2012_Lulic_Analysis of Emergency Physicians' Twitter Accounts

7/31/2019 2012_Lulic_Analysis of Emergency Physicians' Twitter Accounts

http://slidepdf.com/reader/full/2012lulicanalysis-of-emergency-physicians-twitter-accounts 1/7

Analysis of emergency physicians’ Twitter accounts

Ileana Lulic,1 Ivor Kovic2

ABSTRACTBackground Twitter is one of the fastest growing socialmedia networks for communication between users viashort messages. Technology proficient physicians havedemonstrated enthusiasm in adopting social media fortheir work.Objective To identify and create the largest directory ofemergency physicians on Twitter, analyse their useraccounts and reveal details behind their connections.Methods Several web search tools were used toidentify emergency physicians on Twitter withbiographies completely or partially written in English.NodeXL software was used to calculate emergencyphysicians’ Twitter network metrics and createvisualisation graphs.Results The authors found 672 Twitter accounts of self-

identified emergency physicians. Protected accountswere excluded from the study, leaving 632 for furtheranalysis. Most emergency physicians were located inUSA (55.4%), had created their accounts in 2009(43.4%), used their full personal name (77.5%) andprovided a custom profile picture (92.2%). Based on atleast one published tweet in the last 15 days, there were345 (54.6%) active users on 31 December 2011. Activeusers mostly used mobile devices based on the Appleoperating system to publish tweets (69.2%).Visualisation of emergency physicians’ Twitter networkrevealed many users with no connections with theircolleagues, and a small group of most influential userswho were highly interconnected.Conclusions Only a small proportion of registeredemergency physicians use Twitter. Among them existsa smaller inner network of emergency physicians withstrong social bonds that is using Twitter’s full potentialsfor professional development.

INTRODUCTIONSocial media has become an integral part of personal and professional lives of many.1 2  Variablerates of physicians using social networkingwebsites have been reported.3e5 However, it seemsthat those physicians who are comfortable with

computers and mobile technology are highly engaged with social media.4

TwitterTwitter is one of the most popular social medianetworks, which became available in July 2006, andhas since seen a tremendous growth, reaching 100million active users in 2011.6 e8 Twitter allowscommunication between users via short messageslimited to 140 characters, commonly called tweets.Users can publish tweets, using computers ormobile devices, which show up in the streams of the people who are subscribed to them. People who

subscribe to receive tweets from a user are calledfollowers, while the people followed by a user are

referred to as friends. Each Twitter user has a profilepage, which can be customised with a short biog-

raphy, website address, location information,custom profile picture and background.9

Goals of our study Individual emergency physicians (EPs) have com-mented on the benefits of using Twitter.10

However, little is known about the number of EPsusing Twitter, as well as their usage characteristicsand connections. The goal of our study was toidentify EPs on Twitter, analyse their profiles andreveal details behind their mutual connections.

METHODSIdentifying EPs’ Twitter accounts

In order to find EPs’ accounts we used severalstrategies. First, we searched users’ names andbiographies using the web based people searchprovided by Twitter and FollowerWonk andTwiangulate tools.11 12 Search was performed usingall combinations of many variations of keywordslike emergency and emergency medicine withphysician and doctor. After identifying influentialEPs, their followers’ biographies were also exam-ined to find more of their colleagues. The same wasdone with Twitter accounts of organisations andjournals related to emergency medicine, like thosefrom the American College of Emergency Physi-

cians (@EmergencyDocs) and Emergency MedicineJournal (@EmergencyMedBMJ).

Utilising these strategies, we were able to iden-tify 672 Twitter accounts of self-identified EPs withbiographies completely or partially written inEnglish. Among these accounts, 42 (6%) wereprotected. Users of protected accounts do not sharetheir tweets publicly, and were therefore excludedfrom the study. The remaining 632 (94%) Twitteraccounts were included in the study.

Analysing EPs’ twitter accountsThe FollowerWonk tool was used to extract severalparameters of all Twitter accounts.11  Variables

including full name, biography, location informa-tion, custom profile picture, web address in theprofile, number of followers, number of friends,date of account creation, date of last tweet andtotal number of tweets were exported into Mi-crosoft Excel (Microsoft, Redmond, WA, USA).

