1

The International Tobacco Control PolicyEvaluation Project (ITC Project):

Evaluating the Impact of the WHO FrameworkConvention on Tobacco Control

Geoffrey T. FongDepartment of Psychology, University of Waterloo

Ontario Institute for Cancer Research

“Making the World a Smaller Place: InternationalImplementation of Large-Scale Prevention Practices,Policies, and Programs.” Opening Plenary SessionSociety for Prevention Research, Washington, DC

June 1, 2011

2

“Tobacco is the most effectiveagent of death ever developedand deployed on a worldwidescale.”

– John Seffrin, Past President,American Cancer Society and the

International Union Against Cancer

3

Some Statistics on Global Tobacco Use

1.1–1.3 billion people smoke– 82% live in low-and middle-income countries– 500 million will die of tobacco-related causes

20th Century: 100 million tobacco-related deaths

21st Century: 1 billion tobacco-related deaths

5 million will die this year

By 2025: 8 million will die/year

WHO: Tobacco use is the leading preventablecause of death and disability in the world

44

Facing the Challenge

0

70

220

520

Estimated cumulative tobacco deaths1950-2050

300

400

500

2025 205020001950

100

200

Year

Toba

cco

deat

hs (m

illio

ns)

World Bank. Curbing the epidemic: Governments and the economics oftobacco control. World Bank Publications, 1999. p80.

How can we mosteffectively flatten

this curve?

Population-levelinterventions such

as policies

5

What can we do?

Primary prevention at the population level

National-Level Policies

Population Impact = Reach x Effectiveness

What has the greatest potential impact in a country?

6

Framework Convention on Tobacco Control (FCTC)

First-ever health treaty

Unanimously adopted in 2003

173 nations have becomeParties to the FCTC

(The U.S. is one of the fewnations that has NOT ratifiedthe FCTC)

Conferences of the Parties:COP-1: Feb 2006 in GenevaCOP-2: July 2007 in BangkokCOP-3: Nov 2008 in S. AfricaCOP-4: Nov 2010 in Uruguay

7

Policies of the FCTC

Demand-side strategies More prominent warning labels

Elimination of “light/mild” and other deceptions

Bans/restrictions on marketing (advertising, sponsorship)

Protection from exposure to tobacco smoke

Higher taxes

Support for cessation

Education, communication, public awareness

Supply-side strategies Reduce illicit trade

Reduce youth access

88

Potential of Policies to Flatten the Curve

190

340

500

0

70

220

520

Estimated cumulative tobacco deaths1950-2050

300

400

500

2025 205020001950

100

200

Year

Toba

cco

deat

hs (m

illio

ns)

World Bank. Curbing the epidemic: Governments and the economics oftobacco control. World Bank Publications, 1999. p80.

Impact of policiesdepends on twomain factors:

1. Intervention date

2. Effect size

9

The FCTC: World’s First Health Treaty

Are these policies working?

– Do pictorial warnings work better than text warnings?

– Are smoke-free laws being obeyed? What factorsexplain why these laws may have stronger impact inone country than in others?

– Does policy impact vary across different people (dohigher taxes have stronger impact on youth?)

WHY and HOW do policies have their impact?

To answer these questions, we need to evaluateFCTC policies across different countries

10

International: control/comparison countries

Longitudinal cohort: track individuals over time

Include outcome measures for multiple policies (not justthe policy that is changing tomorrow)

Include measures of key intermediate outcomes of apolicy (what a policy should be changing within a personBEFORE the person changes his/her behaviour): “leadingindicators” of policy impact

Include psychosocial mediators and moderators that willhelp us understand how and why the policy had its impact(if indeed it had an impact)

Increasing internal validity for evaluation ofnational-level policies

11

Policy

ModeratorsCountry

Sociodemographics(e.g., age, sex, SES,ethnic background)

Past Behavior(e.g., smoking history,CPD, quit attempts)

Personality(e.g., time perspective)Psychological State

(e.g., stress)Potential Exposure to

Policy(e.g.,employment status)

Policy-SpecificVariables

• Label salience• Perceived cost• Ad/promo awareness• Awareness of alternative products• Proximal behaviors (forgoing a cigarette because of labels)

PsychosocialMediators

• Outcome expectancies • Beliefs & Attitudes • Perceived Risk • Perceived Severity • Self-Efficacy/ Perc. Beh Control • Normalization beliefs • Quit intentions

Policy-RelevantOutcomes

• Quit Attempts• Successful Quitting• Consumption changes

• Brand switching• Tax/price avoidance• Attitude/belief changes (e.g., justifications)

EconomicImpact

PublicHealthImpact

Conceptual Model of the ITC Project

SES as possible moderatorsof tobacco use and of policyimpact: which policies canhelp close the equity gap?

