Twitter as a Sentinel Tool to Monitor Public Opinion on Vaccination: An Opinion Mining Analysis from September 2016 to August 2017 in Italy

"Twitter may be useful to capture real-time changes in public perception about vaccination, potentially providing a fast, low-cost, and easy alternative to traditional polls and surveys."

In Italy, the phenomenon of vaccine hesitancy has led to a drop in vaccination coverage since 2013. Studies on traditional (e.g., newspapers) and social media (e.g., YouTube, Twitter) have found that in the last decade rumours, myths, and disinformation regarding vaccines have been widely broadcasted, resulting in a negative impact on public opinion and people's willingness to be vaccinated. This study presents the results of an opinion mining analysis on vaccination performed on Twitter from September 2016 to August 2017 in Italy.

One purpose of the study is to show how Twitter can be used as a monitoring system to gauge public opinion during vaccination policy changes. In particular, the measles epidemic in Italy in 2017, which led to about 4,885 cases and 4 deaths, stirred a heated political debate. Two events in 2017 are noteworthy: (i) the publication of the National Immunization Prevention Plan (Piano Nazionale Prevenzione Vaccinale, PNPV) 2017-19 on January 19 2017; and (ii) the Legislative Decree n. 73 (June 7 2017) introducing additional compulsory vaccines for school-aged children (Vaccines Decree). Both events were accompanied by strong public debate, also in the social media.

The study is the result of a multi-sectorial approach, applying text mining and machine-learning techniques to tweets' opinion mining. A dataset of 180,620 tweets obtained from the Italian Twitter stream from September 2016 to August 2017 was identified and collected using keywords and hashtags related to vaccination, vaccine-preventable diseases (VPDs), and possible or alleged vaccine side effects. The extracted tweets were then pre-processed in preparation for the automatic classification by means of machine-learning techniques, which are detailed in the article.

During the study period, the number of tweets showed an increasing trend, peaking in the month of July 2017. According to the overall analysis by category, 60% of tweets were classified as neutral, 23% against vaccination, and 17% in favour of vaccination. Overall, during the entire study period, the proportion of subjective tweets (e.g., non-neutral) showed a steady increase, indicating a progressive polarisation of the opinions on vaccination. Tweets expressing opinions against vaccination became predominant over those in favour in the period April-August 2017, with a peak in July 2017.

Vaccine-related events appeared to influence the number and the opinion polarity of tweets. For example, preliminary approval of the Vaccines Decree on June 7 2017 produced a prominent effect in the social discussion in terms of number of tweets (+98.3% vs. baseline, p = .014), with an increase of subjective tweets about vaccination (baseline: 0.41, peak: 0.48, aftermath: 0.46, p < .001 overall), but no effect on negativity rate. The ratification of the Vaccines Decree by the Italian Chamber of Deputies on July 28 2017 resulted in the highest spike in the number of tweets (max tweet count 3,662 on July 28, +130% vs. baseline, p = .03), with moderate effects on neutrality and negative rates.

A qualitative analysis of Word cloud representations of the training datasets (see above) highlighted a higher occurrence of hashtags (in particular #novaccines) and of the word autism in the tweets against vaccination than in the ones in favour of vaccination. In the tweets in favour of vaccinations, instead, the researchers found a higher occurrence of insults to anti-vaccination activists and references to the political world. In both the datasets, the main VPD discussed was measles.

Reflecting on the findings, the researchers note that, although the share of tweets against vaccination showed an increasing trend during the study period, "the increase of public debate on vaccinations and the diffusion of data on the ongoing measles epidemic have already had a positive effect on vaccine perception. The introduction of mandatory vaccinations, despite being generally not well accepted by public, further consolidated this trend leading to an increase in polio and measles vaccines uptake..."

The researchers suggest that this analysis could be relevant in the context of a progressive politicisation of the topic of vaccination, as was seen during 2016 American presidential election. This approach may be particularly beneficial, they say, when implemented in correspondence with key events, such as the adoption of a new health policy (e.g., the Vaccine Decree), as a sentinel system to rapidly gather signals from the public. They conclude that opinion mining analysis based on Twitter is a potentially useful and timely method to assess public opinion toward vaccination and, in the future, may contribute to the development of appropriate communication and information strategies in response to the issue of vaccine hesitancy.