 Another web service, TweetStats, was used toreveal information about users’ preferred platformand day for publishing tweets.13

Visualising EPs’ twitter accounts and theirinterconnectionsTwiangulate tool and NodeXL software were used

to identify connections among EPs’

Twitter ac-counts.12 14 NodeXL was further used to calculate

1Institute of EmergencyMedicine of Primorsko-goranskaCounty, Emergency Medical

Service Rijeka, Rijeka, Croatia2Institute of EmergencyMedicine of Istria County,Emergency Medical ServicePazin, Pazin, Croatia

Correspondence toDr Ivor Kovic, Zavod za hitnumedicinu Istarske zupanije,Hitna medicinska pomoc Pazin,Jurja Dobrile 1, Pazin 52 000,Croatia; ivorkovic@gmail.com

Accepted 28 April 2012

Lulic I, Kovic I. Emerg Med J (2012). doi:10.1136/emermed-2012-201132 1 of 6

Original article EMJ Online First, published on May 25, 2012 as 10.1136/emermed-2012-201132

Copyright Article author (or their employer) 2012. Produced by BMJ Publishing Group Ltd under licence.

group.bmj.comon May 25, 2012 - Published by emj.bmj.comDownloaded from 

7/31/2019 2012_Lulic_Analysis of Emergency Physicians' Twitter Accounts

http://slidepdf.com/reader/full/2012lulicanalysis-of-emergency-physicians-twitter-accounts 2/7

EPs’ Twitter network metrics and create visualisation graphs forbetter understanding of account features and their interconnec-tions based on the following.14

Statistical analysisPearson’s c2 test was used to assess the significance of an asso-ciation between two categorical variables. Continuous variableswere examined for normality of distribution by examination of 

histograms and tests for skewness, rejecting the null hypothesisthat the data is normally distributed. Non-parametric rank sumtests were therefore used to reveal differences between groups.Pearson’s rank correlation coef ficient was used to measure thestrength of a relationship between two continuous variables. Allstatistical values were considered significant when the p valuewas<0.05. Statistical analysis of data was performed using SPSSStatistics for Mac, V.20.0 (IBM Corporation).

RESULTSPrincipal characteristics of 632 EPs’ Twitter accounts arepresented in table 1The largest number of accounts was created in April 2009,

which is shown in figure 1. Among the users who providedlocation information (N¼522, 82.6%), most were located in USA (N¼350, 55.4%). However, the proportion of EPs openingTwitter accounts outside USA has been increasing. In 2007 every twelfth account was opened outside USA, while in 2011 nearly every second account came from another country (8.3% vs47.5%, c2

4¼27.76, p¼0.001). The average Twitter account was23.5612.7 months old. Accounts of EPs from all countriescombined, except from USA (N¼172, 32.9%), were newer thanthose from USA alone (20612.6 vs 26 612.4, U¼21 689,p¼0.001).

Twitter public profile customisationUsers who provided a custom profile picture were more likely toalso provide a website address on their profile (99% vs 88.2%,c

21¼23.07, p¼0.001). The average length of user biography 

was 87645.6 characters, constituting on average 54% of allowed160 characters. Users who included additional information intheir biographies about work (U¼23740, p¼0.001), family (U¼23981, p¼0.01), hobbies (U¼33074, p¼0.001), location

(U¼30322, p¼0.047) or a disclaimer (U¼1840, p¼0.027), hadsignificantly longer biographies. EPs mentioning additional factsabout their work in biography were less likely to reveal infor-mation about their families (12.2% vs 19.1%, c

21¼4.36,

p¼0.037) or hobbies (20.2% vs 34%, c21¼13.99, p¼0.001), but

more likely to provide a full personal name (85.6% vs 74.3%,c

21¼9.56, p¼0.002) and a website address (41.1% vs 21%,

c2

1¼29.49, p¼0.001) on their profile. On the other hand, usersmentioning family information more often shared their hobbiesin their biography (23.5% vs 13%, c2

1¼11.69, p¼0.001) and wereless likely to provide a website address on their profile (13.3% vs19.4%, c2

1¼3.99, p¼0.046).

Twitter activity 

The main features of EPs’ Twitter activity, as well as differencesbetween active and non-active users’ accounts are presented intable 2.

