Article Text

Original research
How do current tobacco warnings compare to the WHO FCTC guidelines: a content analysis of combustible tobacco warnings worldwide
  1. Leah M Ranney1,2,
  2. Sonia A Clark1,
  3. Kristen L Jarman1,
  4. Allison J Lazard2,3,
  5. Sarah D Kowitt1,2,
  6. Jennifer Cornacchione Ross4,
  7. Guido Baler1,
  8. James F Thrasher5,
  9. Adam O Goldstein1,2
  1. 1Department of Family Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
  2. 2Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC, USA
  3. 3Hussman School of Journalism and Media, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
  4. 4Department of Health Law, Policy, and Management, Boston University, School of Public Health, Boston, MA, USA
  5. 5Department of Health Promotion, Education & Behavior, Arnold School of Public Health, University of South Carolina, Columbia, SC, USA
  1. Correspondence to Dr Leah M Ranney; Leah_Ranney{at}


Objective Many countries have adopted warning labels for combustible tobacco products, yet little research exists describing tobacco warning characteristics globally and to what extent they meet the WHO Framework Convention for Tobacco Control (FCTC) Guidelines. This study evaluates characteristics of combustible tobacco warnings.

Design We conducted a content analysis to describe the overall landscape of warnings using descriptive statistics and compared to the WHO FCTC Guidelines.

Setting We searched existing warning databases for combustible tobacco warnings from English-speaking countries. We compiled warnings meeting inclusion criteria and coded for message and image characteristics using a pre-defined codebook.

Primary and secondary outcomes measures Characteristics of combustible tobacco warning text statements and warning images were the primary study outcomes. There were no secondary study outcomes.

Results We identified a total of 316 warnings from 26 countries or jurisdictions worldwide. Of these warnings, 94% included warning text and an image. Warning text statements most often described health effects to the respiratory (26%), circulatory (19%) and reproductive systems (19%). Cancer was the most frequently mentioned health topic (28%). Fewer than half of warnings included a Quitline resource (41%). Few warnings included messages about secondhand smoke (11%), addiction (6%) or cost (1%). Of warnings with images, most were in colour and showed people (88%), mostly adults (40%). More than 1 in 5 warnings with images included a smoking cue (ie, cigarette).

Conclusions While most tobacco warnings followed WHO FCTC guidance on effective tobacco warnings, such as communicating health risks and inclusion of images, many did not include local Quitline or cessation resources. A sizeable minority include smoking cues that could inhibit effectiveness. Full alignment with WHO FCTC guidelines will improve warnings and better achieve the WHO FCTC objectives.

  • Public health
  • Health & safety
  • Health policy

Data availability statement

Data are available upon reasonable request. Please email a reasonable request to the corresponding author for access to the research data.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • This research is the first content analysis to quantitatively describe combustible tobacco warning labels.

  • This research assesses the extent to which current tobacco warnings meet WHO Framework Convention for Tobacco Control guidelines.

  • The study collected only English language warnings.

  • The tobacco warnings were identified through warning database searches; warnings that were not available online were not included.

  • The research did not examine behavioural assessments of warnings by viewers.


Globally, over 8 million deaths are attributed to tobacco use each year, either from direct use or exposure to secondhand smoke.1 In response to the epidemic of tobacco use globally, the WHO Framework Convention on Tobacco Control (FCTC) was formed in 2003 to uphold the right to health of all persons through the promotion of effective tobacco control practices.2 Article 11 of the WHO FCTC treaty specifies that agreeing parties/countries should implement pictorial tobacco warning labels that cover at least 50% of the primary pack surfaces.3 As of January 2021, 112 countries and jurisdictions have pictorial warnings with a size of at least 50% of the principal display on the pack.4 Warning labels on tobacco product packaging communicate risk to both current and potential consumers. Effective warnings can increase knowledge and awareness of tobacco-related risks, increase intentions to quit and reduce tobacco product use overall.5 Therefore, the successful implementation of these recommendations is critical to global tobacco control and upholding the WHO FCTC’s overall objective to prevent and reduce harms caused by tobacco use.6

