Social Listening Applied to Tailor Communication on Immunization in the Republic of Moldova

"Analysing conversations and engagement rates through social listening can help design messages and build communication strategies, as well as develop social marketing strategies to influence behaviour and make vaccination attractive again."

In the Republic of Moldova, anti-vaccine messages have begun to spread, eroding the population's confidence. Vaccination coverage against measles, mumps, and rubella at 12 months decreased from 95% in 2008 to 83% in 2021. This study, funded by the United Nations Children's Fund (UNICEF) in Moldova, involved conducting social listening and media monitoring, particularly focusing on misinformation related to routine vaccination. Research aims included providing a qualitative analysis and assessment of misinformation, identifying potential threats to immunisation activities, and offering recommendations for designing communication messages.

The research team conducted a mixed-methods study by using Talkwalker, a platform for social media monitoring and listening,  and generating qualitative analysis of the identified immunisation narratives through thematic analysis. Talkwalker's Sentiment Analysis covers 186 languages as of this writing and can provide an overall accuracy of up to 90%. These sentiment models can also interpret human emotions and detect basic types of irony and sarcasm.

Of 450 identified comments, 30 were selected for further analysis. Over 5 months (August 1 2023 and January 1 2024), 865 results were obtained, with an engagement rate of 6,300. The peak in results occurred between December 11-18 2023, driven by a measles outbreak in Romania that saw several confirmed cases and one death, as well as the first cases of measles in Moldova after 3 years. The highest level of engagement was recorded between September 27 and October 1 2023, and it included material related to immunisation in general.

The net sentiment for the analysed period was identified by the Talkwalker software as negative (-25.3%), a finding also confirmed by the qualitative analysis of the messages. The emotional palette is dominated by anger (40.8%) and fear (30.1%). Comments on immunisation tended to reflect distrust in authorities and doctors, political and ideological views, and a lack of knowledge about the risks of vaccine-preventable diseases. For example, conversations revealed suspicions about the truthfulness of posts made by health authorities, accusations against authorities and doctors of inducing fear for the purposes of control and manipulation, sarcasm, and anti-vaccination propaganda. All these elements are signs of vaccine hesitancy.

The research revealed that the topic of vaccination is used to promote political ideologies through messages issued by the mass media. For example, anti-vaxxers may express support for pro-Russian politics and portray vaccines as an influencing tool used by Western countries. Disinformation appears to be present in conversations, attempting to discredit reliable studies, although the intent of the poster is unknown.

The analysed conversations are assessed as having a low risk of spreading (virality), though some employ techniques to influence or manipulate behaviours. When media outlets attempt to gain clicks by using shocking titles, the materials tend to become memorable, as they follow simple but effective techniques.

The authors note that there are messages (though still limited) promoting vaccination, but they receive little support from internet users. Among the pro-vaccination messages, some comments contain reactive and "unfriendly" content directed at the hesitant public, which only increases resistance.

A good example of inducing positive associations with the subject of immunisation is reflected in the story of Galina Spataru, who wrote in a post: "To stay healthy, I got vaccinated according to the doctor's recommendations. Vaccination is very important for everyone...". The story has resonated with netizens, particularly those from the community, who did not hesitate to express their appreciation for the former teacher. The text is not too long (about 220 words) and is notable for the way the message was constructed: The introduction includes a short presentation using words that evoke positivity - energy, good mood, and kindness - and is interspersed with quotes from the activist.

Going forward, the researchers suggest:

• When tailoring messages on immunisation, it is worth considering the sentiment generated by certain words and concepts. For instance, positively received terms include 'medication', 'appointment', and 'health', while 'flu', 'vaccinated', 'measles', and 'doctors' generate negative sentiment.
• A compelling story about a single person (e.g., a protagonist promoting vaccination) is far more likely to move an audience to action than the use of data alone. For a message to be effective, both the messenger and the way the message is presented are important. The use of rational arguments and data alone will not be sufficient

In conclusion: "The information vacuum must be filled with messages designed to produce attitude change regarding vaccination."