Did the introduction of pictorial health warnings increase information seeking for smoking cessation?: Time-series analysis of Google Trends data in six countries

Abstract

INTRODUCTION Pictorial health warnings (PHW) can influence smoking cessation rates and precursors thereof. However, both the magnitude and duration of their impact, in national populations, remain uncertain because of limitations of the available data. In this study we used Google Trends data from six European countries to evaluate whether the implementation of PHW was followed by a short-term increase in online searches on smoking cessation. METHODS We applied an interrupted time-series design using ARIMA models. We used weekly or monthly data on the relative search volume (RSV) for search terms about smoking cessation. First, RSV trends were seasonally adjusted and adjusted for autocorrelation. Next, regression models were fitted that included terms for the potential effect of PHW in month 1, months 2–3, and months 4–6 after implementation. RESULTS Our findings for France and the United Kingdom were partly in line with our initial expectations. In France, a 4% increase (95% CI: -2% – 11%) occurred in the first month after implementation, but not later. In the UK, a 3% increase (95% CI: 1% – 6%) in ‘quit smoking’ searches occurred in months 2–3. No increases were observed for any other periods for France, the UK, Ireland, Norway, Denmark or Switzerland. CONCLUSIONS We found no consistent support that the implementation of PHW was associated with increased internet searches for smoking cessation. Further studies are needed to assess and understand the magnitude and duration of populationwide impacts of PHW.