1. Impact of different types of tobacco control advertising content on six-month campaign recall in England Sol Richardson UK Centre for Tobacco Control Studies (UKCTCS) University of Nottingham

2. 27/06/2013 Background In industrialised countries, tobacco use accounts for around half the inequalities in adult mortality While overall rates of tobacco use in the UK declined from 2001 to 2008, no progress was made in multiply disadvantaged groups

3. 27/06/2013 Background Tobacco control mass media campaigns play a key role in encouraging smoking cessation among adults (Bala et al., 2012) Campaigns featuring emotive or graphic content are more effective than those which do not, especially among socioeconomically disadvantaged groups (Durkin et al., 2011) Although many studies have measures recall in children and adolescents, little is known about the relationship between campaign exposure and recall in adults

4. 27/06/2013 Adult Recall Campaigns with highly emotive content, or those delivered using testimonials, tend to generate greater recall Recall tends to rise in line with campaign exposure, although increased exposure to emotive campaigns generated a greater increase in recall Recall tends to rise with level of education. The differences in levels of recall generated by each campaign type tend to narrow with higher levels of education Dunlop et al., 2012 Niederdeppe et al., 2011

5. 27/06/2013 Background Why study campaign recall? To be effective in modifying behaviour, advertisements must be observed and recalled Campaign recall provides a direct measure of the possible effectiveness of campaigns in reaching smokers, and could be considered a proxy measure of effective campaign exposure Can provide insight into the mechanisms through which tobacco control campaigns impact on rates of smoking prevalence and cessation

6. 27/06/2013 Opportunities The UK provides an ideal setting in which to evaluate the effectiveness of different campaign types due to the variety of televised tobacco control advertising content This contrasts with other settings: Australia: Overwhelming majority of campaigns employ graphic images and hard-hitting messages. USA: Fragmented media market and a range of organisations disseminating tobacco control advertising

7. 27/06/2013 Challenges Small individual-level effects of advertising - Large samples and high levels of exposure are therefore required to detect significant effects Confounding factors, including underlying trends, other concurrent policies and initiatives, individual-level characteristics…

8. 27/06/2013 Hypotheses 1.Increased exposure to tobacco control campaigns results in increased recall, and that negative emotive campaigns achieve a greater size of effect 2.Exposure to different types of televised tobacco control campaigns has differential effects in different socioeconomic groups 3.Campaigns designed to elicit negative emotions achieve higher rates and longer durations of recall

9. 27/06/2013 Methods The International Tobacco Control Four Country Survey (ITC-4) An eight-wave prospective longitudinal cohort study of adult smokers in the United Kingdom, Australia, Canada and the United States The selection procedure used stratified random digit dialling to yield a representative sample of the UK smoking population aged ≥18 years who have smoked more than 100 cigarettes in their lifetime and who have smoked at least once in the past 30 days Participants were rewarded with a £7 pharmacy voucher to encourage retention

10. 27/06/2013 Follow-up Smokers residing in England aged 18 or over 3,641 unique individuals recruited in waves 1 to 6 of the study’s UK arm (from January 2002 to February 2008) 1,987 (or 55%) were followed up at least once in waves 4 to 7 between April 2005 and March 2009 and had valid exposure and recall data, giving 3,972 observations over four iterations of the survey Apr-05 4 0 May-05160.4 Jun-05461.8 Jul-05557.7 Aug-0595.2 Sep-05514.6 Oct-05212.1 Nov-05151 Dec-05278.4 Jan-060 Feb-06459.7 Mar-06641.6 Apr-06 5 0 May-060 Jun-060 Jul-060 Aug-060 Sep-06394.6 Oct-0617 Nov-060 Dec-060 Jan-07707.6 Feb-07126.2 Mar-07 6 454 Apr-07100.5 May-07328.4 Jun-07615.7 Jul-0795.1 Aug-070 Sep-070 Oct-07219.1 Nov-0792.7 Dec-0785.4 Jan-08718.1 Feb-08487.4 Mar-08411.5 Apr-08 7 14.6 May-080 Jun-08604.2 Jul-08469.3 Aug-08147.6 Sep-08174.6 Oct-08247.1 Nov-08644 Dec-08272.5 Jan-09608.5 Feb-09723.5 Mar-09984.3

