2. International standards
Pictures
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
2
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
3. Health Warning Labels
Implementation of pictorial warnings
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
3
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
Source: Cigarette health warnings: International status report. Canadian Cancer Society. October 2010.
4. Health Warning Labels
Countries with pictorial warnings
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
4
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
5. Health Warning Labels
Countries with pictorial warnings
Second level
Latvia 2010 Mexico 2010 Mongolia 2010 Pakistan 2010 Switzerland 2010
Third level
Fourth level
Turkey France Malta 2011 Spain 2011* Norway 2011*
Fifth level
2010 2011
Philippines 2011* Urkraine 2012*
Warning Labels stopped you from having a
cigarette.
40
Canada
5
Australia
*Proposed implementation date at time of publication.
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
6. Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
6
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
7. Size
82% Australia*
80% Uruguay
Second level
75% Canada
Third level
65% Madagascar
Fourth level
65% Mauritius
Fifth level
65% Mexico
60% Philippines
60% New Zealand
60% Cook Islands
56% Belgium 40
Warning Labels stopped you from having a
cigarette.
Canada
7
Australia
US
30 UK
Percentage
20
10
Source: Cigarette health warnings: International status report. Canadian Cancer Society. October 2011.
0
Wave 1 Wave 2
8. Position
Front vs. Back
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
8
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
9. Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
9
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
10. Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
10
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
11. Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
11
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
12. Message “rotation”
Second level
2002
Third level
Fourth level
Fifth level
2004
2008 Warning Labels stopped you from having a
cigarette.
40
Canada
12
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
13. Number of messages in rotation
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
13
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
25. Text
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
25
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
26. Health Warning Content
Quit lines
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
26
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
27. Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
27
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
28. Content
Cessation Inserts
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
28
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
29. Health Warning Content
Pack inserts
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
29
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
30. Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
30
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
31. Legal challenges
United States
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
31
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
32. Preliminary injunction
Factual?
“…the evidence here overwhelmingly suggests that the
Second level requirements are not….purely factual
Rule's graphic-image
and uncontroversial disclosures…. Indeed, the fact alone
Third level
that some of the graphic images here appear to be
Fourth level
cartoons, and others appear to be digitally enhanced or
Fifth level
manipulated, would seem to contravene the very definition
of "purely factual.” page 14
Warning Labels stopped you from having a
cigarette.
40
Canada
32
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
Source: Richard J. Leon. Memorandum of Opinion. Nov 7 2011. Civil Case No. 11-1482 (RJL) United States District Court for the District of Columbia.
33. Factual
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
33
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
34. Credibility
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
34
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
35. Content
Credibility
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
35
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
36. Implementation & Legal challenges
India
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
36
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
37. Implementation challenges
“Little” tobacco
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
37
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
38. Implementation challenges
Contraband tobacco
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
38
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
39. Second level
Third level
Fourth level
Fifth level
“Other” tobacco products 40
30
Warning Labels stopped you from having a
39
cigarette.
Canada
Australia
US
UK
Percentage
20
10
0
Wave 1 Wave 2
40. “Other” tobacco products
Waterpipe
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
40
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
41. Links with other media
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
41
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
42. Summary
• Warning labels should be regarded as a
communication campaign.
• Policy evolution: “First” vs. “second”
generation.
• Importance of international collaboration.
• Role of research & evidence.
43. Second level
Third level
Fourth level
Fifth level
www.tobaccolabels.ca
Warning Labels stopped you from having a
cigarette.
40
Canada
43
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
44. Contact
David Hammond PhD
School of Public Health & Health Systems
University of Waterloo
Tel. 519 888 4567 ext.36462
Email dhammond@uwaterloo.ca
Web www.davidhammond.ca
45. Evidence from low and middle income
countries: Latin America & the Caribbean
Ernesto M Sebrié, MD MPH
Department of Health Behavior, Roswell Park Cancer Institute
Buffalo, New York
15 th World Conference on Tobacco OR Health
Singapore, March 21, 2012
46. Pictorial Health Warning Labels in
MEXICO (2010)
LAC
HONDURAS (2011)
VENEZUELA (2005)
PANAMA (2008)
COLOMBIA (2010)
PERU (2010) BRAZIL (2002)
BOLIVIA (2012)
CHILE (2006)
URUGUAY (2006)
47. Cigarette Labeling Studies in Latin
America & the Caribbean:
1. Caribbean Study:
• Jamaica, Barbados, Trinidad & Tobago,
Guyana
2. Central America Study:
• Nicaragua, Honduras (Guatemala,
Panama, El Salvador)
3. South America Study:
• Bolivia
48. GOALS
1. To determine the most effective content of pictorial
HWLs related to 3 themes:
• Health effects
• Environmental effects/ social & economic impact
• SHS & Smokefree support
2. To determine if greater credibility is attributed to the
message from the Minister of Health compared to
other agencies
3. To determine which “marker word” is most effective
in attracting attention and increasing credibility of
the message presented
49. DATA COLLECTION
Mall intercept recruitment technique
• Subjects were recruited from 2 different local public
venues (e.g., mall, market, or equivalent) in the
capital city of each of the countries.
