079829
5 7 2 869 10778
1.0 2.4 1.0 4.3 4.0
3306 6757 2658 8218 6131
Calculi #EndoUrology** #KidneyStones #UroStone*
Mayer-Rokitansky-Küster-Hauser Syndrome Pediatric Cancer Pediatric Urology
Transplant #KidneyTransplant*
Renal Transplant
Neurourology #NeuroUrol* #UroBPH* #UroUTI*
Neurourology Benign Prostatic Hyperplasia Urinary Infections
Uro Recon / Trauma #UroRecon* #UroTrauma*
Reconstructive Urology Urologic Trauma
Sexual Med. / Infertility
TE D
Urologic Epidemiology Urologic Health Policy & Economics Urologic Health Services Research International Urology Journal Club Urology
AC C
Others #UroEpi* #UroHPE* #UroHSR* #urojc #Urology
Andrology Erectile Dysfunction Hypogonadism Erectile Dysfunction Male Infertility Peyronies Disease Premature Ejaculation Priapism Sexually Transmitted Disease
EP
#Andrology* #erectiledysfunction*** #Hypogonadism* #MaleED* #MaleInfertility* #Peyronies #PrematureEjaculation* #Priapism* #STD
M AN U
Pediatrics #MRKH*** #pedcsm #PedUro*
SC
Female Urology #FPMRS* #Incontinence #InterstitialCystitis* #OAB*
RI PT
#adcsm #BladderCancer #blcsm #kcsm #KidneyCancer #PCSM #PenileCancer* #ProstateCancer #scmsm #TesticularCancer #tscsm #uroonc*** #utuc**
Supplementary Table S1:Tweet activity according to specific hashtags over a 1-year time period (QIV 2014 - QIII 2015). * 3-months time-period only: Tracking of this hashtag's Tweet activity has begun with QIII 2015 ** 6-months time-period only: Tracking of this hashtag's Tweet activity has begun with QII 2015 *** 9-months time-period only: Tracking of this hashtag's Tweet activity has begun with QI 2015
ACCEPTED MANUSCRIPT Geo-Location of contributors to the Twitter discussion in urology according to continents
1 2 3 4 5 6
Continent North America Europe Asia South America Africa Australia
n (of 63,219) % 34.796 55% 15,700 25% 4.847 8% 3.058 5% 2.501 4% 2.317 4%
RI PT
Rank
Geo-Location of countries of contributors to the Twitter discussion in urology according to countri
SC
n (of 63,219) % 28.314 45% 9.211 15% 4.884 8% 1.961 3% 1,460 0% 1.447 2% 1.222 2% 984 2% 970 2% 907 1% 866 1% 755 1% 665 1% 506 1% 475 1% 430 1% 395 1% 364 1% 325 1% 275 0% 268 0% 248 0% 243 0% 242 0% 240 0% 240 0% 238 0% 211 0% 198 0% 190 0% 187 0% 183 0% 171 0% 162 0%
M AN U
TE D
EP
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 25 26 27 28 29 30 31 32 33
Country United States of America United Kingdom Canada Australia India Russia Brazil France Germany Mexico Spain Indonesia South Africa Ireland Italy Nigeria Argentina Netherlands Philippines Venezuela Kenya New Zealand Colombia Belgium Slovenia Trinidad and Tobago Finland Switzerland Turkey Iran Pakistan Sweden Saudi Arabia Japan
AC C
Rank
ACCEPTED MANUSCRIPT
SC M AN U TE D
n % (of all tweets) 20.955 6% 19,610 6% 19.169 6% 19.165 6% 18.425 6% 18,370 5% 16.584 5% 13.505 4% 12,320 4% 12.159 4% 3% 11.181 10.592 3% 10.346 3% 10.344 3% 10,340 3% 10,160 3% 9.672 3% 9,350 3% 8.625 3% 8.444 3% 8.421 3% 2% 8.305 2% 6.875 2% 6,820 2% 6.533 2% 6.433 2% 6.247
EP
word prevent cancer follow men condom appreciate awareness patients life test risk prostate free therapy research know quality shows month help check preserves study health story reuters balls
