|Home | About | Journals | Submit | Contact Us | Français|
The internet is a major source of health information and several notable health web sites contain information on the risks associated with cigar smoking. Previous research indicates that internet pages containing health information on cigars have high reading levels and are restricted to text material, which can decrease understanding. We examined the effects of existing text-only (from the United States National Cancer Institute website) versus novel graphic-enhanced information on smokers' perceptions of health risks associated with cigar smoking. The study was a laboratory-based single session of current cigarette smokers (n=102) who viewed cigar smoking risk information on a computer monitor then completed cigar risk questionnaire items. Participants were randomized to view either text-only or graphic-enhanced cigar information. The graphic version contained additional risk information about cigarillos and little cigars. Text-only participants were more likely to underestimate perceived health risks associated with cigar smoking compared to graphic-enhanced participants (47.1% versus 17.7%, p=.001); and, graphic-enhanced participants were more likely to report that they would share the cigar health risk information with friends compared to those viewing text-only, 47.0% versus 27.4%, p=.005. Employing graphics to convey health risks associated with cigar smoking increases understanding and likeliness to share information. Integrating information about little cigars and cigarillos risk in conjunction with large cigar risk information is an effective public health strategy to provide more comprehensive risk information. Utilizing graphics on health information internet pages can increase knowledge and perceived risks of cigar smoking.
The internet is a popular source of health information; half of US adults report searching the internet for health information (Cohen & Stussman, 2010). Several websites have health information on cigars, yet they have high reading levels, no graphics, and ignore little cigars and cigarillos (Dollar, Mix & Kozlowski, 2008). There have been few efforts to correct misperceptions of cigars as a less-hazardous product (Dollar, Mix, & Kozlowski, 2008). People often minimize the health risks of cigars (Baker et al., 2001; Nyman, Taylor, & Biener, 2002). Effective means of communicating risk information are needed on the internet to counter common beliefs that cigars are a safer alternative to cigarettes.
In the U.S, cigar sales, especially of little cigars and cigarillos, have risen dramatically in the past two decades, as cigarette sales declined (United States Department of Agriculture (USDA), 2007; Kozlowski, Dollar & Giovino, 2008). This rise was supported by lower tax rates on cigars and by beliefs that cigars are less dangerous than cigarettes (Campaign for Tobacco Free Kids, 2010).
Although cancer risk for cigar smokers depends on frequency of use and depth of inhalation, the risk is clearly far greater than for non-smokers (Shopland et al., 1998), and approaches level of risk for cigarette smokers for many cancer types (Baker et al., 2000; Shanks & Burns, 1998; Shopland et al., 1998). In the case of little cigars and cigarillos, inhalation patterns approximate cigarette smoking behaviors, and lung cancer risk is comparable to that associated with cigarettes (Boffetta et al., 1999; Henningfield, Hariharan, & Kozlowski, 1996). Compared to large cigars, for which risk is complex because it depends on smoking history and inhalation, the explicit dangers from little cigar smoking is more easily communicated. Yet, there is an absence of risk information on little cigars available on websites (Dollar, Mix, & Kozlowski, 2008).
We conducted a laboratory-based investigation on the effects of graphic-enhanced web-based messages to improve smokers' understanding of cigar smoking risks compared to text-only messages taken from the National Cancer Institute (NCI) website, (see Dollar, Mix, & Kozlowski, 2008) Research on graphic communications (Hammond et al, 2007) led us to hypothesize that participants viewing the graphic version would have a better understanding of the risks. Additionally, we included little cigar and cigarillo risk information in the graphic version.
Participants were 102 current smokers responding to advertisements for a 90 minute study on attitudes and beliefs about tobacco smoking. Eligibility was self-reported smoking a minimum 10 daily cigarettes (non-menthol) for a minimum 5 years; not currently trying to quit or no intent to do so in the next two months; age 21 to 65; English as primary language; no current substance abuse; no visual impairments. The protocol was institutional review board-approved and sessions conducted October 2008 - April 2009.
Participants smoked one of their own brand to standardize time since last cigarette, and completed demographic and smoking-related questionnaires administered using computer-based software (MediaLab v2004.3.14, Empirisoft) presented on 19-inch monitors. Participants viewed cigarette and potential reduced exposure products (PREP) materials (not reported) prior to the cigar module.
Participants were randomized, balanced for sex, to text or graphic conditions presented in Gazetracker (Eye Response Technologies), completed a 7-item Cigar Questionnaire, provided open-ended comments, and received $50.00.