Guidelines from WHO FCTC Article 13 explicitly recommend removing advertising and promotion on tobacco product packaging, including all design features that make tobacco products attractive.7 Research supporting this recommendation concludes that plain packaging with pictorial health warnings increases visual attention to warnings, increases harm perceptions,8 and reduces pack appeal,9 but may not increase the effectiveness of the health warning labels.10 As of October 2020, 17 countries have adopted plain packaging.11

Currently, over 182 Parties/countries and jurisdictions (ie, 90% of the world population) have signed the WHO FCTC treaty in agreement that they will strive to support and ratify measures outlined above.12 To assist Parties in fulfilling the aim of Article 11 and the adoption of these recommendations, the WHO provides guidelines on what requirements should be adopted and proposes best practices in the development of policy measures by a Parties’ Health Authority.3 Key recommendations include having a variety of warning labels that clearly communicate health risks as well as different issues related to tobacco use; such as, advice on quitting, the addictive nature of tobacco and adverse economic outcomes.3

The WHO guidelines also recommend several design elements for tobacco warnings, including: location, size, use of pictorials, colour, rotation, message content (ie, tone of warning), language and source attribution.2 These recommendations reflect evidence-based research and collective understanding of best practices. Current research suggests that larger warnings with pictorial images are more likely to be noticed and more effective in communicating the health risks of smoking.3 13–20 Compared with text-only warnings, pictorial warnings are rated as more personally relevant,21 more likely to draw attention and be remembered,3 promote cessation attempts22 and decrease consumption.23 Additional research indicates tobacco warning labels that offer Quitline numbers and websites are an important source of cessation information; such labels are associated with self-report of using a cessation resource.24 25 Finally, use of the marker word ‘WARNING’ on warning labels appears more impactful than labels without a marker word.26

While research exists that examines WHO FCTC implementation,27–29 and historical evolution and diffusion of warnings,30 no published research describes current tobacco warning characteristics globally and to what extent they meet WHO guidelines.3 Though research indicates graphic warnings can produce unpleasantness for non-smokers and increase smokers’ intentions to quit,31 the ideal type of health effect or number of health effects to include has not been examined on tobacco warnings. A meta-analysis of single exposure experiments from 16 countries concluded that pictorial cigarette pack warnings were more effective than text-only warnings.16 However, pictorial warnings also vary widely across the world.

Visual design principles (ie, images of people, colour, eye gaze) can play an important role for increasing cognitive processing of tobacco warning labels.32 Conducting a review of design principles represented in combustible tobacco warnings is important. The purpose of this study was to systematically describe the key characteristics of existing combustible tobacco warnings to better understand how these warnings compare to existing warning guidelines and research.


Identifying warnings

To compile the warning database, we identified countries that use English as a de facto or official language.33 We then searched databases from the Campaign for Tobacco Free Kids34 and the WHO35 and identified combustible tobacco warnings that include English text statements on 15 October 2019 and 21 July 2020. Our initial tobacco warning data collection was initiated as part of a NIH (National Institute of Health) funded study, and in July 2020 we reviewed the databases a second time to collect any warning updates. We coded the tobacco warning statements for known health topics, such as cancer, addiction, secondhand smoke and chemicals, and added topics discovered during coding with an iterative review process. The research team collaboratively developed a codebook to capture the detailed information about each warning collected. The tobacco warning codebook with the complete list of text and pictorial warning codes is in the online supplemental materials. The coders for this study were public health professionals with advanced degrees who received training on the codebook prior to conducting the tobacco warning coding. With the exception of the USA, all countries included in our sample have formally entered into the force of the FCTC (either through ratification, acceptance, approval, formal confirmation or accession) (online supplemental table A). The research in this manuscript is non-human subject and therefore, the University of North Carolina at Chapel Hill determined it was exempt from IRB (Institutional Reveiw Board) approval.