• Eligible subjects answered a 20 minutes face-to-
face survey (paper or computer based).
• Inclusion Criteria:
• Legal resident
• Able to talk and read English or Spanish
• Smokers (18 & over)/ Non-smokers (18-24)
• Male & Female
52. Members States of the CARICOM
ANTIGUA & GRENADA ST. KITTS & NEVIS
BARBUDA
BAHAMAS GUYANA ST. LUCIA
ST. VINCENT & THE
BARBADOS HAITI
GRENADINES
BELIZE JAMAICA SURINAME
TRINIDAD &
DOMINICA MONTSERRAT
TOBAGO
53. PROTOCOL
Respondents were shown a series of 24 cigarette
mock-ups and asked questions following each one.
Respondents were asked:
On a scale of 1 to 10, with 1 being “NOT AT ALL” and 10
being “EXTREMELY”, please tell me whether this warning
message…
• … grabs your attention
• … is believable
• … would make smokers want to quit
• … would help prevent young people from starting to smoke
Overall, on a scale of 1 to 10, how effective is this health
warning?
54. HIGHEST Rated
TOTAL SMOKERS NON SMOKERS
Theme /
Type
Health effect: Mouth Cancer /
GRAPHIC
Health effect: Mouth Disease
/
GRAPHIC
Health effect: Mouth Cancer /
GRAPHIC
Health effect: Stillbirth /
GRAPHIC
55. LOWEST RATED
TOTAL SMOKERS NON SMOKERS
Theme /
Type
Cessation/ Economic effect /
SYMBOLIC
Health effect: Aesthetic /
GRAPHIC
Cessation /
SYMBOLIC
Cessation /
SYMBOLIC
56. MARKER WORD
Respondents were asked to focus on the word in CAPITAL letters and different
color that appears at the beginning of these messages.
100
80
63 62 63
Warning
60
% Caution
40 29 29 30
Danger
20 11 12 10
0
Overall (n=1,441) Smokers (n=712) Non-smokers (n=729)
Which pack is most effective at getting people to think about the dangers of
smoking?
57. MARKER WORD
Respondents were asked to focus on the word in CAPITAL letters and different
color that appears at the beginning of these messages.
100
80 68 69 66
60 Warning
% Caution
40 27 27 27 Danger
20 11 12 10
0
Overall (n=1,441) Smokers (n=712) Non-smokers (n=729)
Which pack is most effective at getting people to think about the
dangers of smoking?
58. ATTRIBUTION
Respondents were asked to focus on the last words print in smaller font at the end of
the warning.
75
60
46 45 46 45 Minister of Health
43 40
45 Ministry of Health
%
30 Chief Medical Officer
11 12 10 None
15
2 3 1
0
Overall (n=1,441) Smokers (n=712) Non-smokers (n=729)
Which is MOST effective at getting people to think about the dangers of
smoking?
59. ATTRIBUTION
Respondents were asked to focus on the last words print in smaller font at the end of
the warning.
75
60
45 46 46 Minister of Health
43 44
45 40
Ministry of Health
%
30 Chief Medical Officer
11 13 10 None
15 4
3 2
0
Overall (n=1,441) Smokers (n=712) Non-smokers (n=729)
Which is MOST effective at getting people to think about the dangers of
smoking?