Risk information was either text-only or graphic-enhanced with minimal text (see Figures 1A and 1B), presented as two successive screens (i.e., 1A-1 then 1A-2; or 1B-1 then 1B-2), congruent for occupied monitor space. Text-only was an unedited excerpt (243 of 1,525 words) from a NCI webpage (http://www.cancer.gov/cancertopics/factsheet/Tobacco/cigars) entitled “Questions and Answers about Cigar Smoking and Cancer” and was the answer to the question, “How are the health risks associated with cigar smoking different from those associated with smoking cigarettes?” (see Figure 1A). It was selected based on ratings of risk message, harm reduction theme and complexity level (Dollar, Mix, & Kozlowski, 2008).
The graphic version was similar to the text version and employed simple warning or traffic light graphics, derived from a traffic light information campaign in the United Kingdom on healthy eating (http:///www.eatwell.gov.uk/foodlabels/trafficlights/). Additionally, graphics to convey risks related to inhalation and the number of cigars smoked, as well as the risks of little cigar and cigarillo use were included. (Figure 1B).
Age, sex, education, race, ethnicity, preferred cigarette brand, number of years smoking, nicotine dependence [Fagerstrom Test of Nicotine Dependence (FTND; Heatherton, Kozlowski, Frecker, & Fagerstrom, 1991)], history of cigar use, and need for cognition (preference for thinking about complex issues; Cacioppo, Petty, Feinstein, & Jarvis, 1996) were assessed prior to viewing cigar risk information.
A 7-item Cigar Questionnaire test was used to assess participants' perceived risks of cigar smoking and attitude toward the information presented. Cigar items focused upon risk knowledge and the content of the information provided. Participants were instructed to select the response option that best reflects their answer to the following statements: How do cigars and cigarettes compare in risks to the users?; Little cigars are less dangerous than cigarettes; Inhalation of cigar smoke increases risks to health; How likely would you be to tell a cigar smoking friend or family member about what you learned from reading the webpage?; and acceptability: Was the web page on cigars interesting?; Was the web page on cigars annoying?; Did you like or dislike the webpage on cigars? The first item was scored on a 5-point response format (1=cigars are much less dangerous than cigarettes; 2=…a little less dangerous…; 3=…about as dangerous…; 4=… a little more dangerous…; 5=…much more dangerous…). Items 2-5 were scored on a 5-point response format (1=not at all; 2=a little; 3=moderately; 4=very much; 5=extremely); and the last item on a 4-point response format (1=dislike it greatly; 2=dislike it a little; 3=like it a little; 4=like it greatly).
Analysis of Variance (ANOVA) assessed condition effects on cigar risk perception and attitude toward messages, with covariates p<.20 retained in models. Fisher's Exact tests were performed post-hoc to explore dichotomous differences [less versus more (item 1); not at all versus any level (items 2-6); dislike versus like (item 7)].
Participants were 67% male, and 90% White, 5% Asian American, 3% African American; 2% Hispanic irrespective of race; mean age, 30.5 years (SD=8.1). All participants completed High School; 42% college graduates; 74% never married; 37% currently had full-time work, 34% part-time work, and 28% not working. Average number of daily cigarettes was 16.5 (SD=5.6; range 10-40); participants started smoking at age 16.5 (SD=2.7) and had been daily smokers for 12.8 years (SD=6.9). Mean nicotine dependence score (FTND) was 4.1 (SD=2.1; range 0-9).
Eleven participants reported smoking a cigar at least 50 times in their entire life and 6 currently smoked cigars, similar to that observed in the US population (5.8%; Centers for Disease Control and Prevention, 2006). Thirteen participants reported ever smoking little cigars, 9 reported cigarillo use, and 23 reported large cigar use; 72 reported having never tried any type of cigar.
Text-based and graphic-based groups did not differ with respect to gender, education level, number of ever or current cigar smokers, age, FTND or need for cognition, but did smoke more daily cigarettes [18.0 (SD=6.2) versus 15.0 (SD=4.6), p<.01].
Text participants scored lower [2.55 (SD=0.8)] than graphic participants [3.18 (SD=0.9)] on, How do cigars and cigarettes compare in risks to the users, (F=4.897, p=.001), controlling for age, daily cigarette count and need for cognition score. Among text participants, 47.1% underestimated the health risks of cigars by responding that cigars are much less or a little less dangerous than cigarettes. Only 17.7% of graphic participants underestimated the health risks of cigars (Figure 2). A similar trend was found for perceived health risk associated with smoking little cigars/cigarillos compared to cigarettes (F=2.4, p=.12). Fisher's Exact test for item 2, Little cigars are less dangerous than cigarettes?, indicates a significant difference between text and graphic versions when comparing ‘not at all’ versus ‘any’; 82% of those in the graphic group reported little cigars were not less dangerous than cigarettes compared to 57% of the graphic group (Fisher's p-value=0.009).