Text warning statements were coded for a variety of characteristics, including: the country using the warning, the type of tobacco product mentioned in the warning (eg, cigarette) or non-specific tobacco product mentioned (eg, ‘smoking’), the health system described in the warning (eg, respiratory); the number of health effects mentioned within a warning, the presence of Quitline information; presence of a marker word (eg, ‘WARNING’) and any source attribution (eg, government authority) of the warning. We coded the tobacco warning statements by specific topics, such as cancer, addiction, secondhand smoke and chemicals. The pictorial warnings were coded for constructs related to visual communication. Figure 1 depicts an example of how codes were applied for selected characteristics of a pictorial tobacco warning.

Figure 1

Example of how codes were applied to combustible tobacco warnings. Note: Image is from a public database.

Coding process

Each tobacco warning was double coded by two researchers using the codebook developed by the research team. The research team collaboratively reviewed the first round of coded tobacco warnings and revised the codebook to add additional categories and clarify the codebook. To ensure consistent coding across coders, any differences between the two coders were flagged and a third coder independently resolved the discrepancies before the database was finalised. The tobacco warning documents were re-accessed on 11 July 2020, and any updated warnings were added to the database and coded.

Patient and public involvement

Patients and members of the public were not involved in the design nor the conduct of the study.


We generated frequencies and proportions for each coded characteristic. For our analysis of warning statements, text statements that appeared in multiple tobacco warnings within a country or jurisdiction were counted only once. The number of unique text statements (n=274) was used as the denominator in all text characteristic calculations. For coding of pictorial warning characteristics (eg, # of people in warning), only tobacco warnings that had an image were included (n=279). When a code was not relevant—for example, for the code ‘Gender of Subject’, but the pictorial warning had no human present—we coded as ‘no human present’ and removed it from any analyses containing humans. We present descriptive statistics (frequencies and proportion of each code) and report relevant tobacco warning characteristics in the results as they related to the WHO FCTC Framework.


We identified and examined a total of 316 tobacco warnings from 26 countries and jurisdictions (the number of warnings from each ranged from 2 to 24) around the world (figure 2). The 316 tobacco warnings included 274 distinct text statements. To deduplicate text statements, we removed any duplicate text within the same country or jurisdiction. For example, if the text statement ‘Smoking Kills’ appeared in multiple tobacco warnings within the same country or jurisdiction, it was only counted once for analysis. Of these 316 warnings, only 6% did not include an image (n=19) and were text-only. The 297 pictorial tobacco warnings included a text statement paired with one or more images, with the majority of these (78%) including a unique text statement paired with a unique image (n=233). The combustible pictorial tobacco warnings collected for this study were most often from the UK (n=42), Canada (n=37) and Jamaica (n=31).

Figure 2

Countries in blue denote location of English language text and pictorial warnings. Note: aCountries and jurisdictions in blue were included in study. bOnline supplemental table A provides a breakdown the number of warnings each country added to in the study.

Tobacco warning text statements

We identified 274 unique text warning statements among the 316 tobacco warnings. The text statement included on most of these combustible tobacco warnings (87%) referred to tobacco products broadly (eg, smoking/tobacco causes cancer); a minority (10%) mentioned a specific tobacco product, such as ‘cigarette’ or ‘cigar’. A warning marker word appearing prior to the warning statement was present 57% of the time, and the word ‘WARNING’ was the most commonly used marker word (49%). A third of the warnings (35%) had a source attribution present (eg, ‘Ministry of Health’), indicating the scientific origin of the statement. Seven countries in our sample included warnings with a source attribution (Pakistan, USA, Jamaica, Canada, New Zealand, Sri Lanka or Australia). Quitline phone numbers or cessation websites were included on 41% of warnings. Of the 316 warnings examined, 17% (53/316) included the following elements on a single warning: a marker word prior to specifying a tobacco product, cessation resources (ie, quitline phone number or website) and a pictorial that was not a smoking cue.

