This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
Social media has become increasingly important as a source of information for the public and is widely used for health-related information. The outbreak of the coronavirus disease (COVID-19) has exerted a negative impact on dental practices.
The aim of this study is to analyze the nature and diffusion of COVID-19–related oral health information on the Chinese social media site Weibo.
A total of 15,900 tweets related to oral health and dentistry information from Weibo during the COVID-19 outbreak in China (December 31, 2019, to March 16, 2020) were included in our study. Two researchers coded 1000 of the total tweets in advance, and two main thematic categories with eight subtypes were refined. The included tweets were analyzed over time and geographic region, and coded into eight thematic categories. Additionally, the time distributions of tweets containing information about dental services, needs of dental treatment, and home oral care during the COVID-19 epidemic were further analyzed.
People reacted rapidly to the emerging severe acute respiratory syndrome coronavirus 2 threat to dental services, and a large amount of COVID-19–related oral health information was tweeted on Weibo. The time and geographic distribution of tweets shared similarities with epidemiological data of the COVID-19 outbreak in China. Tweets containing home oral care and dental services content were the most frequently exchanged information (n=4803/15,900, 30.20% and n=4478, 28.16%, respectively). Significant differences of public attention were found between various types of bloggers in dental services–related tweets (
Our study overviewed and analyzed social media data on the dental services and oral health information during the COVID-19 epidemic, thus, providing insights for government organizations, media, and dental professionals to better facilitate oral health communication and efficiently shape public concern through social media when routine dental services are unavailable during an unprecedented event. The study of the nature and distribution of social media can serve as a useful adjunct tool to help make public health policies.
The outbreak of the coronavirus disease (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), first identified from Wuhan, Hubei Province, China, has almost swept across the whole world and constituted a public health emergency of “pandemic” proportions [
COVID-19 has posed a particular threat to the practice of dentistry. With the identification of SARS-CoV-2 in the saliva of patients who are infected [
The internet, especially social media, is becoming increasingly important as a source of information for public health issues since it provides free and immediate access to large volumes of data [
Sina Weibo, similar to Twitter, is the most popular online microblog platform in China. Weibo allows its users to tweet or retweet messages optionally with links, pictures, or videos attached. The public reactions of Chinese people to the MERS-CoV and H7N9 outbreaks were significantly strong on Weibo [
A study of “COVID-19-related oral health information” tweets on Sina Weibo was performed. A new anonymous Weibo account was created with only the name, gender, and date of birth provided upon registration. Using a new account without search histories, previous likes, or friends can avoid preferential links promoted by Weibo. Four keywords related to COVID-19 (pneumonia of unknown cause, coronavirus, COVID-19, and epidemic) and two keywords for dentistry (stomatology and dentistry) in Chinese characters were employed to search tweets about COVID-19 and dentistry or oral health on Weibo. Eight independent searches with a combination of one keyword for COVID-19 and the other for dentistry were carried out on March 17, 2020, through the new account. We selected December 31, 2019, as the start date of tweets since, on this day, the pneumonia of unknown cause (the name of COVID-19 at the time) in Wuhan was officially reported to the World Health Organization, and the first group of epidemiologists were dispatched by the Chinese Center for Disease Control and Prevention (CCDC) to support the control of this emerging infectious disease (EID) in Wuhan.
Flow chart of the study. COVID-19: coronavirus disease.
Two researchers (ZT and GC) with expertise in dentistry completed the coding. First, the two coders were asked to pilot the project by coding 1000 (6.28%) of the total 15,900 tweets to develop and refine the coding schemes for thematic categories. Two main thematic categories for all tweets were determined initially: COVID-19–related and oral health–related information. The information related to COVID-19 was further subtyped into five domains: epidemiology, pathology, symptoms, diagnosis, and prevention; the information related to oral health was subtyped into three domains: dental services, needs of dental treatment, and home oral care information. The tweets that were related to oral health and COVID-19 but inappropriate to be sorted into any of the categories were labelled “others.” The definition and examples of each category are shown in
Second, to test the feasibility and reliability of categories, a weighed Kappa test was used to assess interrater and intrarater agreement of coding between two researchers. Two researchers were asked to classify 200 randomly selected tweets and reclassify 2 weeks after the first coding. The results of the weighed Kappa test of the two researchers were 0.983 for interrater agreement and 0.994 (ZT) and 0.983 (GC) for intrarater agreement, which indicated excellent reliability of coding procedure.