Graphic participants reported being more likely to tell a friend or family member about what they had learned from the webpage [3.26 (SD=1.2)] than were text participants [(2.57 SD=1.2), F=2.44, p=.005], controlling for age, daily cigarette count and need for cognition. Half of the text participants (49%) were not at all or a little likely to share what they had learned; 27.4% of text participants were very or extremely likely to share the cigar risk information. Alternately, 25.4% of graphic participants were not at all or a little likely to share the information they had learned; 47% were very or extremely likely to share the cigar risk information (Figure 2).
There were no other significant differences between graphic and text participants (p's > 0.2); they equally understood that inhalation of cigar smoke increases risks to health [4.43 (SD=0.7) versus 4.45 (SD=0.88)]; found the web pages moderately interesting [2.94 (SD=0.84) versus 3.07 (SD=1.00)]; found the web pages not to be annoying [1.45 (SD=0.76) versus 1.51 (SD=0.83)]; and had mild liking toward the web pages [2.78 (SD=0.58) versus 2.90 (SD=0.58)].
Participants who viewed the graphic version of the cigar risk message were more likely to perceive cigar smoking risks at a level consistent with the scientific literature. Fewer graphic participants underestimated the health risks of smoking cigars relative to text participants (17.7% versus 47.1%). The graphic condition emphasized the relative health risk posed by cigar smoking compared to cigarette smoking: specifically, that cigars carry risks equivalent to cigarettes if you have previously smoked cigarettes, inhale or use frequently. The text version included information on the relative risk of cigars to cigarettes that is no less accurate, but as previously reported, the text has a high reading level and is more open to interpretation (Dollar, Mix, & Kozlowski, 2008).
Those viewing the graphic version were also more likely to report that they would share the information with others. The pattern of results indicates that three-quarters of the graphic participants were at least moderately likely to share cigar risk information, compared to half of those the text participants. A 50% increase in intention to share cigar smoking risk information could have meaningful public health implications. Interpersonal health communication, especially from non-health professionals (e.g., family and friends), is linked to smoking cessation attempts (Korhonen et al., 1998). Future research should identify factors that increase sharing risk information, possibly including believability, trust or perceived truth of graphic information (Schapira, Nattinger, & McAuliffe, 2006; Lipkus, Klein, & Rimer, 2001; Lipkus & Hollands, 1999).
Internet sites can reach many individuals with effective health information at a low cost (Strecher, 2007). However, it is critical that the information is accurate and understandable. If main web pages presented understandable graphic-based information then even those giving a cursory inspection might accurately obtain important health information. More detailed information on hyperlinked pages could also be available. Seeking and scanning for health information can lead to increased health knowledge and improved health behaviors (Shim, Kelly, & Hornik, 2006). Knowledge acquired from accurate, succinct statements of smoking risk have been retained up to 9 months later (Kozlowski et al., 2001). Counter marketing the risks of tobacco products marketed as harm-reducing has been effective in tobacco control, notably with light cigarette labeling (Kozlowski et al., 2001), and filter ventilation risks (Kozlowski et al., 2000). Possibly an efficiently and effectively conveyed warning about cigars, that addresses little cigars and cigarillos, could improve risk understanding and reduce the severity of this emerging health threat (Symm et al., 2005).
This study has limitations. Cigar perceptions were assessed only after viewing the web pages, to avoid priming participants with the items we would be assessing. However, our groups may have differed in cigar knowledge prior to our study. Also, the cigar protocol was part of a larger study in which participants also viewed advertisement materials for potential reduced exposure products and cigarettes prior to completing the cigar module; however none were different than what might be encountered daily. There were very few cigar smokers in the study. However, all participants were current cigarette smokers, and therefore at risk to erroneously consider switching to cigar smoking to reduce harm. Further, 70% of the sample reported ever smoking a cigar and those who inhale intensely on a cigar, as one does on cigarettes, are at increased risk from cigars (Baker et al., 2000).
The study explored if perceived risks of cigar smoking might be improved by employing a graphic format, in comparison to existing text-based information from an a major website. Results indicate that those viewing graphic versions of the message were significantly more likely to understand the health risks of cigar smoking, and were more likely to intend to share this information with friends and family compared to those viewing a text-only message. Presenting messages about the health risks of cigars in a graphic format might lead to more accurate perceived risk and increased sharing of risk information.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.