12

Mediational Model(s) of Policy Effects

Proximal Variables(Policy-Specific)

Distal Variables(Psychosocial Mediators)Policy Behavior

Different policies operate differently, but can bedescribed by the same general model

LabelsLabel SaliencePerc EffectivenessDepth of ProcessingEmotion/arousal

Intentionsto Quit

QuitAttempt

Perceived riskPerceived severity

Ad Ban Advertising saliencePositive assns

Intentionsto Quit

QuitAttempt

Denorm beliefsSocial acceptSubjective norms

13

The International Tobacco Control PolicyEvaluation Project (the ITC Project)

Canada United States Australia United Kingdom

Ireland Thailand Malaysia South Korea

China

France

New ZealandMexicoUruguay

NetherlandsGermany Bangladesh

IndiaBhutanBrazil Mauritius

The ITC Survey Research Team: Principal Investigators, Co-Investigators

Canada: Geoffrey Fong, Mary Thompson, David Hammond, Changbao Wu, Mark Zanna,Christian Boudreau, Steve Brown, Sharon Campbell, Paul McDonald, Emmanuel Guindon

United States: Michael Cummings, Andrew Hyland, Richard O’Connor, Gary Giovino, Frank Chaloupka,Jim Thrasher, Ernesto Sébrie, Mohammad Siahpush, Hana Ross, Maansi Travers, Fritz Laux

Australia: Ron Borland, Hua Yong, Lin Li, Melanie Wakefield, David YoungUnited Kingdom: Ann McNeill, Gerard Hastings, Abraham Brown, Louise Hassan, Fiona HarrisIreland: Shane Allwright, Luke Clancy, Fenton Howell, Maurice MulcahyThailand: Buppha Sirirassamee, Warangkana Polprasert, Philip Guest, Steve Hamann,

Prakit Vateesatogkit, Aree Jampaklay, Aree Prohmmo, Chanya SethaputMalaysia: Maizurah Omar, Rahmat Awang, Foong Kin, Ahmad Shalihin Mohd SaminNew Zealand: Nick Wilson, George Thomson, Richard Edwards, Judy LiSouth Korea: Hong-Gwan Seo, Yoo-Seock Cheong, Seung-Kwon Myung, Yeol KimChina: Jiang Yuan, Li Qiang, Yang Yan, Feng Guoze, Xiao LinFrance: Romain Guignard, Pierre Arwidson, François Beck, Jean-Louis Wilquin, Sylviane RatteGermany: Martina Pötschke-Langer, Ute MonsNetherlands: Marc Willemsen, Gera Nagelhout, Bas van den Putte, Hein de VriesMexico: Jim Thrasher, Edna Arillo Santillán, Eduardo Lazcano Ponce, Rosaura Pérez Hernández,

Luz Myriam Reynales, Victor VillabosUruguay: Marcelo Boardo, Eduardo BiancoBrazil: Cristina Perez, Valeska Figuerido, Tania Cavalcante, André Szklo, Paula Johns, Eliane VolchanBangladesh: Nigar Nargis, SM Ashiquzzaman, Ummul Ruthbah, Hussain Ghulum, Iftekharul Huq,

Abu S.M. AbdullahBhutan: Sonam Phuntsho, Ugyen NorbuMauritius: Premduth Burhoo, Bimla Moussa, Deowan Mohee, Véronique Le ClézioIndia: Prakash Gupta, Mangesh Pednekar

15

Content of the ITC Surveys

Unique ITCContent:

170-200 Qsfocusing on

policy impact

Surveillancecontent

Surveillancecontent

Mixed Surveillanceand policy content

Throughoutthe policysectionsthere aremeasuresrelevant tomonitoring

16

17

ITC Surveys are being conducted in: Over 50% of the world’s population Over 60% of the world’s smokers Over 70% of the world’s tobacco users

1818

Warning labels– UK (2003): Text – India (2009): Graphic– UK (2009): Graphic– Thailand (2006): Graphic– Australia (2006): Graphic– Canada (2010): Graphic, Round 2– China (2008): Text– Mexico (2008): Graphic– Uruguay (2006,09): Graphic– Brazil (2008/09): Graphic, Round 3– Malaysia (2008): Graphic