Cancer was the topic mentioned most often (28%) in the warning statements. Fewer warnings mentioned themes of secondhand smoke (11%), addiction (6%) or financial cost (0.7%). The respiratory system, comprised of the mouth, throat and lungs, was the most frequently mentioned health system in the warnings (26%). Over half of warning statements focused on one health effect (57%), 37% of warnings mentioned multiple health effects and 6% of warnings did not mention a specific health effect.

Warnings often included multiple colours of text (62%) and in the backgrounds (42%). When warnings used a single colour for the text, it was either white (23%) or black (10%). Similarly, solid colour backgrounds were mainly black (31%) or white (11%). Table 1 describes the coded characteristics mentioned in the tobacco warning statements and the prevalence of health systems specified in tobacco warning statements.

Table 1

Frequency of tobacco text warning statement characteristics (unique text statements within each country/jurisdiction only; n=274)

Pictorial warnings

There were 297 pictorials identified among the 316 tobacco warnings. The key pictorial warning characteristics are synthesised in table 2. Most (70%) included the presence of a single person. When a person was depicted, it usually was an adult (40%), with children represented less frequently (19%). Many pictorial warnings showed a body part (ie, no face) that was in black and white and/or unclear; thus, we were often not able to determine gender, race or gaze direction of the people shown. When these variables could be discerned from the pictorials, a little more than a third majority of subjects were men (31%), white (34%) and the gaze direction was pointed away from the viewer (36%). When faces were visible on tobacco warnings, they were primarily white males (34%). More than 1 in 5 pictorial tobacco warnings (23%) included a smoking cue (ie, cigarette, smoke from cigarette).

Table 2

Frequency of tobacco pictorial warning characteristics (exclude text-only warnings; n=297)


Our study sought to characterise combustible tobacco warnings used in different countries and examine how consistent these tobacco warnings are with existing WHO Guidelines and research. WHO guidelines recommend several design elements for tobacco warnings, including use of pictorials, colour, message content (ie, tone of warning), language and source attribution.2 While our research shows that many key WHO recommendations are being followed, many combustible tobacco warnings diverge in important ways from these guidelines and best practice research. Study findings indicate the WHO recommendations that are most frequently implemented include a variety of health effects to communicate health risks, broad use of colour depictions and pictorial warnings to illustrate effects. The majority of warning statements analysed mentioned at least one or more specific health effects that are associated with tobacco use which directly adheres to WHO guidelines.

Pictorial warnings identified for our study were primarily from the UK, Canada and Jamaica. Like Canada, which was the first country to implementation photographic pictorial warnings,30 the UK and Jamacia signed and ratified with the WHO FCTC early in 2004 and 2005, respectively (see online supplemental table A). Therefore, it is understandable that these countries would have a wide variety of unique pictorial warnings. However, the number and frequency by which new health warning labels are released is not standardised; thus, variation in the number of warnings released is expected. Lastly, our findings were limited to the database reviewed. Future research will need to understand the variation in the number of warnings available within countries requiring pictorial tobacco warnings. Our analysis of pictorial warnings incorporated some favourable visual design aspects including colour and human imagery.32 36 37 Visual design principles in the warnings appear to increase cognitive processing. For example, images of the human body may increase the likelihood that viewers will attend to the visual; gaze direction of the image can influence direction of viewer attention32 and the gender and sex of the image can influence visual attention and memory.38 Many tobacco warnings in our study contained human imagery showing an eye gaze direction away from the viewer. Advertising research indicates that images of faces capture consumers’ visual attention, and eye gaze toward the object of interest is an effective tool for driving viewers’ attention toward specific products and intentions to purchase products.39 Over a third of warnings in our study that showed faces directed the eye gaze away from viewer. This may be a missed opportunity to increase the impact of tobacco warnings for smokers. More research in this area is needed to fill this gap.