66. Bolivia Cigarette Labeling Study
Aim: To pre-test the effectiveness of pictorial-based HWLs printed on
cigarette packages to raise awareness on SHS and to increase support
for 100% smokefree policies in Bolivia (synergy of policies)
Examples of 4 types of pictorial HWLs on SHS & lung cancer to be tested
67. ACKNOWLEDGEMENTS
RPCI Dep. Health Behavior International Partners
• Michael Cummings • Center for Tobacco Epidemic
Research of Uruguay
• Mark Travers
• InterAmerican Heart Foundation
• Maansi Travers
• Ministry of Health of Panama
• Richard O’Connor
• National Cancer Institute of Brazil
• Andrew Hyland • National Institute of Public Health of
• Cheryl Rivard Mexico
• Essie Torres • Smokefree Alliance of Argentina
Univ. of Waterloo • Heart Foundation of Jamaica
• David Hammond • Heart & Stroke Foundation of
Barbados
Univ. of South Carolina
• Trinidad & Tobago Cancer Society
• James Thrasher
• Guyana Chest Society
68. Ernesto Sebrie, MD MPH
ernesto.sebrie@roswellpar
k.org
Roswell Park Cancer Institute
69. How to design the content of
pictorial warning labels:
A case study of Mexico
James F. Thrasher
70. Mexican health warnings labels (HWLs)
3rd largest in the world (65% of pack)
(2004-2010) (2010-present)
50% of the back (text only) 30% of the front (picture & text)
4 messages 100% of the back and one side
(text only)
2 new HWLs every 3 months,
fastest rotation in the world
71. 1st round of Mexican
Pictorial Health Warning Labels (HWLs)
72. 3-stage process to select HWL content
Phase 1 Phase 2 Phase 3
Objective: Objective: Objective:
Most Most Confirm
effective effective findings
HWL imagery HWL textual about HWL
content imagery &
Method: text
Field Method:
experiments Field Method:
experiments Focus groups
4
73. International Pack Study
Parallel studies in 7 diverse countries:
Second level
Mexico India South
Korea
Third level
US Germany
Fourth level
China Bangladesh
Fifth level
Examine effective content for pictorial warnings
Examine individual differences in responses
Examine cultural, national level differences
P01 CA138389, Tobacco packaging and labeling policies: 40
30
Warning Labels stopped you from having a
5
cigarette.
Canada
Australia
US
UK
Percentage
Expanding the evidence on novel policies
20
10
0
Wave 1 Wave 2
74. Adaptation of Health Warnings
Second level
Third level
Fourth level
Fifth level
Warning Labels stopped you from having a
cigarette.
40
Canada
6
Australia
US
30 UK
Percentage
20
10
0
Wave 1 Wave 2
75. Health warning topic “sets”
5 or 6 warnings tested for each topic
Text Symbolic Human suffering
Graphic External Graphic Internal Testimonial
77. Field experiments
Topics
Brief Survey Rate 5 to 7 Rank
Addiction
• Sociodemographics warnings from warnings
• Smoking-related Throat cancer selected block within a block:
perceptions & Mouth cancer in random order
behavior • Which
Death Evaluate for most
• Warning labels Premature aging each warning: motivates
Stroke you to stop
• Attention
Random smoking?
Impotence
• Credibility
assignment Lung cancer • Which
• Relevance most
Gangrene
motivates
• Negative
Heart attack you to not
emotional
Toxicity start
arousal
smoking?
Enphysema
• Impact
Premature birth
SHS in children
Hammond D, Thrasher JF, Reid J et al. Cancer Causes & Control. 2012.
78. Samples
• Mexico:
• Sample
– 535 16 to 18 years old, smokers and nonsmokers
– 527 adult smokers
• Mode
– Intercept surveys in Mexico City (July 2010)
– In-person administration
– Stimulus presentation by laptop
Hammond D, Thrasher JF, Reid J et al. Cancer Causes & Control. 2012.
79. Graphic more effective than human suffering,
which is more effective than symbolic warnings
(p<.001)
Second level
Third level
Fourth level
Fifth level
Human Symbolic
Graphic vs. vs.
suffering 40
30
Warning Labels stopped you from having a
11
cigarette.
Canada
Australia
US
UK
Percentage
20
10
Hammond D, Thrasher JF, Reid J et al. Cancer Causes & Control. 2012.
0
Wave 1 Wave 2
80. Adding testimonial information
increased effectiveness (p<.001)
Second level
Third level
Fourth level
Fifth level
Testimonial vs. Non-Testimonial
Warning Labels stopped you from having a
cigarette.
40
Canada
12
Australia
US
30 UK
Percentage
20
10
Hammond D, Thrasher JF, Reid J et al. Cancer Causes & Control. 2012.
0
Wave 1 Wave 2
82. Health Beliefs
Does smoking cause impotence in male smokers?
% “Yes”
Mutti S, Hammond D, Reid J, Thrasher JF. Jnl of Health Communicaiton. In press.
84. 2nd phase: Selection of textual content for
pictorial HWLs
Phase 1 Phase 2 Phase 3
Objective: Objective: Objective:
Most Most Confirm
effective effective findings
HWL imagery HWL textual about HWL
content imagery &
Method: text
Field Method:
experiments Field Method:
experiments Focus groups
85. Characteristics of Mexican pictorial HWLs
30% of front with
image and text
“Qualitative” focus
on toxic
constituents
Call to action with 100% of one side, reinforcing
1800 number the central message
86. Didactic/scientific vs. testimonial content
Didactic /scientific Testimonial - other Testimonial - self
Breathing tobacco “My husband was a “Tobacco smoke is a silent
smoke causes the smoker, and while he was killer. Without thinking, I
arteries of your heart to still young, he died from breathed it in, unable to
clog. The clogging a heart attack. I was left feel the damage it had
stranded to take care of done…until it gave me a
damages your heart and
the family all by myself.” heart attack.”
can kill you.