– Mauritius (2009): Graphic

ITC Evaluation of FCTC Policies (Partial List)

Product policies– UK (EU): 10-1-10 regulation– US/Canada: Reduced ignition propensity– All countries: product; product x behavior

Light/mild bans – UK (2003) – Australia (2005) – Canada (2006) – Brazil + others (2009+)

Smoke-free– Ireland (2004) – Mexico (2008+)– Scotland (2005) – Brazil (2008+)– England (2007) – Bangladesh (2009+)– Uruguay (2006) – Bhutan (2009+)– France (2007/08) – India (partial)– Germany (2008+) – Mauritius (2009) – China (2008/09 + Olympics) – Netherlands (Part 2–2008/09)

Advertising/Promotion– UK (2003): Comprehensive

– Canada (2003): Last part of Comp.– Thailand (2006): POS bans– Mexico (2008): Comprehensive

– Canada (2008+): Re-emergence of“descriptive” ads & possible new ban

– China (2011): Comprehensive– Many other countries: Partial

Taxation– All countries

Illicit trade – China (2008): prevalence

– Additional in Canada/U.S.: close to reserves – Bhutan (2009+): total ban on sales in country

19

“Tobacco use is unlike other threats to global health.Infectious diseases do not employ multinationalpublic relations firms. There are no front groups topromote the spread of cholera. Mosquitoes haveno lobbyists.”

– WHO Zeltner Report (2000)

2020

ITC Project Scientific Papers

2121

ITC China Project Supplement in Tobacco Control

2222

ITC Project Cessation Supplement inNicotine & Tobacco Research

1. Released October 1, 2010

2. 11 articles

3. Examples:

– Gibson et al.: Evaluation of the UK National Health Service (NHS) Stop Smoking Services

– Reid et al.: SES disparities in quit intentions, quit attempts, and smoking abstinence among smokers in US, UK, Canada, Australia

– Wilson et al.: Increased recognition of NZ quitline following introduction on pictorial labels, equalizing pre-intro SES differences in noticing

2323

Creating Guidelines and Standards for Research

IARC Cancer PreventionHandbook:

Methods for EvaluatingTobacco Control Policies

(February 2009)

• Best practices in the evaluation of tobacco control policies

• Preparation of Handbook led by ITC Project investigators

• Conceptual Model of the ITC Project used to frame the evaluation methods of the Handbook

2424

ITC National Reports

2525

ITC National Reports: France

2626

ITC National Reports: France (Feb 2009)

Preface from theMinister of Health

and Sports,Roselyne Bachelot-

Narquin

2727

ITC National Reports: Bangladesh (April 2010)

Preface from the Prime Minister,Sheikh Hasina

2828

ITC National Reports: Bhutan (May 2011)

Bhutan’s Minister of Health, Zangley Dukpa,at the release of the ITC Bhutan Report

2929

ITC Policy Reports

30

ITC Labels Report–WNTD 2009

The ITC Project released a report on pictorial warnings forWorld No Tobacco Day (May 31, 2009): 12 page report on ITC Project

findings in the domain of warning labels (in English and Chinese)

31

ITC Cessation Report–Prepared for the FCTCConference of the Parties, Nov 2010

32

ITC Working Paper on Tax in Bangladesh (2010)

Impact ofvarying taxlevels on

tobacco use

Impact ofvarying tax

levels on taxrevenue

3333

The Use of ITC Data to Evaluate theImpact of FCTC/Tobacco Control Policies

343434

Evaluating Graphic Warnings in Thailand

2005 2006

Thailand

Malaysia

353535

Enlarging the labels and adding graphic images led tosubstantial increases in the percentage of Thai smokers

reporting that warnings make them think about health risks

363636

Enlarging the labels and adding graphic images also led tolarge increases in the percentage of Thai smokers reporting

that warnings make them more likely to quit

37

ITC Four Country Survey—Change in Avoiding Labels

Significance: avoiding labels predicts future quit attempts.

Example of the mediational model for warnings: Pictorialwarnings are linked to future quit attempts because they firstcreate reactions that are associated with future quitting

Borland et al.Tob Control(2009)

UK: TextJan 2003

Aust: pictorialMar 2006

38

Smoking Prevalence in Bars/Pubs Before & After Ban in Ireland (04),Scotland (06), UK (07), France (08), Netherlands (08), Germany (07-08)

Jan04

Oct04

Jan05

Oct05

Jan06

Oct06

Jan07

Mar07

Oct07

Oct08

98 98 98 93 96

Mar08

9892

• • •

• • •

Ireland UK (Non-Scotland) Scotland France Netherlands Germany

•

••

Dec08

UKApr,Jul

07

SCMar06

FRJan08

93

NLJul08

Apr09

IEMar04

DE 07-08

5 4 2 5 4

31

52

3939

4040

Using ITC Findings to Demonstrate:1. Need for strong(er) policies

2. Current policies are inadequate

“Evidence to Power”

4141

Dec 9, 2010: Hearing held by the Houseof Commons Health Committee on theGovt’s decision to shelve the revision.