Based on at least one published tweet during the last 15 days,there were 345 (54.6%) EPs actively using Twitter on 31December 2011. Among the users who opened their accounts in2007 and 2008, 70% were active, compared with 54% of theusers with accounts created in 2010 and 2011 (c2

4¼10.53,p¼0.032). A larger proportion of users from USA were not activewhen compared with EPs from all other countries (74.4% vs25.6%, c

21¼10.61, p¼0.001). Active users more frequently 

mentioned facts about hobbies in their biographies (61.9% vs49.9%, c2

1¼8.85, p¼0.003), as well as more often included

custom pictures (97.1% vs 86.4%, c2

1¼

25.03, p¼

0.001) andwebsite addresses (64.3% vs 48.6%, c2

1¼14.87, p¼0.001) ontheir profile. Those physicians mentioning work information intheir biographies were more often not active in the last 15 days(48.1% vs 57.2%, c2

1¼4.35, p¼0.037). When it comes to the number of followers, those EPs

mentioning work in their biographies had a higher total numberof followers (U¼35885, p¼0.017) and followers among otherEPs (U¼35633, p¼0.008). Users mentioning only their hobbiesalso had a higher total number of followers (U¼38837, p¼0.001), but not a higher number of followers among EPs(U¼43 699, p¼0.069). A significantly lower number of followersamong EPs was found for users mentioning family informationin their biography (U¼24626, p¼0.025).

Users with more followers on Twitter were also themselvesfollowing more users (r¼0.98, p¼0.001). The same was also trueamong EPs, where those who were following more of their peers,also had more of them following back (r¼0.66, p¼0.001). EPswith older accounts had more followers overall (r¼0.10,p¼0.013), more followers among their colleagues (r¼0.19,p¼0.001), more friends in total (r¼0.08, p¼0.046) and amongEPs (r¼0.092, p¼0.021). They also had a higher total number of tweets (r¼0.152, p¼0.001), and their last tweet was publishedmore recently (r¼0.173, p¼0.001). EPs who were following(r¼0.175, p¼0.001) or were followed (r¼0.205, p¼0.001) by more of their colleagues, had a higher total number of tweets(r¼0.124, p¼0.002) and their last tweet was published more

recently (r¼

0.161, p¼

0.001), when compared with EPs with lessconnections to and from their peers.

Table 1 Principal characteristics of emergency physicians’ Twitteraccounts (N¼632)

Characteristics Number (%)

 Year created

2007 14 (2.2)

2008 69 (10.9)

2009 274 (43.4)

2010 124 (19.6)

2011 151 (23.9)

Country

USA 350 (55.4)

UK 38 (6.0)

Canada 24 (3.8)

Australia 23 (3.6)

Others 87 (13.7)

No data 110 (17.4)

Full personal name

 Yes 490 (77.5)

Custom profile picture

 Yes 583 (92.2)

Website address on the profile

 Yes 241 (38.1)

Information in the biography

Work 181 (28.6)

Family 108 (17.1)

Hobby 247 (39.1)

Location 138 (21.8)

Religion 18 (2.8)

Disclaimer 10 (1.6)

2 of 6 Lulic I, Kovic I. Emerg Med J (2012). doi:10.1136/emermed-2012-201132

Original article

 group.bmj.comon May 25, 2012 - Published by emj.bmj.comDownloaded from 

7/31/2019 2012_Lulic_Analysis of Emergency Physicians' Twitter Accounts

http://slidepdf.com/reader/full/2012lulicanalysis-of-emergency-physicians-twitter-accounts 3/7

Further analysis of solely active users revealed that amongthese EPs, mobile devices were most popular for publishing onTwitter (figure 2). Early adopters, who created their accounts in

2007, predominantly used desktop applications to post tweets(desktop applications 45.5%, web 36.4%, mobile devices 18.2%),while the percentage of users mostly using mobile devices rosethroughout the years and peaked among those with accountsopened in 2011 (desktop applications 7.5%, web 32.6%, mobiledevices 60%) (c2

2¼23.37, p¼0.001). In addition to owning theoldest accounts, users primarily using desktop applications hada higher number of total followers (c2

2¼25.4, p¼0.001),followers among EPs (c2

2¼10.75, p¼0.005), friends (c22¼6.92,

p¼0.031) and total number of tweets (c22¼16.75, p¼0.001),

than users using other platforms to publish on Twitter. Apple iOS devices (Apple Inc., Cupertino, CA, USA) were

used most often by those predominantly using mobile devices to

post tweets (iPhone: N¼105, 67.3%; iPad: N¼12, 1.9%). Mobiledevices based on Android operating system (Google, Mountain

 View CA, USA) were used more often by the US based EPs

compared with their peers from all other countries (23.6% vs3.4%, c22¼12.57, p¼0.002). On the contrary, physicians outside

USA used BlackBerry mobile devices (Research In Motion,Ontario, Canada) more often (15.5% vs 5.6%, c2

2¼12.57, p¼0.002). Users were most active on Twitter during the middle of the week (figure 3).