When faces were visible on tobacco warnings, they were primarily white males (34%). Two studies found pictorial tobacco warnings produce minimal differences in ratings across racial groups suggesting that pictorial tobacco warnings are one of the few tobacco control policies that do not widen health disparities.21 40 However, one study that compared the perceived effectiveness of FDA (US Food and Drug Administration)-proposed graphic warning labels reported that lesbian, transgender or gender non-conforming persons evaluated the graphic warnings as less effective.41 While research in this area suggest inconsistencies, our study finding is concerning considering that large tobacco-related health disparities exist. The lack of personal characteristics like race and gender in tobacco warnings directly differs from the WHO recommendation that personalising tobacco warnings help increase believability and relevance. Thus, the inability to discern personal characteristics may decrease message relevance by limiting the users’ ability to identify with the person shown in the warning.42 At the same time, however, the inclusion of these characteristics may cause some people who do not share these characteristics to perceive the warning as less relevant. This tension creates some challenges when selecting the most effective imagery for generating relevance and impact, helping to explain why many warnings do not include identifiable personal characteristics.

Guidance by WHO suggests that colour enhances the noticeability of warnings and suggests that contrasting colours be used for the background of the text.2 Colour is an aspect of all visually represented messages, with colour combinations and polarity impacting reading speeds.43 Data on specific colours for warning text and background are sparse, though a recent study of e-cigarette warnings on advertisements reported warning recall was greater for warnings that featured black text on white or yellow backgrounds.44 Almost one-third of the warnings in our study used multiple colours for text and background, consistent with the contrasting colours recommendation and visual design research.

Unfortunately, almost a quarter of tobacco warnings presented a smoking cue, such as a cigarette, which may lessen the intended effect of the warnings. Pictures of cigarettes often increase cravings to use tobacco rather than inhibit tobacco use,45 46 as found in experimental research on responses to the first round of pictorial warnings in Brazil. As a result, Brazil eliminated cigarette imagery from its second round of pictorial warnings.47 Until further research elucidates the effects of smoking cues on tobacco warnings, it is prudent to eliminate smoking cues in tobacco warnings.

Surprisingly, slightly less than half of the messages had a Quitline resource and only 15% mentioned quitting as a key warning topic. Including cessation information on warnings is vital and increases awareness and utilisation.23 25 For example, Health Canada introduced new tobacco packaging warning labels featuring a prominently displayed toll-free number for a Quitline in 2012, and the average monthly call volume increased by 160% during the first 7 months.48 However, many countries, particularly low and middle income countries, do not have such services to promote through warnings,49 which may help explain our results. Further, marker words were present in only slightly over half of the tobacco warnings we examined. However, the use of a marker word may influence a viewer’s attention. Using a marker word like ‘WARNING’ before the health warning text can increase message recall, message effectiveness, fear and thinking about harms.26 44

Several areas of warning content suggest areas for future research. For instance, tobacco warning content should serve as a mechanism to communicate health risks, and WHO Guidelines recommend using a range of health warning content that elicits negative emotional associations with tobacco use to increase warning effectiveness.3 Over half of the tobacco warnings in our study mentioned a single health effect, and over a third mentioned multiple health effects. At present there are no published studies examining the effectiveness of warnings with multiple health effects (ie, heart disease and lung cancer) compared with a single health effect (ie, heart disease). Therefore, it is unknown if mentioning multiple health effects like many of the warnings in our sample, increases warning effectiveness. Future research may be able to obtain additional information about health warning labels by conducting a content analysis comparing the FCTC implementation database from country reports and the WHO database. Additionally, tobacco warnings can be assigned to each country or a group of countries that implement the warning, and then observations can be calculated on a country-level rather than a warning-level to demonstrate how WHO FCTC is implemented in each country. Finally, research that solely focuses on the visual aspects of tobacco warnings may garner important information to increase tobacco users’ attention to warnings and assist with cognitive processing of the warning text.