Third, after agreement for the coding of the tweets was confirmed, all 15,900 tweets were randomly separated into two groups and classified by two researchers. The tweets were coded to more than one category if containing miscellaneous information. Tweets in each thematic category were divided into two types according to the numbers of followers (less than 1000 followers and 1000 or more followers).
The tweets of each day were counted and compared with COVID-19 daily new cases and deaths in China for time distribution analysis. Among 15,900 tweets, 1682 tweets with location information were analyzed for geographic distribution and compared with the regional distribution of total COVID-19 cases by March 16, 2020. The epidemiological data were obtained from the official website of the CCDC [
Meanwhile, the tweets related to oral health information were further analyzed. Specifically, the time distribution of tweets containing information about the risks of COVID-19 transmission during dental procedures, notices of stopping all or part of dental services, need for dental treatment, home oral care, protective measures during dental services, and notices of restoring dental services during the COVID-19 epidemic were analyzed. The public needs for dental treatment and home oral care information tweeted on Weibo were further categorized and counted. The Kruskal-Wallis test was used to determine the differences of public reactions (numbers of likes, shares, and comments) to tweets from different types of bloggers (governments, media, dental clinics or hospitals, dentists or dental nurses, online health platforms, and others). In addition, the Mann-Whitney U test was used to compare the public reactions to tweets related to oral health information tweeted by the same types of bloggers.
All statistical analyses were carried out with Microsoft Excel (Microsoft Corporation) and SPSS software 18.0 (SPSS Inc), and
As shown in
Further analyses were conducted on 1682 tweets whose geographic distribution could be identified and compared to the regional distribution of total cases. The geographic location was optional for the users when posting tweets on Weibo. Therefore, only 1682 tweets included geographic location information. As demonstrated in
Time distribution of tweets and new cases and deaths of COVID-19 in China. COVID-19: coronavirus disease.
The geographic distributions of Weibo tweets (left) and total coronavirus disease cases (right) in the region.
Among 15,900 tweets included in our study, 79.81% (n=12,690) were oral health–related information and 38.86% (n=6180) contained background knowledge of COVID-19. As shown in
Some information was frequently tweeted or retweeted on Weibo and the top five pieces of widely diffused information were selected for evaluating the public reactions (
Thematic distributions of tweets with COVID-19–related oral health information. COVID-19: coronavirus disease.
Public reactions to highly tweeted information on Weibo.
Most highly tweeted information | Count, n | Public reactions | ||
|
|
Likes, mean (SD) | Shares, mean (SD) | Comments, mean (SD) |
The news propagandizing aerosol as a transmission route of COVID-19a | 1406 | 321.28 (11,225.52) | 24.85 (581.45) | 19.70 (539.29) |
Risks of COVID-19 spread by dental clinics due to the aerosol created by dental handpieces | 659 | 4.72 (56.91) | 1.92 (14.34) | 1.31 (7.29) |
Refutation of the misinformation that gargling with saltwater or mouthwash can prevent COVID-19 | 468 | 105.99 (1010.50) | 27.27 (312.07) | 9.98 (62.31) |
Refutation of the misinformation that eating garlic can kill the novel coronavirus in the oral cavity | 389 | 50.25 (639.83) | 17.62 (237.45) | 7.60 (84.71) |
Refutation of the misinformation that oral spray/disinfectants can prevent COVID-19. | 372 | 0.25 (1.88) | 0.43 (2.74) | 0.05 (0.32) |
aCOVID-19: coronavirus disease.
The time distributions of tweets related to misinformation for COVID-19 prevention and its refutations. COVID-19: coronavirus disease.