ITC Canada Survey data show that everyindicator of label impact has declined

dramatically over the past 7 years

Dec 30, 2010: Health Ministerreverses decision—the revision of thewarnings will continue. New warningsare expected by the end of 2011.

Sep 28, 2010: Health Canadaannounces that the ongoing initiativeto revise the 10-year-old tobaccowarnings is being shelved

4242

On May 1, 2009, China increased cigarette taxesby 6-11%, claiming that they were doing so inaccordance with the FCTC.

BUT prices have still not increased (Jun 1, 2011).

Affordability of cigarettes continues to increase.

The ITC China Survey shows clearly that pricecontinues to be no barrier to smoking and is nota factor in motivating quitting.

43

China’s warning labels changed in Oct 2008

OLD warning:Side of pack

Only one message

NEW warnings:Front/back of pack

Two messages

4444

ITC China Survey: Label change had trivial impact

4545

ITC China Survey: Label change had trivial impact

46

47

< 10% of adult smokersunderstood bothEnglish warnings

4848

Smoking Prevalence in Restaurants in the 6 ITC cities in China compared toOther Countries Before and After Smoke-Free Laws: Ireland (04), Scotland (06),France (08), Germany (07-08), Netherlands (08), Mexico City (08), Mauritius (09)

•

84•

IEMar04

2

59

SCMar06

0

71

2

FRJan08

83

NLJul08

5

86

29

DE07‐08

•

•

MXCApr08

9

78

62

16

MUMar09

•

China87‐96%

••••••••• ••

% o

f Res

taur

ants

in w

hich

ther

e w

as A

NY

smok

ing

4949

Bangladesh: Decline in beliefs about health harmsof smoking: 2009–10

• There has been avery significant dropin beliefs aboutharms of smoking

• There is a strongneed to inform thepublic about theharms of smoking

• Warning labels arethe most cost-effective method foreducating the public

5050

Implications for beliefs about harms in Bangladesh

Tuberculosis 231,000

Lung Cancer 1,743,000

Stroke 1,806,000

Heart Disease 1,218,000

Mouth Cancer 3,654,000

Impotence 3,297,000

Number of smokers in Bangladesh (out of 21 million) whofrom 2009 to 2010 no longer believe that smoking causes...

5151

Evidence of poor implementation of Dutch smoking ban and strongpro-tobacco attitudes compared to smokers in other ITC countries

ITC National Reports: The Netherlands (2010)

52 52

The Dutch Health Ministercontinues her campaign toremove funding for tobaccocontrol and to weaken smoke-free laws. So the ITC Projectissues a second report inMarch 2011 showing thatDutch smokers are the leastknowledgeable and leastconcerned of all 19 ITCcountries about SHS hazards.

Geoffrey Fong gives aspeech in Xi’an China inApril 2011 to the leadingnational group of tobaccocontrol experts in China:“Things are bad in China,but at least they aren’t asbad as they are in theNetherlands.”

BMJ article—Apr 9, 2011

5353

ITC Project: system for understanding the natural historyof tobacco use over time and impact of population-levelinterventions (e.g., policies)

Evaluation = Surveillance + Design

Allows strong opportunities for Evidence to Power

ITC as a possible model for other domains of health wherepopulation-level interventions (policies and programs) arebeing considered (e.g., obesity, physical activity).

Summary

54

“Just as surely as the laws of gravityoperate in Mumbai as they do in Lyon,the principles of causality—and themethods employed to make moreconfident judgments about causalrelations—are constrained by neitherlocation nor content domain.”

– Fong and Thompson, IARC Handbook,Methods for Evaluating Tobacco Control Policies (2009)

55

ITC Project Research Support

Core support provided by theU.S. National Cancer Institute

(P01 CA138389)

Additional core fundingprovided by the

Canadian Institutes ofHealth Research

QuickTime™ and aTIFF (Uncompressed) decompressor

are needed to see this picture.

ITC Project Research Organizations

QuickTime™ and a decompressor

are needed to see this picture.