Interconnections between EPs on TwitterFeatures and interconnections of 632 EPs’ accounts are visualisedin figure 4. Each Twitter account is represented with a circle orsquare. An account is represented with a circle if its user isfollowing less than 50 EPs within the network, or with a square,if its user is following more than 50 EPs. Each individual account

Figure 1 Distribution of number of Twitter accounts created per month, from January 2007 to December 2011.

Table 2 Usage characteristics of emergency physicians’ Twitter accounts and differences between active and non-active users

Characteristics

Median (range) Stat

All (N[632) Active (N[345) Non-active (N[287) U p Value

Account age (months) 26 (0e58) 27 (0e58) 26 (0e57) 47 275 0.328*

Last tweet (days) 11 (0e2283) 2 (0e15) 155 (16e2283) 0.000 0.001y

Followers (N) 43.5 (0e69 187) 77 (2e69 187) 20 (0e8829) 25 591 0.001y

Friends (N) 66.5 (0e73 734) 104 (1e73 734) 28 (0e9347) 23 950 0.001y

Friends am ong eme rge ncy p hysicians (N) 1 (0e165) 1 (0e165) 0 (0e53) 35 772 0.001y

Friends of emergency physicians (N) 1 (0e126) 2 (0e126) 0 (0e46) 36 713 0.001y

Total tweets (N) 81.5 (0e38 862) 286 (1e38 862) 20 (0e11 308) 14 920 0.001y

Tweets per month (N) 5.16 (0e1615) 15.63 (0e1138) 1.17 (0e1615) 13 972 0.001y

*Manne

Whitney U test reveals no significant difference between active and non-active accounts.yManneWhitney U test reveals significant difference between active and non-active accounts.

Lulic I, Kovic I. Emerg Med J (2012). doi:10.1136/emermed-2012-201132 3 of 6

Original article

 group.bmj.comon May 25, 2012 - Published by emj.bmj.comDownloaded from 

7/31/2019 2012_Lulic_Analysis of Emergency Physicians' Twitter Accounts

http://slidepdf.com/reader/full/2012lulicanalysis-of-emergency-physicians-twitter-accounts 4/7

has a colour assigned to it on a continuous scale from green tored, depending on the number of followers among EPs. The sizeof the accounts represents the total number of followers. Linesconnecting accounts represent connections between them onTwitter. The darker line signifies that one user is following theother. On the contrary, the lighter line signifies that theconnection is reciprocated and that both the users are followingeach other.

There were 213 (34%) accounts, depicted on the edge of thefigure 4, which were isolated from the Twitter network of EPs,meaning that they had no incoming or outgoing connectionswith other EPs. Most of these users were following a small

number of Twitter users, and had a small number of followersand total tweets. Some, on the other hand, had a very largenumber of followers, but all of them came from outside thenetwork of EPs. In fact, among 35 users with more than 1000total followers, there were 7 (20%) without a single incoming oroutgoing connection with another colleague. On the contrary,

18 accounts in the middle of the graph had numerous connec-tions, at least 50, coming in from other EPs. These were themost influential users in the network, who were also greatly interconnected. Some of these users are represented withsquares, meaning that they had more outgoing connections withother colleagues outside the small group of influential users. Themost distinguishing feature of most influential EPs was the factthat the great majority of them (N¼15, 83%) maintained theirown medical blogs.

DISCUSSION

Principal findings We managed to identify almost 700 EPs on Twitter with biog-raphies written in English. All of them are followed by ourresearch Twitter account (@research_er) and also listed ona dedicated web page.9 15 In order to facilitate future research inthis subject, we will continue to add new EPs to the list, as well

Figure 2 Distribution of number ofactive emergency physicians primarilyusing different platforms to publishtweets.

Figure 3 Distribution of number ofactive emergency physicians primarilypublishing tweets on different days ofthe week.

4 of 6 Lulic I, Kovic I. Emerg Med J (2012). doi:10.1136/emermed-2012-201132

Original article

 group.bmj.comon May 25, 2012 - Published by emj.bmj.comDownloaded from 

7/31/2019 2012_Lulic_Analysis of Emergency Physicians' Twitter Accounts

http://slidepdf.com/reader/full/2012lulicanalysis-of-emergency-physicians-twitter-accounts 5/7

as expand it with those tweeting in languages other thanEnglish. A majority of the EPs in our sample were from USA.They represent only 1.6% of all board-certified EPs in USA,which are estimated to be more than 20 000.16  The same is truefor Australia which has approximately 1330 EPs,17 but only 1.7%were found on Twitter. According to the latest available statis-tics, 77% of adults in USA use the internet, and 13% of them useTwitter.18 19 If we assume the same for EPs in USA, there shouldbe around 17 000 of them using the internet, and 2210 usingTwitter. Based on these results, we can conclude that EPs aremore reluctant to adopt Twitter than the general adult popu-

lation, or that there are many more currently using this servicewho were not identified by us. However, these numbers aresubstantially larger than those presented by Chretien andcolleagues.20 In 2010 they identified 523 physicians tweetingprimarily in English, with EPs accounting for only 6.2%.20 If thisratio of EPs versus physicians of all specialties on Twitter iscorrect, there might currently be more than 11 000 physiciansusing Twitter. This is a significantly greater number than the1327 physicians presently listed in the largest directory of physicians on Twitter,21 but is also most likely to be under-estimated. Furthermore, the number of EPs on Twitter seemsto be steadily increasing, although not as fast as the numberof all users.22

 When comparing the main characteristics of EPs’

Twitteraccounts with accounts of all users, EPs were more likely to

provide location information (82.6% vs 73%) and web address(62% vs 45%) in their profile.22 This might suggest they are morecomfortable disclosing information about themselves on Twitterthan the average user.

 Among the active users, mobile devices were the most popularplatform for publishing tweets. Usage of mobile devices hasshown great growth during the years, which is consistent withthe data for all Twitter users.23 Interestingly, users with theoldest accounts were predominantly using desktop applications.It might be the case that desktop users, who are more active andinfluential, chose desktop applications because of their benefits

over other platforms, such as better integration with otherapplications on the computer. Apple iOS mobile phones andtablets were by far most popular among those EPs using mobiledevices to tweet. Such a high dominance of iPhones and iPadsamong physicians over other mobile devices has also beenreported by others.24

 When the Twitter network of EPs was analysed and visualisedfor mutual connections, we found three groups of users. Thefirst group consisted of users with no connections to other EPsin the network. It is most likely the case that these users, someof which were very active and had a large numbers of followersand friends, were not using Twitter for work, but were ratherusing it to connect with families and friends who might share

similar hobbies and interests. Such users distinguished them-selves in Twitter biographies by providing information about

Figure 4 Visualisation of emergencyphysicians’ Twitter accounts featuresand their interconnections.

Lulic I, Kovic I. Emerg Med J (2012). doi:10.1136/emermed-2012-201132 5 of 6

Original article

 group.bmj.comon May 25, 2012 - Published by emj.bmj.comDownloaded from 

7/31/2019 2012_Lulic_Analysis of Emergency Physicians' Twitter Accounts

http://slidepdf.com/reader/full/2012lulicanalysis-of-emergency-physicians-twitter-accounts 6/7

their work less frequently and hobbies more frequently thanthose EPs with more connections with their colleagues. ThoseEPs who provided more information about their work ina biography were more often non-active. However, our criteriafor distinguishing active users was based merely on days sincethe last tweet, so it might be the case that these users were usingTwitter to listen to other users to a greater extent, than to sharetheir own tweets. The second group was represented by 2.8% of 

most influential EPs in the network who had many followersamong their colleagues. These users formed a small innernetwork characterised by many mutual connections. Most of them also followed EPs outside this small group, but there wereonly a few who were very generous in terms of following morethan 50 of their colleagues. They represented a bridge betweenthe most influential and all other users in the Twitter network of EPs. The vast majority of most influential users among EPs werealso medical bloggers, who previously reported sharing practicalknowledge and skills as one of their major motivators for blog-ging.25 They seem to be using their blogs to gain more followerson Twitter and vice versa. Finally, the third and largest group of EPs had a mixture of friends outside and inside the Twitternetwork of EPs, which might indicate that they were dividingattention between work related and personal topics.