Our findings also indicate that tobacco warnings stated various health effects associated with tobacco use, but almost always used in broad terms rather than mentioning specific tobacco products. While WHO specifically recommends personalising tobacco warnings to help with believability and relevance, it is unclear whether this recommendation should be interpreted to use precise terms for tobacco products in warning such as ‘cigarette smoke’ or ‘cigar use’. More research is needed on how people interpret broad product terms (eg, tobacco) compared with specific terms (eg, cigarettes). For instance, are warnings that specifically refer to cigarettes disregarded by those who smoke cigars?

Research is inconclusive about adding source attribution to warnings to bolster credibility.50 51 If the country or jurisdiction require a source attribution, WHO Guidelines recommend that the source be a credible expert and the statement should not detract from the warning. While more research is needed to clarify whether adding a source to tobacco warnings increases intentions to quit, we found seven countries (Pakistan, USA, Jamaica, Canada, New Zealand, Sri Lanka or Australia) included one or more warnings with a source attributed (eg, ‘Health Canada’).

Finally, every tobacco warning should feature as many key recommendations and attributes from best practice research as feasible. However, very few tobacco warnings in our study included multiple attributes (ie, pictorial (non-smoking cue), cessation resources, marker word). Research is needed to determine what combination of tobacco warning design elements are the most effective.

Our study has several limitations including that we collected only English language warnings, and we have no data regarding which pictorial warnings and characteristics were associated with positive (eg, increase intentions to quit, quit attempts) and negative (eg, increase cravings) behaviours by viewers. In addition, most warnings were identified through warning database searches; warnings that were not available online may not be represented. Since some of the tobacco warnings were from countries where English is the second language, we did not measure plain or concise language. Also, the clarity of some tobacco warning images was poor, making it difficult to identify and code all the warning characteristics. Finally, the tobacco warnings were not viewed on packages, hence warning size and location could not be established and no information on warning rotation was assessed; however, the size of tobacco warnings has been increased over time in many countries.52 The features discussed in this study are best practices but provide no guarantee of increased effectiveness of pictorial warnings. Tobacco warnings are best informed by best practice and evidence of warning effectiveness in the population where they will be implemented.


Our systematic evaluation of English language combustible tobacco warnings confirms that the landscape of tobacco warnings can improve considerably to follow proposed WHO FCTC Guidelines and best practice research. While our findings show adherence to many key recommendations, we found few warnings advised quitting, provided Quitline resources, used a marker word such as ‘WARNING’, and too many warnings showed smoking cues. Increasing the strength of tobacco warning labels should apply evolving evidence-based design principles while carefully adhering to the WHO guidelines to improve tobacco warning attention, engagement and recall among tobacco users.32

Data availability statement

Data are available upon reasonable request. Please email a reasonable request to the corresponding author for access to the research data.

Ethics statements

Patient consent for publication

Ethics approval

Not applicable.


We would like to acknowledge Dr. Margaret Gourlay for contributing her time and ideas for this manuscript.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors AL, AG, KJ, JFT, LR, JC and SK contributed to conceptualisation. SK, KJ, LR, GB and SC contributed to data curation. KJ, SC, AL, AG, JFT, JC and SK contributed to methodology. SC, KJ and LR contributed to formal analysis and investigation. KJ, LR and AG contributed to project administration. LR, SC and KJ contributed to writing-original draft preparation. GB, SC, KJ, SK, AL, JFT, AG and JC contributed to writing-review and editing. AG contributed to funding acquisition. KJ and LR contributed to supervision. LR is the acting guarantor.

  • Funding Research reported in this publication was supported by National Cancer Institute and US Food and Drug Administration, Center for Tobacco Products (CTP) under Award Number R01CA240732.

  • Disclaimer The content is solely the responsibility of the authors and does not necessarily represent the official views of the funders or National Institutes of Health.

  • Map disclaimer The inclusion of any map (including the depiction of any boundaries therein), or of any geographic or locational reference, does not imply the expression of any opinion whatsoever on the part of BMJ concerning the legal status of any country, territory, jurisdiction or area or of its authorities. Any such expression remains solely that of the relevant source and is not endorsed by BMJ. Maps are provided without any warranty of any kind, either express or implied.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.