Tweets with different types of oral health–related information were distributed differently during the COVID-19 epidemic. As shown in
When the dental services were not available from the end of January to early March, many bloggers complained of oral problems and sought dental care on Weibo. There were a steady number of tweets (around 40-80 tweets/day;
Interestingly, numerous tweets with home oral care content were found on Weibo when the majority of dental services were not available for the public, much more than the tweets seeking oral health care as previously mentioned (4803 vs 2973 tweets). Information about daily oral care, how to deal with dental emergencies at home, and online consultation services shared similar proportions among these tweets (n=1684/4803, 35.06%; n=2092, 43.56%; and n=2029, 42.24%, respectively;
As for the public responses to dental services and home oral care–related tweets from different types of bloggers, the number of likes, shares, and comments for tweets from governments, media, dental clinics and hospitals, dentists and dental nurses, online health platforms, and other nondental bloggers were counted and analyzed (only bloggers with >1000 followers were included;
Time distributions of oral health–related tweets during the COVID-19 epidemic. COVID-19: coronavirus disease.
Needs for dental treatment during COVID-19 epidemic.
Needs of dental treatment | Number of tweets (n=2793), n (%)a |
Toothache or wisdom tooth problem | 1132 (40.53) |
Oral ulcer | 264 (9.45) |
Orthodontic problem | 536 (19.19) |
Implants or prostheses | 31 (1.11) |
Pediatric oral diseases | 81 (2.90) |
Oral cancer | 41 (1.47) |
Others or not specific | 788 (28.21) |
aThe sum value of all parts is over 100% because some tweets mentioned more than one need of dental treatment.
Thematic distributions of tweets with home oral care information.
Comparison of public reactions to tweets with dental services and home oral care information from different types of bloggers.
Public reactions | Blogger categories | ||||||||||||
|
Governments | Media | Dental clinics/hospitals | Dentists/dental nurses | Online health platform | Others | |||||||
|
|||||||||||||
|
DSa | 492 (16.05) | 805 (26.26) | 563 (18.37) | 363 (11.84) | 0 (0) | 842 (24.47) | ||||||
|
HOCb | 98 (2.95) | 87 (2.62) | 1049 (31.60) | 874 (26.33) | 660 (19.88) | 551 (16.6) | ||||||
|
|||||||||||||
|
DS, mean (SD) | 11.91 (68.00) | 54.36 (1053.84) | 1.66 (8.21) | 2.47 (47.20) | N/Ac | 11.47 (78.58) | ||||||
|
HOC, mean (SD) | 14.35 (128.38) | 168.10 (1110.52) | 1.27 (11.09) | 26.82 (377.06) | 4.05 (64.39) | 5.69 (38.82) | ||||||
|
–6.858 | –5.49 | –3.259 | –0.896 | N/A | –14.75 | |||||||
|
<.001 | <.001 | .001 | .37 | N/A | <.001 | |||||||
|
|||||||||||||
|
DS, mean (SD) | 2.75 (8.44) | 4.63 (33.37) | 0.88 (5.75) | 0.49 (9.37) | N/A | 2.37 (14.81) | ||||||
|
HOC, mean (SD) | 1.94 (8.27) | 32.48 (182.30) | 0.88 (6.75) | 5.34 (42.89) | 1.28 (17.34) | 3.81 (49.31) | ||||||
|
–5.943 | –5.234 | –0.565 | –1.045 | N/A | –10.06 | |||||||
|
<.001 | <.001 | .57 | .30 | N/A | <.001 | |||||||
|
|||||||||||||
|
DS, mean (SD) | 3.97 (15.79) | 7.43 (74.11) | 0.70 (2.59) | 0.42 (8.08) | N/A | 4.15 (23.85) | ||||||
|
HOC, mean (SD) | 2.87 (23.22) | 25.36 (154.88) | 0.33 (1.56) | 5.55 (51.73) | 0.59 (8.85) | 2.39 (12.42) | ||||||
|
–5.925 | –3.81 | –3.279 | –0.236 | N/A | –12.68 | |||||||
|
<.001 | <.001 | .001 | .81 | N/A | <.001 |
aDS: dental services.
bHOC: home oral care.
cNot applicable.
