The effect of plain cigarette packaging on attention to hea»warnings

Abstract

BackgroundWhile the Australian Government pushes ahead with the introduction of standardised packaging for tobacco products, the UK Government completed a consultation on this issue in August, 2012. There are several potential benefits, including scope for increasing the effectiveness of health warnings. A recent review suggests that standardised packaging increases the salience of health warnings. Notably, however, the review identified only one study using the most direct measure available to investigate attention: eye movements. The sharpest vision is produced by the central retina, so we use rapid (saccadic) eye movements to fixate on different parts of an image and accrue visual information. By use of eye tracking technology commonly used in psychological research to record eye movements, fixation times on different areas of an image can be calculated as a proxy for attention. Branding is of course designed to be eye-catching, taking advantage of our eyes' tendency to seek areas of high image contrast. Consistent with this fact, removal of branding in the aforementioned study increased the number of saccades towards health warnings, at least for non-smokers or non-daily smokers. However, this result was obtained with graphic warning labels whereas, in the UK, front-of-pack warnings are text only. We aimed to examine the effect of plain packaging on text health warnings by presenting participants with a mocked-up display of branded and plain packaged cigarette packs and asking them to select a pack for purchase. Methods28 adults were recruited by convenience sampling (six men, age 18–40 years, mean age 23·29 years), and consisted of daily (n=9), weekly (n=7), non (n=8), and ex (n=4) smokers. Participants viewed displays containing six packets of cigarettes, with each of the two current UK text health warnings present on three packets. Displays always contained the same six brands, but brand position and assignment of the health warnings was randomised in every trial. Each participant received six consecutive presentations, which alternated between all branded and all plain displays (with order counterbalanced across participants). A mouse pointer appeared 5 s after display onset to allow brand selection, which also ended the trial. A video eye tracker recorded eye movement data. The key measure was the proportion of time participants spent fixating health warnings compared with the rest of the packet, which was either plain or contained the branding. FindingsAs expected, most time was spent looking at branding. Crucially, however, fixation times on warning labels increased in the unbranded compared with the branded presentations (18·1%vs 15·8%). This difference was not uniform across the experiment, but rather was significant only in the first pair of trials (paired-sample t test, p=0·043). InterpretationThe finding suggests that in contexts without recent habituation to warning labels (eg, first-time smokers) plain packaging helps to redirect the eyes towards text-only health warnings. However, the effect appears transient relative to that obtained with graphic warnings. Hence we suggest that a change to plain packaging might be most profitable if accompanied by a shift to front-of-packet graphic warning labels. FundingNone.