11. 27/06/2013 Socio-Economic Status Around a third of responses for income were missing Respondents were classified according to level of education: - Low: Compulsory education up to age 16 - Middle: A-level or Level 3 vocational qualifications, any course of higher education below degree level - High: Any degree-level university education or above

12. 27/06/2013 Campaign Exposure: TVRs Television ratings (TVRs) / Gross rating points (GRPs): - Expressed as the percentage of all potential viewers capable of receiving broadcast signals in a given media market exposed to an advertisement during a given time period e.g. 1000 TVRs = each person has viewed an ad 10 times, or 50% have viewed it 20 times On an individual level, actual exposure will vary according to a range of factors, including frequency and time of television viewing

13. 27/06/2013 Campaign Exposure From April 2005 to March 2009 ranged from 0 to 1,051 TVRs, with a mean of 276.6. The total exposure over the period was 13,276 TVRs There was no discernible long-term trend in tobacco control TVRs, although campaign exposure tended to peak in January TVRs: 2004-2012

14. 27/06/2013 Campaign Exposure Intensity

15. 27/06/2013 Characterisation of campaigns Advertising content was categorised according to: -Theme -Informational content -Emotional content -Style of delivery Coded by Tessa Langley (University of Nottingham) and Michelle Sims (University of Bath)

16. 27/06/2013 Characterisation of campaigns For the purposes of our study, campaigns were categorised as having either: “Positive” emotional content: eliciting happiness, satisfaction or hope (48.3%) “Negative” emotional content: eliciting fear, sadness, guilt, anger or disgust (46.5%) … or no emotional content at all (5.2%)

17. 27/06/2013 Outcome Variable Question: “In the last 6 months, have you noticed advertising or information that talks about the dangers of smoking, or encourages quitting on television?” Operationalised as a binary variable

18. 27/06/2013 Analysis We generated mutually adjusted variables for exposure to each campaign type in the 1-3, 4-6 and 1-6 months before survey Generalised estimating equations (GEE) for binary outcomes We tested the associations between exposure to different campaign types campaign recall, stratified by level of education

19. 27/06/2013 Results All Participants (n = 3972) Campaign Category TVRs*PeriodOR (95% CI)p All Campaigns1-6 months ago1.25 (1.10 - 1.41)< 0.001 Elicits Negative Emotions1-6 months ago1.42 (1.24 - 1.62)< 0.001 Elicits Positive Emotions1-6 months ago0.89 (0.72 - 1.09)0.256

20. 27/06/2013 Results Low Education (n = 1240)Mid Education (n = 1942)High Education (n = 750) ModelCampaign Category TVRs*PeriodOR (95% CI)p p p 1All Campaigns1-6 months ago1.10 (0.90 - 1.34)0.3681.39 (1.15 - 1.68)0.0011.28 (0.97 - 1.68)0.079 2 Elicits Negative Emotions1-6 months ago1.25 (1.01 - 1.56)0.0421.60 (1.30 - 1.95)< 0.0011.43 (1.03 - 1.99)0.035 Elicits Positive Emotions1-6 months ago0.74 (0.53- 1.03)0.0700.98 (0.70 - 1.36)0.8870.97 (0.60 - 1.55)0.886 *in units of 1,000 TVRs, adjusted for gender, age group, cohort of recruitment