Since the COVID-19 outbreak exerted a negative impact on dental practices [
It was interesting to note that the time distribution of the Weibo data was approximately consistent with the trend of both daily new cases and deaths. Our finding is similar to a previous study on H7N9–related tweets on Weibo, which identified a positive correlation between the number of daily tweets and the cumulative case fatality rate of H7N9 [
Additionally, the peaks of tweets were influenced by some milestone events of COVID-19, including the official confirmation of human-to-human transmission of COVID-19, lockdown of Wuhan, and the death of Dr Wenliang Li. The possible explanation for this phenomenon is that social media engages more public attention during outbreaks of EIDs, especially when important news about the epidemic is released [
Although our results show that the geographical distribution of the tweets was roughly consistent with the distribution of total COVID-19 cases, Hubei Province was an exception. As the first epicenter of a newly identified infectious disease in the world, people may have been overwhelmed, and the concerns for oral health problems and demands for dental services were unavoidably suppressed. This phenomenon should attract more attention from public health policy makers.
Social media should not only provide true and useful information for the public, but also possess self-correction function for misinformation [
Social media can amplify the spread of contents compared to traditional mass media [
The analysis of the COVID-19–related oral health contents on Weibo provided an overview of the supply and demand of dental services during the COVID-19 outbreak. The public concern about the risks of spreading coronavirus by aerosols was disseminated earlier than the official notices of stopping part of or all dental services. With insufficient dental services, the public needs for oral care could not be satisfied, and an increased number of users complained of oral diseases or sought for consultation on Weibo. Our study showed dynamic changes of information related to supply and demand of dental services during this period, indicating that social media can serve as a useful tool for the monitoring of medical and health demands during an unprecedented time [
Social media plays an increasingly important role in health policy making [
In response to perceived unmet dental care needs from the online community, home oral care information was highly tweeted on Weibo to satisfy the public needs for both daily oral care and dental emergencies. Remote dental consultations were also achieved through Weibo, not only providing diagnosis and suggestions for patients with oral health problems but also avoiding risks of coronavirus transmission. This correlation between dental service disruption and an increased use of social media as a means of communication is not uncommon during a disaster or emergency [
In our study, the online community reacted variously to the tweets from different bloggers. The tweets posted by governments and media attracted more responses from the public due to their authority [
Several limitations of our study should be acknowledged. First, there was inevitable bias of information in the data collection process. For example, Weibo is more popular among young people rather than the aged and is more accessible for economically developed regions. Therefore, the impact of age and regional distribution of users may need to be considered when interpreting the results. As a real time social media, some tweets were deleted by bloggers and some bloggers’ accounts were suspended by Sina Weibo, which led to direct information loss. In the geographic analysis of tweets, around 90% did not provide location information, thus affecting the overall objectivity and accuracy of results to a certain extent. Second, our study did not provide any data on the characteristics of Weibo users who viewed and shared these tweets, and the complete diffusion route of tweets were not extracted and analyzed. Therefore, the audience of the information and the diffusion scale of tweets could not be accurately evaluated. Third, the information provided by online consultation services was not available, and thus, the quality of online consultations and the effects on the patients remained unknown.
To the best of our knowledge, this is the first study to comprehensively overview and analyze social media data on the dental services and oral health information during the COVID-19 epidemic in China. Based on our results, it is evident that social media users reacted immediately to the emerging SARS-CoV-2 threat to dental practices. Social media not only contributed to public health surveillance and policy making but also served as a bridge between oral health information providers and the patients. The findings illustrate the relationship between social media information with the supply and demand of dental services during the outbreak of the COVID-19 epidemic in China. In addition, the study provides insights for government organizations, media, and dental professionals to efficiently affect and shape public awareness, and disseminate dental public health information through social media.
Thematic distribution of tweets related to oral health/dentistry during the coronavirus disease epidemic, December 31, 2019, to March 16, 2020.
Chinese Center for Disease Control and Prevention
coronavirus disease
emerging infectious disease
Middle East respiratory syndrome–related coronavirus
severe acute respiratory syndrome coronavirus 2
This work was supported by the University of Hong Kong research output prize 2019.
None declared.