AC C
Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
RI PT
Top 100 Words used in Tweets on urology-related topics
SC M AN U
2% 2% 2% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1%
TE D
5.647 5.208 5,170 4.865 4,840 4.788 4.675 4,620 4.606 4,400 4.364 4.339 4,290 4,290 4,180 4,180 4.125 4,070 4.064 4.046 3,960 3.889 3.659 3,630 3.612 3.575 3,520 3,520 3,490 3.465 3.465
EP
support text screening treatment sexual diagnosis common thx diagnosed people good pads early talk self stone supporting you're cause need quoted today october kidney learn monthly raise information use sex testing
AC C
28 29 30 31 32 33 34 35 36 37 38 39 40 40 41 41 42 43 44 45 46 47 48 49 50 51 52 52 53 54 54
RI PT
ACCEPTED MANUSCRIPT
SC M AN U
1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1%
TE D
3.408 3.405 3.355 3.216 3.135 3.135 3,080 3,080 3.025 2,970 2.915 2.895 2,860 2,860 2.814 2.763 2,750 2.741 2.723 2.695 2.695 2.695 2.677 2,640 2,640 2.585 2.585 2.505 2.475 2.475 2.439
EP
great fight reminder hiv join women think age including safe healthy testicular challenge big active video avoiding abena bladder std reduce expression thanks kub read sexually right delivery young care urology
AC C
55 56 57 58 59 59 60 60 61 62 63 64 65 65 66 67 68 69 70 71 71 71 72 73 73 74 74 75 76 76 77
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1%
RI PT
2,420 2,380 2.365 2.365 2.365 2.365 2.365 2.343 2.342 2.341 2,310
SC
researchers san tested online oct better linked results nominate htt aggressive
M AN U
78 79 80 80 80 80 80 81 82 83 84
AC C
EP
TE D
Supplementary Table S3: Top 100 Words used in Tweets on urology-related topics
ACCEPTED MANUSCRIPT
Top 10 tweets by physicians having the most retweets on urology-related topics Tweet text #BladderCancer is no longer just 1 disease - Paradigm shift #GU15 @ASCO Hospital nocturnal phone calls by specialty: #urology up there nzma.org.nz/__data/assets/pdf_… #prostatecancer diagnoses down 28% 1 yr after USPSTF grade D @JUrology @urogeek @UroCancerMD jurology.com/ar5cle/S0022-534… @LoebStacy at #EAU15 on #ProstateCancer: 'new Epstein/ISUP grade groups coming soon' @AmerUrological @EUplatinum Schroder receives a touching standing ovation at #AUA15 plenary for his contributions to #prostatecancer screening Walsh's "confession" on changing paradigm: shifting prostatectomy to high risk #ProstateCancer #pcwc15 Unusual pattern of #bladdercancer - "Chordoid"-like pattern. #gupath Moving toward cytoreductive prostatectomy for M1 #prostatecancer? Pros/cons by @DrDanielMoon-need more data #pcwc15 Clear cell RCC hemangioma-like areas, potential pitfall for renal mass biopsy ncbi.nlm.nih.gov/m/pubmed/2336… #kidneycancer .@theNCI Annual Report on Cancer: #prostatecancer had largest mortality decline of all cancers 1.usa.gov/1BLNNzg
RI PT
User name @mehrazinmd @loebstacy @danbarocas @karitikkinen @loebstacy @loebstacy @gleason4plus5 @loebstacy @williamson_sr @loebstacy
SC
number of retweets 28 28 26 26 23 22 21 17 16 16
M AN U
Rank 1 1 2 2 3 4 5 6 7 8
AC C
EP
TE D
Supplementary Table S4:Top 10 tweets by physicians having the most retweets on urology-related topics
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT