Analysis of Instagram® Posts Referring to Cleft Lip

International Journal of Environmental Research and Public Health

Article Analysis of Instagram® Posts Referring to Cleft Lip

Magdalena Sycinska-Dziarnowska 1, Piotr Stepien 2, Joanna Janiszewska-Olszowska 3,*, Katarzyna Grocholewicz 1, Maciej Jedlinski 1 , Roberta Grassi 4 and Marta Mazur 5

1 Department of Interdisciplinary Dentistry, Pomeranian Medical University in Szczecin, 70111 Szczecin, Poland; [email protected] (M.S.-D.); [email protected] (K.G.); [email protected] (M.J.) 2 Department of Technology and Education, Koszalin University of Technology, 75453 Koszalin, Poland; [email protected] 3 Department of General Dentistry, Pomeranian Medical University in Szczecin, 70111 Szczecin, Poland 4 Department of Surgical, Medical and Experimental Sciences, University of Sassari, 07100 Sassari, Italy; [email protected] 5 Department of Oral and Maxillofacial Sciences, Sapienza University of Rome, 00161 Rome, Italy; [email protected] * Correspondence: [email protected]

Received: 17 June 2020; Accepted: 8 October 2020; Published: 12 October 2020

Abstract: Background: Social media has become a source of medical information. Cleft lip and palate is a visible congenital anomaly. The aim of the study was to analyze Instagram® posts on the topic of cleft lip. Methods: Instagram® posts with “#cleftlip” from March 2014–March 2017 were accessed. Separate lists of expressions (hashtags, meaningful words, words with emojis or emojis alone) were prepared for primary posts and for replies. Thirty expressions statistically most frequent in primary versus secondary posts and 30 in secondary versus primary posts were identified (Group 1) as well as 30 English words or hashtags (Group 2), non-English words or hashtags (Group 3) and emojis (Group 4). The frequencies of expressions were compared (Z-test for the difference of two population proportions). Results: There were 34,129 posts, (5427 primary posts and 28,702 replies), containing 62,163 expressions, (35,004 in primary posts). The occurrence of all expressions was 454,162, (225,418 in primary posts and 228,744 in replies). Posts with positive expressions such as “beautiful”, “love”, “cute”, “great”, “awesome” occurred more often than these with negative ones. In replies all emojis were positive. Conclusions: Numerous Instagram® posts referring to cleft lip are published and do provoke discussion. People express their solidarity and sympathize with persons affected by cleft.

Keywords: cleft palate; internet; social media

1. Introduction Instagram® (Facebook, Inc., Menlo Park, CA, USA) is an online photo-sharing application and service, where users may share pictures or videos. It was launched in October 2010 as a free mobile application. On 21 December 2016 it was announced that its community has grown to more than 600 million users and the last 100 million of Instagrammers joined in the past six months [1]. On 26 April 2017 it was announced on Instagram® website that its popularity has grown to 700 million Instagrammers [1]. Since 2012, when 13% people used this service, a signiﬁcant increase of usage is observed. According to a national survey carried out between 7 March and 4 April 2016 on 1520 adults, about 32% of online adults (e.g., American people who currently use the Internet, according to the same survey it was 86% of the population) or 28% of all adult Americans report using Instagram®—roughly the same share as in 2015, when 27% of online adults used the application. Instagram® use was especially

Int. J. Environ. Res. Public Health 2020, 17, 7404; doi:10.3390/ijerph17207404 www.mdpi.com/journal/ijerph Int. J. Environ. Res. Public Health 2020, 17, 7404 2 of 11 high among younger adults; 59% Instagrammers were people between 18–29 year, whereas 8% only were older than 65. According to Pew Research Center women were more likely to use this service (38%) than men (26%). Half of Instagram® users access the platform daily, 35% of them several times a day [2]. Instagram® reﬂects major social trends, especially among young population. People use Instagram® for communication and entertainment or to express thoughts, moods and feelings. Nowadays, social media play an important role in searching for information by medical caregivers or patients who might look for answers to their questions online [3–6]. Several papers pertaining to the use of social media in the context of a medical problem (diabetes, Zika virus, arthroplasty) could be found. By interacting with other users, people with medical problems may provide or gain support and share information. Numerous social media network analyzers are available online. One of them is Netlytic [7]. It allows to automatically summarize and gather data from online conversations found on social media sites [7]. A few questionnaire studies on the use of social media in the context of cleft lip has been published in the recent years [8,9]. No studies based on social media surveillance for cleft lip could be found. The aim of the study was to analyze the frequency of individual meaningful words, emojis, emoticons or hashtags and to compare their frequency in Instagram® posts and their replies.

2. Materials and Methods

2.1. Instagram® Surveillance The analysis of the content of Instagram® posts was initiated by querying the hashtag #cleftlip using the Netlytic service (netylic.org, Toronto, ON, Canada), an open-sourced software. All tagged messages with the #cleftlip hashtag on Instagram® were downloaded and exported to a spreadsheet. The download was started on 17 February 2017; it enabled data collection from Instagram® search every hour until 4 March 2017. The data gathered covered all posts from public profiles published by Instagram®. The first captured post was published on 11 March 2014 and the last one on 4 March 2017. Thus, the data contained posts published for almost three years. The posts collected were divided into primary posts and secondary posts (replies). Primary posts consisted of pictures and their descriptions posted by Instagram® users. Secondary posts were responses to primary posts written by other users or by the author of primary post. Two separate lists of all expressions (hashtags, meaningful words, words combined with emojis or emojis alone) present in the posts were prepared for primary posts and for secondary ones. For the purpose of this article both emojis and emoticons were later stated as emojis. A spreadsheet macro was written to assign each expression the number of occurrences in all primary and secondary posts. Capitalization of letters was ignored as well as dots, commas, exclamation marks and question marks. Then, both lists were grouped into one, containing expressions sorted in descending order by the number of occurrences in primary posts (x1), and each expression was also assigned the number of occurrences in secondary posts (x2). The total list consisted of 62,163 expressions. The proportion pˆ1 of the number of occurrences of each expression in the total number of occurrences of all expressions in primary posts (n1) was calculated. A similar procedure was applied to secondary posts in order to calculate proportion pˆ2 from x2 and the total number of occurrences of all expressions in secondary posts (n2): x1 pˆ1 = (1) n1 x2 pˆ2 = (2) n2 On the basis of the calculated absolute value of the test statistic z (see the section “Statistical analysis”), 30 expressions statistically most frequent in the primary versus secondary posts were Int. J. Environ. Res. Public Health 2020, 17, x 3 of 12 Int. J. Environ. Res. Public Health 2020, 17, x 3 of 12 On the basis of the calculated absolute value of the test statistic z (see the section “Statistical On the basis of the calculated absolute value of the test statistic z (see the section “Statistical analysis”), 30 expressions statistically most frequent in the primary versus secondary posts were analysis”), 30 expressions statistically most frequent in the primary versus secondary posts were Int.identified J. Environ. and Res. Publictabularized. Health 2020 Simila, 17, x rly, 30 expressions statistically most frequent in the secondary3 of 12 identified and tabularized. Similarly, 30 expressions statistically most frequent in the secondary versus primary posts were identified. All together the 60 expressions have been placed in Table 1 and versusOn primary the basis posts of thewere calculated identified. absolute All together value th ofe 60the expressions test statistic have z (see been the placed section in Table“Statistical 1 and designated as Group 1 of expressions. The same procedure has been applied to English words or designatedanalysis”), 30as expressionsGroup 1 of expressions.statistically mostThe samefrequent procedure in the primaryhas been versusapplied secondary to English posts words were or hashtagsInt. J. Environ. (Group Res. Public2, Table Health 2),2020 non-English, 17, 7404 words or hashtags (Group 3, Table 3) and emojis (Group 4, 3 of 11 hashtagsidentified (Group and tabularized. 2, Table 2), Similanon-Englishrly, 30 words expressions or hashtags statistically (Group most 3, Table frequent 3) and in emojis the secondary (Group 4, Table 4). There were no emojis statistically more frequent found in primary versus secondary posts, Tableversus 4). primary There wereposts nowere emojis identified. statistically All together more frequent the 60 expressions found in primary have been versus placed secondary in Table posts,1 and thus Table 4 consists of 30 emojis from secondary posts only. thusdesignatedidentified Table 4 andas consists Group tabularized. of 1 30of emexpressions.ojis Similarly, from secondary The 30 same expressions postsprocedure only. statistically has been applied most frequent to English in thewords secondary or hashtags (Group 2, Table 2), non-English words or hashtags (Group 3, Table 3) and emojis (Group 4, hashtagsversusTable primary (Group 1. Group posts2, one—expressionsTable were 2), identified.non-English with Allthe words most together orstatistically ha theshtags 60 significant expressions(Group 3, difference Table have 3) beenin and the emojis placedfrequency (Group in of Table 4,1 and TableTable 4). There 1. Group were one—expressions no emojis statistically with the moremost statistically frequent found significant in primary difference versus in the secondary frequency of posts, designatedoccurrence. as Group 1 of expressions. The same procedure has been applied to English words or thus occurrence.Table 4 consists of 30 emojis from secondary posts only. hashtags (Group 2, Table2), non-English words or hashtags (Group 3, Table3) and emojis (Group 4, Expressions Dominating in Secondary Table4Expressions). There were Dominating no emojis statisticallyin Primary Posts more frequentExpressions found inDominating primary versus in secondarySecondary posts, TableExpressions 1. Group one—expressions Dominating in withPrimary the most Posts statistically Posts significant difference in the frequency of Nº Posts thusNº occurrence. Table4 consists of 30 emojis from secondary postsPosts only. Expression () Expression () Expression () Expression () Expression () Expression () Table 1. Group one—expressions with the most statisticallyExpressions significant Dominating difference in in theSecondary frequency 1 Expressions#cleftlip Dominating 4110 in Primary681 Posts53.136 beautiful 173 1647 32.753 1 Expressions#cleftlip Dominating 4110 in Primary681 Posts53.136 beautiful 173 1647 32.753 Nº1 of occurrence.#cleftlip 4110 681 53.136 Postsbeautiful 173 1647 32.753 2 #cleftpalate 1662 359 31.150 love 442 1791 26.535 2 #cleftpalateExpressionExpressions Dominating1662 in Primary359 () Posts31.150 ExpressionsExpressionlove Dominating442 in1791 Secondary ()26.535 Posts 3 Nº #cleftstrong 1459 425 25.889 thank 158 1172 26.239 31 #cleftstrong#cleftlipExpression 14594110x1 425681x 2 25.88953.136abs( z) Expressionbeautifulthank x1581731 11721647x2 26.23932.753abs(z ) 4 1#cleftlipandpalate #cleftlip 572 4110 55 681 21.823 53.136 beautifulcute 17365 1647848 24.609 32.753 42 #cleftlipandpalate#cleftpalate 1662572 35955 21.82331.150 cutelove 44265 1791848 24.60926.535 5 2#cleft #cleftpalate 756 1662 209 359 19.062 31.150 love😍 442104 1791835 22.528 26.535 53 #cleftstrong#cleft 1459756 209425 19.06225.889 thank😍 104158 1172835 22.52826.239 3 #cleftstrong 1459 425 25.889 thank 158 1172 26.239 6 #1in700 383 29 18.389 ❤ 84 630 19.268 64 4#cleftlipandpalate #cleftlipandpalate#1in700 383572 572 2955 5518.38921.823 21.823 cute cute❤ 658465 630848 848 19.26824.609 24.609 7 cleft 1254 575 17.881 great 124 701 18.835 75 5#cleftclef #cleftt 1254756 756 575209 209 17.88119.062 19.062 grea😍 t 104124104 701835 835 18.83522.528 22.528 8 #cleftcutie 768 247 17.841 nice 18 439 18.760 86 6#cleftcutie#1in700 #1in700 768383 383 24729 29 17.84118.389 18.389 nice❤ 841884 439630 630 18.76019.268 19.268 9 lip 907 344 17.570 awesome 34 478 18.634 97 7cleflip t cleft 1254907 1254 344575 575 17.57017.881 17.881 awesomegrea greatt 12412434 478701 701 18.63418.835 18.835 10 8#cleftproud #cleftcutie 563 768 168 247 15.867 17.841 thanks nice 1870 524 439 17.561 18.760 108 #cleftproud#cleftcutie 563768 168247 15.86717.841 thanksnice 7018 524439 17.56118.760 11 9#cleftie lip 284 907 25 344 15.552 17.570 awesomelindo 3414 373 478 17.389 18.634 119 #cleftielip 284907 34425 15.55217.570 awesomelindo 1434 373478 17.38918.634 10 #cleftproud 563 168 15.867 thanks 70 524 17.561 12 #smile 303 39 15.135 gorgeous 14 370 17.310 1210 11#cleftproud#smile #cleftie 303563 284 16839 25 15.13515.867 15.552 gorgeousthanks lindo 141470 370524 373 17.31017.561 17.389 13 #dogs 167 6 12.852 bless 11 360 17.277 1311 12#cleftie#dogs #smile 167284 303 256 3912.85215.552 15.135 gorgeouslindobless 141114 360373 370 17.27717.389 17.310 14 13#puppy #dogs 184 167 13 612.835 12.852 bless:) 1136 411 360 16.811 17.277 1412 #puppy#smile 184303 1339 12.83515.135 gorgeous:) 3614 411370 16.81117.310 15 14#love #puppy 281 184 60 1312.823 12.835 adorable :) 3638 416 411 16.807 16.811 1513 #dogs#love 281167 606 12.82312.852 adorablebless 3811 416360 16.80717.277 16 15#cleftbabies #love 155 281 5 6012.446 12.823 adorablesweet 38105 568 416 16.710 16.807 16 #cleftbabies 155 5 12.446 sweet 105 568 16.710 1614 16#cleftbabies#puppy #cleftbabies 155184 155 135 512.44612.835 12.446 sweet sweet:) 10510536 568411 568 16.71016.811 16.710 17 today 304 80 12.337 wow 5 319 16.657 1715 17today#love today 304281 304 8060 8012.33712.823 12.337 adorablewow wow 5385 319416 319 16.65716.807 16.657 18 #beautiful 150 5 12.225 😍😍😍 6 301 16.069 1816 18#cleftbabies#beautiful #beautiful 150155 150 5 512.22512.446 12.225 😍😍😍sweet 61056 301568 301 16.06916.710 16.069 19 #labioleporino 218 39 11.907 😘 16 323 15.867 1917 19#labioleporino #labioleporinotoday 218304 218 3980 3911.90712.337 11.907 wow😘 16165 323319 323 15.86716.657 15.867 20 mission 155 12 11.665 precious 14 292 15.125 2018 20#beautifulmission mission 155150 155 125 1211.66512.225 11.665 precious precious😍😍😍 14146 292301 292 15.12516.069 15.125 21 #plasticsurgery 208 38 11.564 😊 34 326 14.556 2119 21#plasticsurgery#labioleporino #plasticsurgery 208218 208 3839 3811.56411.907 11.564 😊😘 343416 326323 326 14.55615.867 14.556 22 22#dog #dog 134 134 7 711.246 11.246 god 8787 435 435 14.226 14.226 2220 mission#dog 134155 127 11.24611.665 preciousgod 8714 435292 14.22615.125 23 23#selfie #selfie 127 127 5 511.152 11.152 😍😍 99 238 238 13.880 13.880 2321 #plasticsurgery#selfie 127208 385 11.15211.564 😍😍😊 349 238326 13.88014.556 Int.24 J. Environ.24 Res.children Public children Health 2020, 17235, 235x 63 6310.760 10.760 amazing amazing 171171 584 584 13.820 13.8204 of 12 2422 children#dog 235134 637 10.76011.246 amazinggod 17187 584435 13.82014.226 25 25#surgery #surgery 159 159 23 2310.705 10.705 good good 155155 552 552 13.759 13.759 2625 #squishyfacecrew#surgery 148159 2123 10.37010.705 handsomegood 15534 292552 13.49413.759 2523 26#surgery #squishyfacecrew#selfie 159127 148 235 2110.70511.152 10.370 handsomegood😍😍 341559 552238 292 13.75913.880 13.494 2724 27childrensurgery surgery 507235 507 26463 264 10.02910.760 10.029 amazingcutie cutie 3117131 283584 283 13.44113.820 13.441 2825 28#surgerynew new 299159 299 11523 115 10.7059.980 9.980 good👍 91559 222552 222 13.34613.759 13.346 29 29#charity #charity 157 157 30 309.918 9.918 cool cool 1818 224 224 12.557 12.557 30 #babiesofinstagram 115 10 9.909 looks 59 306 12.085 30 #babiesofinstagram 115 10 9.909 looks 59 306 12.085 x1—number of occurrences in primary posts, x2—number of occurrences in secondary posts, abs(z)—absolute value of test statistic.—number of occurrences in primary posts, —number of occurrences in secondary posts, ()—absolute value of test statistic.

Table 2. Group two—English words or hashtags with the most statistically significant difference in the frequency of occurrence.

English Words or Hashtags Dominating in English Words or Hashtags Dominating Primary Posts in Secondary Posts Nº Word or Word or Hashtag () () Hashtag 1 #cleftlip 4110 681 53.136 beautiful 173 1647 32.753 2 #cleftpalate 1662 359 31.150 love 442 1791 26.535 3 #cleftstrong 1459 425 25.889 thank 158 1172 26.239 4 #cleftlipandpalate 572 55 21.823 cute 65 848 24.609 5 #cleft 756 209 19.062 great 124 701 18.835 6 #1in700 383 29 18.389 nice 18 439 18.760 7 cleft 1254 575 17.881 awesome 34 478 18.634 8 #cleftcutie 768 247 17.841 thanks 70 524 17.561 9 lip 907 344 17.570 gorgeous 14 370 17.310 10 #cleftproud 563 168 15.867 bless 11 360 17.277 11 #cleftie 284 25 15.552 adorable 38 416 16.807 12 #smile 303 39 15.135 sweet 105 568 16.710 13 #dogs 167 6 12.852 wow 5 319 16.657 14 #puppy 184 13 12.835 precious 14 292 15.125 15 #love 281 60 12.823 god 87 435 14.226 16 #cleftbabies 155 5 12.446 amazing 171 584 13.820 17 today 304 80 12.337 good 155 552 13.759 18 #beautiful 150 5 12.225 handsome 34 292 13.494 19 mission 155 12 11.665 cutie 31 283 13.441 20 #plasticsurgery 208 38 11.564 cool 18 224 12.557 21 #dog 134 7 11.246 looks 59 306 12.085 22 #selfie 127 5 11.152 luck 6 150 10.979 23 children 235 63 10.760 lovely 16 168 10.607 Int. J. Environ. Res. Public Health 2020, 17, 7404 4 of 11

Table 2. Group two—English words or hashtags with the most statistically signiﬁcant diﬀerence in the frequency of occurrence.

English Words or Hashtags Dominating in English Words or Hashtags Dominating in Nº Primary Posts Secondary Posts

Word or Hashtag x1 x2 abs(z) Word or Hashtag x1 x2 abs(z) 1 #cleftlip 4110 681 53.136 beautiful 173 1647 32.753 2 #cleftpalate 1662 359 31.150 love 442 1791 26.535 3 #cleftstrong 1459 425 25.889 thank 158 1172 26.239 4 #cleftlipandpalate 572 55 21.823 cute 65 848 24.609 5 #cleft 756 209 19.062 great 124 701 18.835 6 #1in700 383 29 18.389 nice 18 439 18.760 7 cleft 1254 575 17.881 awesome 34 478 18.634 8 #cleftcutie 768 247 17.841 thanks 70 524 17.561 9 lip 907 344 17.570 gorgeous 14 370 17.310 10 #cleftproud 563 168 15.867 bless 11 360 17.277 11 #cleftie 284 25 15.552 adorable 38 416 16.807 12 #smile 303 39 15.135 sweet 105 568 16.710 13 #dogs 167 6 12.852 wow 5 319 16.657 14 #puppy 184 13 12.835 precious 14 292 15.125 15 #love 281 60 12.823 god 87 435 14.226 16 #cleftbabies 155 5 12.446 amazing 171 584 13.820 17 today 304 80 12.337 good 155 552 13.759 18 #beautiful 150 5 12.225 handsome 34 292 13.494 19 mission 155 12 11.665 cutie 31 283 13.441 20 #plasticsurgery 208 38 11.564 cool 18 224 12.557 21 #dog 134 7 11.246 looks 59 306 12.085 22 #selﬁe 127 5 11.152 luck 6 150 10.979 23 children 235 63 10.760 lovely 16 168 10.607 24 #surgery 159 23 10.705 like 296 652 10.197 25 #squishyfacecrew 148 21 10.370 job 29 189 10.183 26 surgery 507 264 10.029 prayers 20 164 10.016 27 new 299 115 9.980 happy 277 611 9.886 28 #charity 157 30 9.918 super 36 197 9.871 29 #babiesoﬁnstagram 115 10 9.909 hi 10 125 9.384 30 #happy 114 10 9.855 lol 26 162 9.311

x1—number of occurrences in primary posts, x2—number of occurrences in secondary posts, abs(z)—absolute value of test statistic. Int. J. Environ. Res. Public Health 2020, 17, 7404 5 of 11

Table 3. Group three—non-English words or hashtags with the most statistically signiﬁcant diﬀerence in the frequency of occurrence.

Non-English Words or Hashtags Dominating in Primary Posts Nº Word or Hashtag Translation Language x1 x2 abs(z) 1 #labioleporino #cleftlip Spanish 218 39 11.907 2 #fissuralabiopalatina #cleftpalate Spanish 74 9 7.555 3 #maternidade maternity Portuguese 65 6 7.391 4 #maedeprimeiraviagem #firsttimemom Portuguese 52 7 6.213 5 idag today Swedish 54 9 6.035 6 #fissuradapelacecilia #ceciliasfissure Portuguese 50 9 5.691 7 #pormaissorrisosemvergonha #formoreshamelesssmile Portuguese 39 9 4.648 8 hoje today Portuguese 49 17 4.311 9 blev became Danish 31 7 4.176 10 #abogoiás #imfromgoias Portuguese/Spanish 37 12 3.892 11 bibir lip Indonesian 34 11 3.736 12 #maedemenina #motherofgirl Portuguese 30 9 3.649 13 labial lip Spanish 24 6 3.538 14 pappa dad Swedish 22 5 3.510 15 dudak lip Turkish 38 15 3.492 16 mamãe mom Portuguese 52 25 3.478 17 operasi operation Indonesian 45 21 3.325 18 bayi baby Indonesian 28 10 3.202 19 vamos come on Spanish 25 9 3.011 20 paladar palate Spanish 25 9 3.011 21 pacientes patients Spanish 20 6 2.979 22 fick got Swedish 30 13 2.892 23 hendido cleft Spanish 21 7 2.888 24 akan will Indonesian 30 14 2.715 25 dias days Portuguese 27 12 2.687 26 semanas week Portuguese 17 6 2.513 27 hari day Indonesian 27 13 2.502 28 labio lip Spanish 36 20 2.479 29 fissura fissure Portuguese 28 14 2.456 30 #associacaoreface #faceassociation Portuguese 60 41 2.348 1 lindo pretty Portuguese/Spanish 14 373 17.389 2 linda pretty Portuguese/Spanish 11 160 10.817 3 gracias thank you Spanish 15 141 9.536 4 deus God Portuguese 41 193 9.258 5 sehat healthy Indonesian 6 67 6.761 6 muito much Portuguese 78 197 6.436 7 coisa thing Portuguese 9 57 5.547 8 mais more Portuguese 117 222 4.878 9 amor love Portuguese/Spanish 26 80 4.785 Int. J. Environ. Res. Public Health 2020, 17, 7404 6 of 11

Table 3. Cont.

Non-English Words or Hashtags Dominating in Primary Posts Nº Word or Hashtag Translation Language x1 x2 abs(z) 10 terus continue Indonesian 6 37 4.436 11 anak child Indonesian 34 88 4.395 12 buat create Indonesian 7 33 3.829 13 selalu always Indonesian 9 37 3.826 14 certo of course Italian 8 31 3.404 15 mesmo same Portuguese 17 47 3.392 16 est east French/Italian 9 32 3.306 17 bara only Swedish 5 24 3.288 18 sempre all time Italian 33 71 3.269 19 obrigada thanks Portuguese 18 44 2.949 20 kasih love Indonesian 8 26 2.826 21 gran great Spanish 5 20 2.777 22 mejor best Spanish 8 25 2.703 Int.Int. J.J. Environ.Environ. Res.Res. PublicPublic HealthHealth 2020,, 17,, xx 7 of 12 Int. J. Environ. Res. Public Health 2020,, 17,, xx 7 of 12 23 mau want Indonesian 11 30 2.681 24 skrg now Indonesian 6 21 2.655 27 nasceu was born Portuguese 12 31 2.604 25 cara dear Italian 5 19 2.639 28 tenhotenho I have Portuguese 10 27 2.523 2628 querotenho I have wantPortuguese Portuguese 10 5 1927 2.6392.523 2729 nasceunya new was bornSwedish Portuguese 39 12 3170 2.6042.500 28 tenho I have Portuguese 10 27 2.523 30 sama same Indonesian 19 41 2.493 29 nya new Swedish 39 70 2.500 —number of occurrences in primary posts, —number of occurrences in primary posts, 30 sama—number of sameoccurrences Indonesianin secondary 19 41 2.493posts, —number of occurrences in secondary posts, () ( ) x1—number()—absolute of occurrences value in primary of test posts, statistic.x2—number of occurrences in secondary posts, abs z —absolute value of test statistic.

Table 4. Group four—emojis with the most statistically significant difference in the frequency of Table 4. Group four—emojis with the most statistically signiﬁcant diﬀerence in the frequency occurrence. of occurrence. Emojis Dominating in Secondary Posts Nº Emojis Dominating in Secondary Posts NºNº Emoji () EmojiEmoji x1 x2 abs (z) () 11 😍 104104 835835 22.528 22.528

22 ❤ 8484 630630 19.268 19.268 33 :):) 3636 411411 16.811 16.811

44 😍😍😍 66 301301 16.069 16.069 55 😘 1616 323323 15.867 15.867 66 😊 3434 326326 14.556 14.556 77 😍😍 99 238238 13.880 13.880 88 👍 99 222222 13.346 13.346 99 💕 5252 198198 8.531 8.531 1010 😃 66 9797 8.518 8.518 11 😄 5 93 8.452 12 ❤❤ 10 102 8.225 13 👌 10 94 7.785 14 ☺ 5 70 7.122 15 💖 5 67 6.931 16 😂😂😂 13 78 6.390 17 😉 10 70 6.311 18 💗 15 81 6.301 19 💜 9 67 6.266 20 😀 11 71 6.224 21 💙 65 171 6.214 22 🙌 13 72 5.990 23 ☺ 6 51 5.625 24 💕💕 5 47 5.505 25 😁 16 71 5.481 Int.Int. J.J. Environ.Environ. Res.Res. PublicPublic HealthHealth 20202020,, 1717,, xx 77 ofof 1212 Int.Int. J.J. Environ.Environ. Res.Res. PublicPublic HealthHealth 2020,, 17,, xx 7 of 12 Int.Int. J.J. Environ.Environ. Res.Res. PublicPublic HealthHealth 20202020,,, 1717,,, xxx 77 ofof 1212 27 nasceu was born Portuguese 12 31 2.604 27 nasceu was born Portuguese 12 31 2.604 2727 nasceunasceu waswas bornborn PortuguesePortuguese 1212 3131 2.6042.604 28 tenho I have Portuguese 10 27 2.523 28 tenhotenho II havehave Portuguese 10 27 2.523 2828 tenhotenho II havehave PortuguesePortuguese 1010 2727 2.5232.523 29 nya new Swedish 39 70 2.500 29 nya new Swedish 39 70 2.500 2929 nyanya newnew SwedishSwedish 3939 7070 2.5002.500 30 sama same Indonesian 19 41 2.493 30 samasama samesame IndonesianIndonesian 19 41 2.493 3030 samasama samesame IndonesianIndonesian 1919 4141 2.4932.493 —number of occurrences in primary posts, —number of occurrences in primary posts, —number of occurrences in primary posts, —number—number ofof occurrencesoccurrences inin primaryprimary posts,posts, —number of occurrences in secondary posts, —number of occurrences in secondary posts, —number—number ofof occurrencesoccurrences inin secondarysecondary posts,posts, () —absolute—number value of test statistic.of occurrences in secondary posts, ()—absolute value of test statistic. ()()—absolute—absolute valuevalue ofof testtest statistic.statistic. Table 4. Group four—emojis with the most statistically significant difference in the frequency of Table 4. Group four—emojis with the most statistically significant difference in the frequency of TableTable 4.4. GroupGroup four—emojisfour—emojis withwith thethe mostmost statisticallystatistically significantsignificant differencedifference inin thethe frequencyfrequency ofof occurrence.Table 4. Group four—emojis with the most statistically significant difference in the frequency of occurrence. occurrence.occurrence. Emojis Dominating in Secondary Posts EmojisEmojis DominatingDominating inin SecondarySecondary PostsPosts Nº Emojis Dominating in Secondary Posts Nº NºNº Emoji () Nº Emoji () EmojiEmoji ()() Emoji () 1 😍 104 835 22.528 1 😍 104 835 22.528 11 😍😍 104104 835835 22.52822.528 2 ❤ 84 630 19.268 2 ❤ 84 630 19.268 22 ❤❤ 8484 630630 19.26819.268 3 :) 36 411 16.811 3 :):) 36 411 16.811 33 :):) 3636 411411 16.81116.811 4 😍😍😍 6 301 16.069 4 😍😍😍 6 301 16.069 44 😍😍😍😍😍😍 66 301301 16.06916.069 5 😘 16 323 15.867 5 😘 16 323 15.867 55 😘😘 1616 323323 15.86715.867 6 😊 34 326 14.556 6 😊 34 326 14.556 Int. J. Environ. Res. Public Health 202066, 17, 7404😊😊 3434 326326 14.55614.556 7 of 11 7 😍😍 9 238 13.880 7 😍😍 9 238 13.880 77 😍😍😍😍 99 238238 13.88013.880 8 👍 9 222 13.346 8 👍 Table 4. Cont9 . 222 13.346 88 👍👍 99 222222 13.34613.346 9 💕 52 198 8.531 9 Emojis💕Dominating 52 in Secondary198 Posts 8.531 Nº99 💕💕 5252 198198 8.5318.531 10 😃 x 6 x 97abs z 8.518 10 Emoji😃 1 6 2 97 ( ) 8.518 1010 😃😃 66 9797 8.5188.518 1111 😄 55 9393 8.452 8.452 11 😄 5 93 8.452 1111 😄😄 55 9393 8.4528.452 1212 ❤❤ 1010 102102 8.225 8.225 12 ❤❤ 10 102 8.225 1212 ❤❤❤❤ 1010 102102 8.2258.225 1313 👌 1010 9494 7.785 7.785 13 👌 10 94 7.785 1313 👌👌 1010 9494 7.7857.785 14 ☺ 5 70 7.122 1414 ☺ 55 7070 7.122 7.122 1414 ☺☺ 55 7070 7.1227.122 15 💖 5 67 6.931 1515 💖 55 6767 6.931 6.931 1515 💖💖 55 6767 6.9316.931 16 😂😂😂 13 78 6.390 1616 😂😂😂 1313 7878 6.390 6.390 1616 😂😂😂😂😂😂 1313 7878 6.3906.390 17 😉 10 70 6.311 1717 😉 1010 7070 6.311 6.311 1717 😉😉 1010 7070 6.3116.311 18 💗 15 81 6.301 1818 💗 1515 8181 6.301 6.301 1818 💗💗 1515 8181 6.3016.301 19 💜 9 67 6.266 1919 💜 99 6767 6.266 6.266 1919 💜💜 99 6767 6.2666.266 20 😀 11 71 6.224 2020 😀 1111 7171 6.224 6.224 2020 😀😀 1111 7171 6.2246.224 21 💙 65 171 6.214 2121 💙 6565 171171 6.214 6.214 2121 💙💙 6565 171171 6.2146.214 22 🙌 13 72 5.990 22 🙌 13 72 5.990 222222 🙌🙌 131313 727272 5.990 5.9905.990 23 ☺ 6 51 5.625 23 ☺ 6 51 5.625 232323 ☺☺ 666 515151 5.625 5.6255.625 24 💕💕 5 47 5.505 24 💕💕 5 47 5.505 242424 💕💕💕💕 555 474747 5.505 5.5055.505 Int. J. Environ. Res. Public Health25 2020 , 17, x 😁 16 71 5.481 8 of 12 Int. J. Environ. Res. Public Health25 2020 , 17, x 😁 16 71 5.481 8 of 12 252525 😁😁 161616 717171 5.481 5.4815.481

2626 😂😂 1717 7070 5.267 5.267

2727 💛 66 4444 5.060 5.060 2828 ♡ 1313 5757 4.886 4.886 2929 😎 77 4141 4.599 4.599 3030 😂 55 3333 4.268 4.268 x x abs(z) 1—number of occurrences—number in primary posts,of 2—numberoccurrences of occurrences in secondaryin posts, primary—absolute valueposts, of test statistic. —number of occurrences in secondary posts, ()—absolute value of test statistic. Names, surnames, question marks, conjunctions, colons, pronouns and dashes were ignored as well as non-meaningful expressions. Emojis occurring one by one were considered as an independent Names, surnames, question marks, conjunctions, colons, pronouns and dashes were ignored as type of emoji. For example, three hearts that occurred one by one were treated independently, as a well as non-meaningful expressions. Emojis occurring one by one were considered as an independent more expressive form. Non-English expressions were recognized using available online dictionaries type of emoji. For example, three hearts that occurred one by one were treated independently, as a and translators (in most cases by Google Translator). If a non-English expression consisted of more expressive form. Non-English expressions were recognized using available online dictionaries several words (e.g., #pormaissorrisossemvergonha), they were separated (#por mais sorrisos sem and translators (in most cases by Google Translator). If a non-English expression consisted of several vergonha) and translated with Google Translator (#for more shameless smile) and then concatenated words (e.g., #pormaissorrisossemvergonha), they were separated (#por mais sorrisos sem vergonha) (#formoreshamelesssmile). and translated with Google Translator (#for more shameless smile) and then concatenated (#formoreshamelesssmile). 2.2.(#formoreshamelesssmile). Statistical Analysis

2.2.The Statistical frequencies Analysis of selected expressions in the posts and their replies were compared using a z-test for the diﬀerence of two population proportions. The test statistic were applied for expressions The frequencies of selected expressions in the posts and their replies were compared using a z- meeting the following conditions: min(n1pˆ1, n1(1 pˆ1)) 5 and min(n2pˆ2, n2(1 pˆ2)) 5. test for the difference of two population proportion− s.≥ The test statistic were− applied≥ for expressions min( ̂ , (1−̂ ))≥5 min( ̂ , (1−̂ ))≥5 meeting the following conditions: min(̂̂ ,(1−̂̂ ))≥5 and min(̂̂,(1−̂̂))≥5. : = The null hypothesis : = (the proportions in both statistical populations are equal) and the following alternative hypotheses were checked: : ≠ • two-tailed : ≠ (the proportions in both populations differ); : > : < • one-tailed : >or : <(proportion among the expressions occurring in primary posts was higher than the proportion of expressions occurring in secondary posts and the proportion of expressions occurring in primary posts was lower than the proportion of expressions occurring in secondary posts respectively). Three tests were made for the following hypotheses:

Hypotheses 1.H0 : p1 = p2, H1 : p ≠ p Hypotheses 1.H0 : p1 = p2, H1 : p1 ≠ p 2

Hypotheses 2.H : p = p H : p > p Hypotheses 2.H0 : p1 = p2, H1 : p1 > p2

∈ Hypotheses 3.H0 : p1 = p2, H1 : p1 ∈ p2

̂ >̂ Test 1 was performed for each expression, test 2 for expressions for which ̂̂ >̂̂, and test 3 for ̂ <̂ expressions for which ̂̂ <̂̂. For each expression the pooled proportion ̅̅ was calculated:

+ ̅̅ = ̅ = + (3) +

Value of test statistic was calculated as: Int. J. Environ. Res. Public Health 2020, 17, 7404 8 of 11

The null hypothesis H0 : p1 = p2 (the proportions in both statistical populations are equal) and the following alternative hypotheses were checked:

two-tailed H : p , p (the proportions in both populations diﬀer); • 1 1 2 one-tailed H : p > p or H : p < p (proportion among the expressions occurring in primary • 1 1 2 1 1 2 posts was higher than the proportion of expressions occurring in secondary posts and the proportion of expressions occurring in primary posts was lower than the proportion of expressions occurring in secondary posts respectively).

Three tests were made for the following hypotheses:

Hypothesis 1. H0 : p1 = p2,H1 : p1 , p2

Hypothesis 2. H0 : p1 = p2,H1 : p1 > p2

Hypothesis 3. H : p = p ,H : p p 0 1 2 1 1 ∈ 2

Test 1 was performed for each expression, test 2 for expressions for which pˆ1 > pˆ2, and test 3 for expressions for which pˆ1 < pˆ2. For each expression the pooled proportion p was calculated:

x + x p = 1 2 (3) n1 + n2

Value of test statistic was calculated as:

pˆ pˆ2 z = 1 − q (4) p(1 p) 1 + 1 − n1 n2 The level of statistical signiﬁcance was set at α = 0.05 and for each test p-value was calculated. By comparing the p-value with the statistical signiﬁcance level, it was stated whether there was enough evidence for rejecting H for H (p-value < α) or not (p-value α). The datasets used and analyzed 0 1 ≥ during the current study are available from the corresponding author on reasonable request.

3. Results The total number of posts downloaded was 34,129 including 5427 primary posts and 28,702 secondary posts. There were 62,163 expressions found in all posts, including 35,004 in primary posts. The total occurrence of all expressions was 454,162, including 225,418 in primary posts and 228,744 in secondary posts. The test statistic conditions have been met for n1 = 147,323 occurrences of expressions in primary posts and n2 = 161,270 in secondary posts. As a result, 59,812 expressions (96.22% of total) had to be removed due to statistically unitary character, so 2351 expressions were tested. Finally, there were 4702 tests performed: 2351 for two-tailed alternative hypothesis and 2351 for one-tailed alternative hypotheses. The summary of the number of test cases decisions for the entire population is presented in Table5. 1337 expressions were more frequent in primary posts and 1014 expressions were more frequent in secondary posts. As many as 1148 expressions were statistically significantly more frequent in primary or in secondary posts. However, 57.64% of the expressions were as often found in primary as in secondary posts, and 42.36% of expressions showed statistically significantly different frequencies. Out of the expressions that were more frequent in primary than in secondary posts in 51.53% there was no statistically significant dominance over the frequency of occurrences in replies, and 48.47% showed a statistically significantly higher rate of occurrence. Out of the expressions that occurred more frequently in secondary posts than in primary posts, in 50.69% cases there was no statistically significant dominance over the incidence of primary posts, and 49.31% had a statistically significant higher incidence. Int. J. Environ. Res. Public Health 2020, 17, 7404 9 of 11

Table 5. Summary of test results.

H0: p1 = p2 H0: p1 = p2 H0: p1 = p2 H1: p1 , p2 H1: p1 > p2 H1: p1 < p2 NE IE Total NE IE Total NE IE Total 1355 996 2351 689 648 1337 514 500 1014 57.64% 42.36% 100.00% 51.53% 48.47% 100.00% 50.69% 49.31% 100.00% IE—there is enough evidence to reject H0, NE—there is not enough evidence to reject H0.

It is seen that posts with expressions associated with positive emotions such as beautiful, love, cute, great, awesome occurred more often than with negative ones. In primary posts hashtags were much more frequent than in secondary posts. Among 30 expressions with the most significant difference in frequency of occurrences from the primary posts there were seven words that were not hashtags, for example: lip, cleft and surgery—so words related to the subject of our study. Hashtags used in the posts, help linking the photos and messages up to other subject on Instagram® featuring the same topic. In the first group of expressions in primary posts dominated hashtags and words associated with cleft lip problem. What was surprising there were hashtags related to animals for example #dog or #puppy. In the same group in secondary posts, expressions with positive meaning were overwhelming. Among non-English words or hashtags a domination of Portuguese and Spanish languages was evident. Also Indonesian language was frequent especially among non-English words from secondary posts. In group four, what was very surprising, no emojis were found that appeared more frequently in primary posts. The emojis appearing more frequently in secondary posts versus primary posts were the very popular ones, expressing emotions (heart, smile). In replies all emojis were highly positive.

4. Discussion This is the first study of Instagram® posts dealing with the subject of cleft lip. The present study is based on the largest number of Instagram® posts on a medical problem of all papers found. In the study by Fung et al. [10] 616 Instagram® posts with hashtag #zikavirus were manually coded. Karimkhani et al. [11] analyzed 50 newest posts referring to dermatology. According to thematic content Pila et al. [12] analyzed a sample of 600 Instagram® posts with hashtag #cheatmeal. Another study of 649 Instagram® posts on total knee arthroplasty and 638 posts on total hip arthroplasty was conducted at the Department of Orthopedic Surgery in Cleveland and Houston [6]. Tiggemann and Zaccardo [13] were looking for fitspiration hashtag on Instagram® regarding to body type and activity. A group of the first 600 posts with #fitspiration hashtag were coded for textual and photo content. Yi-Frazier et al. [14] asked twenty teenagers 14 to 18 years old with diabetes type 1 to use Instagram® and post photos on diabetes-related themes. Twelve participants were highly engaged; the whole study lasted for three consecutive weeks. In the study by Chung et al. [15] 16 women were interviewed who consistently shared and recorded on Instagram® what they had eaten. Another study gathered 476 social media posts tagged with #fitspo among the four platforms. Relevant 415 of 476 posts (87.2%) were analyzed. The majority of posts were accessed from Instagram® (360/415, 86.8%) [16]. Medical problems analyzed by previous studies were Zika virus, skin problems, problems with diabetes and in more detail concerning care of diabetes foot [3,5,10,11,17]. Very popular subjects concerning social media were associated with diet, fitness or nutrition [12,13]. Referring to orthodontics, three studies could be found, one—pertaining to patient experience to orthodontic brackets versus Invisalign® [18], another one was a qualitative analysis of orthodontic-related posts on Twitter [19] the last one showed how social media improve knowledge among patients with fixed orthodontic appliance [4]. Int. J. Environ. Res. Public Health 2020, 17, 7404 10 of 11

In the present study, hashtags dominated in primary versus secondary posts. It may be explained by the fact that use of hashtags helps the author of a message to link post with the group of desired subjects. Pew Research stated that Instagram® is very popular among non-white users. According to this demographic statistics in 2014 Hispanic origin people represent 34% of Instagram® online adult users in the United States of America [20]. This is visible in our study among non-English words or hashtags group (Table4) represented most often by Spanish and Portuguese languages. The statistically significant difference in the occurrence of emoji in secondary versus primary posts indicates that replies gave positive responses or comments. This means that many individuals expressed their solidarity and sympathized with persons affected by cleft. The authors find this aspect as very optimistic. What was remarkable during the review was the fact that there were posts pertained to animals (dogs, cats and a post on a squirrel). This shows that owners of animals with cleft problem also post on Instagram®. It is worth noticing that the number of replies (28,702) was much higher than of the primary posts (5427). This indicates a great interest in the posts concerning cleft lip. No comparison of primary versus secondary posts that might be used for comparison could be found in other studies analyzing social media. From the fact that for primary posts, the total occurrence of all expressions was 225,418 (in 5427 posts) and for secondary posts 228,744 (in 28,702 posts), we may assume that the replies were shorter. The possible explanation could be that the primary posts contained detailed descriptions and the replies were spontaneous (often short and emotional) reactions to them. Personalized medicine may create optimal treatment for the group of patients with cleft lip. Each cleft is unique in terms of its morphology. Every individual affected may suffer from different medical and psychological problems. It is evident that people affected by a cleft interact with one another via Instagram®. As social media become frequently used as a source of medical information, professionals should be aware of content available through Instagram® and consider using it as a means to provide health education. In the future, a more detailed surveillance of Instagram® posts with #cleftlip hashtag may help us better understand motivation, experiences and expectations of patients with clefts. In this way we may provide more accurate interdisciplinary and holistic treatment for affected persons.

5. Conclusions (1) Numerous Instagram® posts referring to cleft lip are published and provoke discussion. (2) In Instagram® posts two groups of meaningful expressions can be identified: one that appear more frequently in primary posts than in secondary posts and the other appearing more often in replies than in primary posts. (3) Expressions that occur more frequently in secondary posts than in primary ones do not contain offensive words, they are positive. People express their solidarity and sympathize with persons affected by cleft. (4) Hashtags occur more frequently in primary posts than in replies.

Author Contributions: Conceptualization, M.S.-D. and J.J.-O.; Methodology,M.S.-D.; Software, M.S.-D.; Validation, M.S.-D., P.S. and J.J.-O.; Formal Analysis, M.S.-D., P.S. and J.J.-O.; Investigation, M.S.-D.; Resources, M.S.-D.; Data Curation, M.S.-D., P.S.; Writing—Original Draft Preparation, M.S.-D. and J.J.-O.; Writing—Review & Editing, M.S.-D., J.J.-O., P.S., K.G., M.J., R.G. and M.M.; Visualization, M.S.-D.; Supervision, Not applicable; Project Administration, Not applicable; Funding Acquisition, Not applicable. All authors have read and agreed to the published version of the manuscript. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Conflicts of Interest: The authors declare no conflict of interest. Int. J. Environ. Res. Public Health 2020, 17, 7404 11 of 11

References

1. Instagram Press Page. 2017. Available online: https://instagram-press.com (accessed on 29 May 2017). 2. Pew Research Center. Social Media Update. 2016. Available online: http://www.pewinternet.org/2016/11/11/ social-media-update-2016/ (accessed on 5 April 2017). 3. Greene, J.A.; Choudhry, N.K.; Kilabuk, E.; Shrank, W.H. Online social networking by patients with diabetes: A qualitative evaluation of communication with Facebook. J. Gen. Int. Med. 2011, 26, 287–292. [CrossRef] [PubMed] 4. Al-Silwadi, F.M.; Gill, D.S.; Petrie, A.; Cunningham, S.J. Effect of social media in improving knowledge among patients having fixed appliance orthodontic treatment: A single-center randomized controlled trial. Am. J. Orthod. Dentofacial. Orthop. 2015, 148, 231–237. [CrossRef][PubMed] 5. Abedin, T.; Al Mamun, M.; Lasker, M.A.; Ahmed, S.W.; Shommu, N.; Rumana, N.; Turin, T.C. Social Media as a Platform for Information About Diabetes Foot Care: A Study of Facebook Groups. Can. J. Diabetes 2017, 41, 97–101. [CrossRef][PubMed] 6. Ramkumar, P.N.; Navarro, S.M.; Haeberle, H.S.; Chughtai, M.; Flynn, M.E.; Mont, M.A. Social Media and Total Joint Arthroplasty: An Analysis of Patient Utilization on Instagram. J. Arthroplast. 2017.[CrossRef] [PubMed] 7. Gruzd, A. Netlytic: Software for Automated Text and Social Network Analysis. 2016. Available online: http://Netlytic.org (accessed on 10 February 2017). 8. Greives, M.R.; Anderson, C.L.; Dean, R.A.; Scerbo, M.L.; Doringo, I.L.; Bebbington, M.W.; Teichgraeber, J.F. Survey of Parent Experiences in Prenatal Visits for Infants With Cleft Lip and Palate. Cleft Palate Craniofac J. 2016, 54, 668–673. [CrossRef][PubMed] 9. Khouri, J.S.; McCheyne, M.J.; Morrison, C.S. #Cleft: The Use of Social Media amongst Parents of Infants with Clefts. Cleft Palate Craniofac J. 2017, 55, 974–976. [CrossRef] 10. Fung, I.C.; Blankenship, E.B.; Goff, M.E.; Mullican, L.A.; Chan, K.C.; Saroha, N.; Duke, C.H.; Eremeeva, M.E.; Fu, K.W.; Tse, Z.T. Zika-Virus-Related Photo Sharing on Pinterest and Instagram. Disaster Med. Public Health Prep. 2017. [CrossRef][PubMed] 11. Karimkhani, C.; Connett, J.; Boyers, L.; Quest, T.; Dellavalle, R.P.Dermatology on Instagram. Dermatol. Online J. 2014, 20, 13030/qt71g178w9. 12. Pila, E.; Mond, J.M.; Griffiths, S.; Mitchison, D.; Murray, S.B. A thematic content analysis of #cheatmeal images on social media: Characterizing an emerging dietary trend. Int. J. Eat. Disord. 2017, 50, 698–706. [PubMed] 13. Tiggemann, M.; Zaccardo, M. Strong is the new skinny: A content analysis of #fitspiration images on Instagram. J. Health Psychol. 2016, 23, 1003–1011. [CrossRef][PubMed] 14. Yi-Frazier, J.P.; Cochrane, K.; Mitrovich, C.; Pascual, M.; Buscaino, E.; Eaton, L.; Panlasigui, N.; Clopp, B.; Malik, F. Using Instagram as a Modified Application of Photovoice for Storytelling and Sharing in Adolescents With Type 1 Diabetes. Qual. Health Res. 2015, 25, 1372–1382. [CrossRef][PubMed] 15. Chung, C.F.; Agapie, E.; Schroeder, J.; Mishra, S.; Fogarty, J.; Munson, S.A. When Personal Tracking Becomes Social: Examining the Use of Instagram for Healthy Eating. In Proceedings of the SIGCHI Conference Hum Factor Comput Systems, Denver CO, USA, 6–11 May 2017. [CrossRef] 16. Carrotte, E.R.; Prichard, I.; Lim, M.S. “Fitspiration” on Social Media: A Content Analysis of Gendered Images. J. Med. Internet Res. 2017, 19, e95. [CrossRef][PubMed] 17. Fu, K.W.; Liang, H.; Saroha, N.; Tse, Z.T.; Ip, P.; Fung, I.C. How people react to Zika virus outbreaks on Twitter? A computational content analysis. Am. J. Infect. Control 2016, 44, 1700–1702. [CrossRef][PubMed] 18. Noll, D.; Mahon, B.; Shroff, B.; Carrico, C.; Lindauer, S.J. Twitter analysis of the orthodontic patient experience with braces vs Invisalign. Angle Orthod. 2017, 87, 377–383. [CrossRef][PubMed] 19. Rachel Henzell, M.; Margaret Knight, A.; Morgaine, K.C.; Antoun, J.S.; Farella, M. A qualitative analysis of orthodontic-related posts on Twitter. Angle Orthod. 2014, 84, 203–207. [CrossRef][PubMed] 20. Pew Research Center. Social Media Update. 2014. Available online: http://www.pewinternet.org/2015/01/09/ social-media-update-2014/ (accessed on 5 April 2017).

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).