Celia Juárez César Guerrero
Contains particles that enter Contains particles that enter Contains particles that enter
your blood stream, form blood your blood stream, form blood your blood stream, form blood
clots and can block your clots and can block your clots and can block your
arteries arteries arteries
You can quit smoking. Call us You can quit smoking. Call us You can quit smoking. Call us
01800 966 3863 01800 966 3863 01800 966 3863
Thrasher JF, Arillo-Santillán E, Villalobos V, et al. Cancer Causes & Control. 2012.
87. Field experiments in 8 public places
Brief survey Condition 1: Random Ranking task
• Sociodemographics 4 blocks of packs presentation of HWLs
• Smoking on same health of packs within block:
perceptions & outcome within each
behavior • Which
Condition 2: block
• HWL motivates
4 blocks of packs Assessment you most
on same health of each HWL: to quit?
Randomization outcome
• Attention • Which
Condition 3: motivates
• Credibility
4 blocks of packs you most
on same health • Relevance to not start
outcome smoking?
• Emotional
Condition 4: arousal
4 blocks of packs • Impact
on same health
outcome
Thrasher JF, Arillo-Santillán E, Villalobos V, et al. Cancer Causes & Control. 2012.
88.
89. Sample characteristics
Youth sample Adult smoker sample
Characteristics % (n) % (n)
Age (average) 20.6 (range 18 – 24) 32.8 (range 25 – 80)
Sex
Male 50% (266) 62% (330)
Female 50% (263) 38% (199)
Education
Secondary or less 28% (146) 18% (100)
Technical school 8% (44) 14% (76)
High school 57% (302) 42% (223)
University + 7% (37) 25% (130)
Smoking behavior
Nonsmoker 49% (258) N/A
Non-daily smoker 30% (159) 43% (225)
Daily smoker, < 5 /day 12% (63) 25% (131)
Daily smoker, 5+ / day 10% (53) 33% (173)
Intend to quit in next 6 months 31% (77 / 252) 31% (165)
90. Overall impact of scientific vs. testimonial
7.5
Scientific
Testimonial 1
7.2
7.0 Testimonial 2
6.8 6.8 6.8 6.8 6.8
6.5 6.6 6.6
6.5 6.4 6.4 6.5
6.3 6.2 6.3 6.3
6.2 6.2
6.0
5.5
5.0
Lung cancer Heart disease Premature Stroke Addiction Throat
birth cancer
Thrasher JF, Arillo-Santillán E, Villalobos V, et al. Cancer Causes & Control. 2012.
91. Adjusted ORs* of selecting a testimonial as most
effective, lowest vs. highest educational attainment
4
3.5
3 3.0
2.9
2.5
2 2.0 1.9
1.7
1.5
1.3
1
0.5
Lung Heart Prem ature Stroke Addiction Throat
cancer disease birth cancer
*adjusted for age, sex, smoking status, level of consumption, perceived risk, and quit intentions
92. 3rd phase: confirm best combination of
imagery & textual content
Phase 1 Phase 2 Phase 3
Objective: Objective: Objective:
Most Most Confirm
effective effective findings
HWL imagery HWL textual about HWL
content imagery &
Method:
text
Field Method:
experiments Field Method:
experiments Focus groups
93. Focus groups
Mixed methods
Individual ratings
Group discussion
Callback two days later
12 groups, 4 in each city:
Mexico City Guadalajara Monterrey
Population Non- Non- Non-
Smokers Smokers Smokers
characteristics smokers smokers smokers
18 - 25 - 18 - 25 - 18 - 25 - 18 - 25 - 18 - 25 -
18 - 24
24 65 24 65 24 65 24 65 24 65
Female (n=80) 7 17 4 0 6 14 6 5 16 5 0
Male (n=67) 4 11 9 1 6 10 5 6 8 6 1
Total (n=147) 11 28 13 1 12 24 11 11 24 11 1
95. Conclusions
Graphic imagery was rated as more effective than
other types of imagery
Consistent across samples and countries
Didactic / scientific textual content was more
effective and credible than testimonial content
Amount of testimonial content may matter
Testimonials work best among lower educated smokers
Testimonials may be more effective with stronger TC
environment, or when complemented by other media
Textual content on toxic constituents effective
Quantitative and focus group results were consistent
96. What do we need?
A role for positively framed quit messages?
People who intend to quit have strongest responses
Habituation and “wear out”
Wear out quicker for some imagery than others
Cycles of wearin and wearout?
Matching/targeting of messages?
Industry responses
Image banks and sharing of high-quality pictorial
imagery among countries