21. 27/06/2013 Results: Duration of Recall ModelCampaign Category TVRs*Period OR (95% CI) (n = 3,972) p 1 All Campaigns1-3 months ago1.38 (1.04 - 1.83)0.025 All Campaigns4-6 months ago1.15 (0.91 - 1.45)0.239 2 Elicits Negative Emotions1-3 months ago1.48 (1.07 - 2.04)0.019 Elicits Negative Emotions4-6 months ago1.44 (1.12 - 1.85)0.004 Elicits Positive Emotions1-3 months ago0.75 (0.43 - 1.32)0.319 Elicits Positive Emotions4-6 months ago1.18 (0.48 - 2.94)0.717

22. 27/06/2013 Limitations The outcome measure, recall, was based on self-reporting False reporting may have occurred due to recall of campaign exposure through other media such as billboards, radio, newspapers and the internet We were unable to control for the timing of the launch phase of individual campaigns We could not evaluate unprompted recall of specific content

23. 27/06/2013 Key Findings Increased exposure to campaigns with negative emotive content in the six months prior to each survey was associated with higher recall, with the greatest effect in the higher SES groups Increased exposure to “positive” campaigns was ineffective at generating recall in any SES group “Negative” campaigns were also found to achieve longer durations of recall (over 3 months)

24. 27/06/2013 Conclusions Campaigns designed to elicit negative emotions achieved higher rates and longer durations of recall Previous UK tobacco control campaigns may have been ineffective in reducing inequalities due to lower rates of recall in low socioeconomic status groups There may have been missed opportunities to reduce smoking prevalence over the last decade due to high expenditures on campaigns carrying “positive” messages

25. 27/06/2013 Acknowledgements Research team Nottingham: Sarah Lewis, Tessa Langley, Lisa Szatkowski Bath: Michelle Sims, Anna Gilmore, Ruth Salway KCL: Ann McNeill

26. 27/06/2013 Acknowledgements Funding This project was funded by the National Prevention Research Initiative. For more information about the NRPI, visit http://www.npri.org.uk.http://www.npri.org.uk The NPRI Funding Partners relevant to this award are (in alphabetical order): Alzheimer's Research Trust; Alzheimer's Society; Biotechnology and Biological Sciences Research Council; British Heart Foundation; Cancer Research UK; Chief Scientist Office, Scottish Government Health Directorate; Department of Health; Diabetes UK; Economic and Social Research Council; Health and Social Care Research and Development Division of the Public Health Agency (HSC R&D Division); Medical Research Council; The Stroke Association; Wellcome Trust; and Welsh Assembly Government.

27. 27/06/2013 Thank you for your attention Any Questions?

28. 27/06/2013 References Jha P, Peto R, Zatonski W, Boreham J, Jarvis MJ, Lopez AD. Social inequalities in male mortality, and in male mortality from smoking: indirect estimation from national death rates in England and Wales, Poland, and North America. Lancet. 2006;368(9533):367-70. Hiscock R, Bauld L, Amos A, Platt S. Smoking and socioeconomic status in England: the rise of the never smoker and the disadvantaged smoker. J Public Health (Oxf). 2012;34(3):390-6. Bala M, Strzeszynski L, Cahill K. Mass media interventions for smoking cessation in adults. Cochrane Database Syst Rev. 2008(1):CD004704. Durkin SJ, Wakefield MA, Spittal MJ. Which types of televised anti-tobacco campaigns prompt more quitline calls from disadvantaged groups? Health Educ Res. 2011;26(6):998- 1009.

29. 27/06/2013 References (Continued) Dunlop SM, Perez D, Cotter T. The natural history of antismoking advertising recall: the influence of broadcasting parameters, emotional intensity and executional features. Tob Control. 2012 ;0:1-8. Niederdeppe J, Farrelly MC, Nonnemaker J, Davis KC, Wagner L. Socioeconomic variation in recall and perceived effectiveness of campaign advertisements to promote smoking cessation. Soc Sci Med. 2011;72(5):773-80. For more information on the ITC-4 Survey methodology: Thompson ME, Fong GT, Hammond D, Boudreau C, Driezen P, Hyland A, et al. Methods of the International Tobacco Control (ITC) Four Country Survey. Tob Control. 2006;15 Suppl 3:iii12-8