Limitations and future studiesIdentifying the types of users on Twitter is challenging. Twittersearch does not offer many advanced options and is known tofail at times. Since at the time of research no directories of EPson Twitter existed, biographies shared by users on their profilesoffered the only means of identification. However, some userschoose to leave their biographies empty or do not mention theirprofession at all. Therefore many EPs potentially using Twittermight have passed unnoticed. On the other hand, since users arethe only ones responsible for writing their biographies, somemight have been falsely identified as EPs. Our study can there-

fore draw conclusions solely regarding Twitter accounts of userswho self-identified as EPs, and cannot offer information aboutthose accounts maintained by EPs who chose not to disclosetheir profession. Furthermore, we included only Twitter userswith biographies completely or partially written in the Englishlanguage, so we cannot draw conclusions about EPs usingTwitter in other languages. Users with protected accounts werealso excluded from our analysis, and we have not provided any information about their use of Twitter. In this study we did notanalyse the content of the tweets published by EPs. Such futureanalysis could reveal more information about the way Twitter isused among EPs, as well as provide more insight regarding theirmutual social connections. Visualisation of the network of conversations might offer more detailed information about

characteristics and strength of social bonds between EPs onTwitter. Finally, usage characteristics of other popular socialmedia networks by EPs, including Facebook,26  should also bestudied in depth to reveal distinctive communication patternsand identify their potentials for emergency medicine.

CONCLUSIONSThere were more EPs on Twitter than expected based on avail-able data. However, this number still represents only a smallfraction of registered EPs. Among the Twitter network of EPsthere was a small group of influential users using its fullpotential to create strong mutual connections and use Twitter

for their work. Hopefully other EPs will soon realise the benefitsof Twitter and join their colleagues for the sake of advancingemergency medicine.

Contributors IL contributed to the concept and design of the study and wrote thepaper. IK conceived the study, performed statistical analysis and edited the paper.

Competing interests None.

Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES1. Social media. Wikipedia, the Free Encyclopedia. http://en.wikipedia.org/wiki/ 

Social_media, http://www.webcitation.org/64dwbVP7e (accessed 12 Jan 2012).2. Madden M, Zickuhr K. 65% Of Online Adults Use Social Networking Sites.

Washington, DC: Pew Internet & American Life Project, 2011. http://www.pewinternet.org/ w /media//Files/Reports/2011/PIP-SNS-Update-2011.pdf, http:// www.webcitation.org/64dwZ6s5g (accessed 12 Jan 2012).

3. Bosslet GT, Torke AM, Hickman SE, et al . Patient-doctor relationship online socialnetworks results a national survey. J Gen Intern Med  2011:1168e74.

4. Modahl M, Tompsett L, Moorhead T. Doctors, Patients & Social Media. Waltham,MA: QuantiaMD, 2011. http://www.quantiamd.com/q-qcp/ DoctorsPatientSocialMedia.pdf, http://www.webcitation.org/64dwUz7x2 (accessed12 Jan 2012).

5. Manhattan Research. Taking the Pulse! U.S. v11.0 Study . New York, NY:Manhattan Research, 724 LLC, 2011.

6. About Twitter. Twitter . https://twitter.com/about, http://www.webcitation.org/ 

64dwCzMu2 (accessed 12 Jan 2012).7. Twitter. Wikipedia, the Free Encyclopedia. http://en.wikipedia.org/wiki/Twitter,http://www.webcitation.org/64dwHQ7Uy (accessed 12 Jan 2012).

8. Twitter Blog: One Hundred Million Voices. http://blog.twitter.com/2011/09/one-hundred-million-voices.html, http://www.webcitation.org/64dwLAAFm (accessed12 Jan 2012).

9. EM Research on Twitter . https://twitter.com/#!/research_er, http://www.webcitation.org/64e2drJ7d (accessed 12 Jan 2012).

10. Berger E. This sentence easily would fit on twitter: emergency physicians arelearning to “tweet”. Ann Emerg Med  2009;54:23e5.

11. Follower Wonk: Twitter Analytics, Follower Segmentation, Social Graph Tracking, And  More. http://followerwonk.com/, http://www.webcitation.org/64hTY2xC6 (accessed15 Jan 2012).

12. Twiangulate: Analyzing The Connections Between Friends And Followers . http:// twiangulate.com/search/, http://www.webcitation.org/64hTd4gKn (accessed 15 Jan2012).

13. TweetStats: Graphin’ Your Stats. http://tweetstats.com/, http://www.webcitation.org/64hTiy6SC (accessed 15 Jan 2012).

