Table 1.
Summary of the general characteristics of the selected studies.
| References | Country | Topic | Campaign | Method of evaluation | Results (main outcomes) | Key limitations |
|---|---|---|---|---|---|---|
| Schlichthorst et al. (27) | Australia | Suicide | “Man Up” campaign, which links masculinity and suicide (hashtags #MANUP, #ABCMANUP, #LISTENUP and #SPEAKUP) | Twitter statistics (followers, likes, retweets and impressions metrics) | Hashtags grew substantially during the campaign broadcast. The most frequent content was related to help-seeking, masculinity and expression of emotions. Very effective in disseminating information and promoting real-time conversations | Metrics Tweet screening Biased sample |
| Harding et al. (28) | Ghana | Breastfeeding | Breastfeed4Ghana Campaign | Online cross-sectional survey (n = 451) | Acceptability was high but 61% of the audience did not remember the purpose of the campaign. Exposure was not associated with increased breastfeeding awareness | Metrics Survey limitations |
| Castillo et al. (29) | Canada | Dementia | Dissemination of digital content on pain in dementia, with the hashtag ##SeePainMoreClearly | Twitter statistics (metrics and impressions) | Hashtag received more than 5,000,000 impressions and was used in 31 countries. There was a greater number of posts on the topic during the campaign broadcast period | Metrics Lack of post-test |
| Grantham et al. (30) | Canada | Nutrition | Campaign of a dietician for 16 weeks, with the hashtag #eatwellcovid19 | Twitter statistics (metrics and follower testimonials) | Two types of followers: those who appreciated listening to stories submitted by followers, and those who appreciated evidence-based information | Metrics Campaign design |
| Moukarzel et al. (31) | World | Breastfeeding | World Breastfeeding Week 2020 (WBW) Campaign | Social network analysis (users and topics of conversation) | Increased conversation during the campaign. Formation of identifiable communities based on geolocation, interests and profession. Identification of influencers as a “bridge” between the public and the scientific community | Lack of behavioral assessment |
| Viguria et al. (32) | Spain | Eating disorders | Eating Disorder Awareness Week and Wake Up Weight Watchers campaigns, through #wakeupweightwatchers, #eatingdisorderawarenessweek, #eatingdisorderawareness, and #EDAW | Twitter statistics (impressions of collected and sorted tweets) | During the campaign there were more tweets about the topic, comparing the official hashtags with the control hashtag, which is used throughout the year (#eatingdisorder). Medical and awareness content was low. A large percentage of tweets did not promote preventive or help-seeking behaviors | Biased sample Lack of behavioral assessment |
| Sundstrom et al. (33) | United States | Vaccination | Campaign aimed at parents, to raise awareness about the human papillomavirus (HPV) vaccine | Twitter statistics (metrics and impressions) | More than 370,000 total impressions were reached, with pro- and anti-vaccine comments using personal experiences. Comments with misinformation were responded to and corrected by the users themselves | Not generalizable |
| Lenoir et al. (34) | United Kingdom | Cancer | Campaign #SmearForSmear to encourage women to take a selfie showing their lipstick going over the edge and post it, to raise awareness of cervical cancer | Twitter statistics, coding of tweets by topic and analysis of the content of the messages | More than half of the users posted the required photo, and almost a third of the tweets were awareness-raising. The awareness messages were linked to the factors “female gender”, “women who experienced an abnormal smear test” and “UK inhabitants” | Data biases Lack of behavioral assessment |
| Lee et al. (35) | Korea | Cancer | Korean Society of Coloproctology colon cancer campaign | Twitter statistics for the keywords “colorectal cancer,” “colorectal cancer awareness campaign,” “gold ribbon,” and/or “love handle" | The majority of the content of the tweets analyzed was spam, with only 12.6% of the messages sharing information. The impact of the campaign among Twitter users was questionable | Small sample size Data biases |
| Booth et al. (36) | Canada | Mental Health | Bell Let's Talk campaign on mental health awareness and utilization of available preventive services | Record of monthly mental health visits in Ontario outpatient clinics | Twitter inclusion in the campaign was associated with increased utilization of mental health and psychiatric services. Especially significant was the increase in adolescents aged 10–17 years | Data limitations |
| Wittmeier et al. (37) | Canada | Hirschsprung's Disease | Shit Happens campaign to engage family members affected by the disease | Twitter statistics (metrics and reach) | Assessment of responsiveness showed that within 2 h of posting, a question could receive 143 views and 20 responses, increasing to 30 responses after 5 h | Biased sample Not representative |
| Harding et al. (38) | Ghana | Breastfeeding | Campaign to promote safe breastfeeding | Twitter statistics (metrics and impressions) | At the start of the campaign, the materials received an average exposure of 60 users. Reach on Twitter was not significant, while it was on Facebook. | Data limitations Small sample size |
| Gough et al. (39) | United Kingdom | Cancer | Dissemination of messages on the effects of sunlight and the prevention of skin cancer | Pre- and post-intervention household survey; Twitter statistics (metrics and reach), and coding of tweets by topic | There were a total of 417,678 tweet impressions. Shocking messages generated the most impressions, while humorous messages generated the most engagement. The survey revealed an increase in skin cancer awareness, and a change in attitudes about UV rays and tanning | Not representative Survey limitations |
| Ayers et al. (40) | United States | Tobacco | Great American Smokeout campaign to encourage smoking cessation | Twitter statistics (metrics and impressions) using a quasi-experimental design | There was a 28% increase in tweets related to the topic compared to the rest of the year | Metrics |
| Jawad et al. (41) | United States | Tobacco | ShishAware campaign warning of the dangers of pipe smoking | Twitter statistics (metrics and impressions) | Twitter enabled the most organization-based contact, but Facebook was the most interactive medium. There is no data on the effects on awareness, knowledge and attitude of users | Data limitations Lack of behavioral assessment |
| Friedman et al. (42) | United States | STDs | GYT: Get Yourself Tested campaign to reduce stigma and promote communication and testing for sexually transmitted diseases (STDs) | Twitter statistics and affiliate data for Planned Parenthood and infertility prevention clinics | It is estimated that the campaign reached over 52,000 youth. Subsequent years saw a 71% increase in STD testing, although cases of positivity remained stable | Data limitations |
| Fung et al. (43) | China | Hand washing | Global Handwashing Day Campaign | Qualitative content analysis of messages | Social networks serve as amplifiers of content provided by traditional media | Data limitations |
| Chung (44) | United States | Tobacco | Tips From Former Smokers is a smoking cessation campaign from the Centers for Disease Control and Prevention (CDC) | Twitter statistics (metrics and impressions) | The role of non-profit entities in disseminating the message launched by government authorities is noted. Two-way interactions with users were minimal | Data limitations Lack of behavioral assessment |