Investigating the Impact of the New York State Flavor Ban on e-Cigarette-Related Discussions on Twitter: Observational Study.

Background: On May 18, 2020, the New York State Department of Health implemented a statewide flavor ban to prohibit the sales of all flavored vapor products, except for tobacco or any other authorized flavor.

Objective: This study aims to investigate the discussion changes in e-cigarette-related tweets over time with the implementation of the New York State flavor ban.

Methods: Through the Twitter streaming application programming interface, 59,883 e-cigarette-related tweets were collected within the New York State from February 6, 2020, to May 17, 2020 (period 1, before the implementation of the flavor ban), May 18, 2020-June 30, 2020 (period 2, between the implementation of the flavor ban and the online sales ban), July 1, 2020-September 15, 2020 (period 3, the short term after the online sales ban), and September 16, 2020-November 30, 2020 (period 4, the long term after the online sales ban). Sentiment analysis and topic modeling were conducted to investigate the changes in public attitudes and discussions in e-cigarette-related tweets. The popularity of different e-cigarette flavor categories was compared before and after the implementation of the New York State flavor ban.

Results: Our results showed that the proportion of e-cigarette-related tweets with negative sentiment significantly decreased (4305/13,246, 32.5% vs 3855/14,455, 26.67%, P<.001), and tweets with positive sentiment significantly increased (5246/13,246, 39.6% vs 7038/14,455, 48.69%, P<.001) in period 4 compared to period 3. "Teens and nicotine products" was the most frequently discussed e-cigarette-related topic in the negative tweets. In contrast, "nicotine products and quitting" was more prevalent in positive tweets. The proportion of tweets mentioning mint and menthol flavors significantly increased right after the flavor ban and decreased to lower levels over time. The proportions of fruit and sweet flavors were most frequently mentioned in period 1, decreased in period 2, and dominated again in period 4.

Conclusions: The proportion of e-cigarette-related tweets with different attitudes and frequently discussed flavor categories changed over time after the implementation of the New York State ban of flavored vaping products. This change indicated a potential impact of the flavor ban on public discussions of flavored e-cigarettes.