14. NodeXL: Network Overview, Discovery and Exploration for Excel . http://nodexl.codeplex.com/, http://www.webcitation.org/64hTnjpZq (accessed 15 Jan 2012).

15. TwittER researchER: Directory of Emergency Physicians on Twitter . http:// emergencytwitter.ivor-kovic.com/, http://www.webcitation.org/64iW0FyOg(accessed 15 Jan 2012).

16. Camargo CA Jr, Ginde AA, Singer AH, et al . Assessment of emergencyphysician workforce needs in the United States, 2005. Acad Emerg Med 2008;15:1317e20.

17. The Australasian College for Emergency Medicine (ACEM). http://www.acem.org.au,http://www.webcitation.org/64grg41Rz (accessed 14 Jan 2012).

18. International Telecommunication Union. Yearbook of StatisticsdTelecommunication/ICT Indicators 2001-2010. 37th edn. Genevea, Switzerland, 2011.

19. Madden M, Zickuhr K. Twitter Update 2011. Washington, DC: Pew Internet &American Life Project, 2011. http://pewinternet.org/ w /media//Files/Reports/2011/ Twitter%20Update%202011.pdf, http://www.webcitation.org/64gc27PRp (accessed14 Jan 2012).

20. Chretien KC, Azar J, Kind T. Physicians on twitter. JAMA 2011;305:566e8.

21. Twitter Doctors. http://twitterdoctors.net, http://www.webcitation.org/64eA5Dcra(accessed 12 Jan 2012).22. Sysomos Inc. Twitter Statistics For 2010. http://www.sysomos.com/insidetwitter/ 

twitter-stats-2010/, http://www.webcitation.org/64gdfArwD (accessed 14 Jan2012).

23. Giga OM. Twitter CEO: We Have 100M Active Users . http://gigaom.com/2011/09/ 08/twitter-ceo-we-have-100m-active-users/, http://www.webcitation.org/64h4cnkzF(accessed 14 Jan 2012).

24. Hirsch R. Physician Mobile Use Grows 45%; Apple Dominates Android And  Blackberry . Reston VA: Bulletin Healthcare, 2011. http://www.bulletinhealthcare.com/PressReleases/MobileData.pdf, http://www.webcitation.org/64h7y2ILJ(accessed 14 Jan 2012).

25. Kovic I, Lulic I, Brumini G. Examining the medical blogosphere: an online survey ofmedical bloggers. J Med Internet Res 2008;10:e28.

26. Welcome to Facebook. Facebook . http://www.facebook.com/, http://www.webcitation.org/665wsvax2 (accessed 12 Jan 2012).

PAGE fraction trail=6

6 of 6 Lulic I, Kovic I. Emerg Med J (2012). doi:10.1136/emermed-2012-201132

Original article

 group.bmj.comon May 25, 2012 - Published by emj.bmj.comDownloaded from 

7/31/2019 2012_Lulic_Analysis of Emergency Physicians' Twitter Accounts

http://slidepdf.com/reader/full/2012lulicanalysis-of-emergency-physicians-twitter-accounts 7/7

doi: 10.1136/emermed-2012-201132published online May 25, 2012Emerg Med J 

Ileana Lulic and Ivor KovicaccountsAnalysis of emergency physicians' Twitter

 http://emj.bmj.com/content/early/2012/05/24/emermed-2012-201132.full.html

Updated information and services can be found at:

These include: 

References http://emj.bmj.com/content/early/2012/05/24/emermed-2012-201132.full.html#ref-list-1

This article cites 4 articles

P<P Published online May 25, 2012 in advance of the print journal.

serviceEmail alerting

the box at the top right corner of the online article.Receive free email alerts when new articles cite this article. Sign up in

Notes

(DOIs) and date of initial publication.publication. Citations to Advance online articles must include the digital object identifiercitable and establish publication priority; they are indexed by PubMed from initialtypeset, but have not not yet appeared in the paper journal. Advance online articles areAdvance online articles have been peer reviewed, accepted for publication, edited and

 http://group.bmj.com/group/rights-licensing/permissions

To request permissions go to:

 http://journals.bmj.com/cgi/reprintform

To order reprints go to:

 http://group.bmj.com/subscribe/

To subscribe to BMJ go to:

 group.bmj.comon May 25, 2012 - Published by emj.bmj.comDownloaded from