The goal of this study was to examine the content and quality of WATIs that are most likely to be encountered by smokers looking for online cessation assistance. We were also interested to see whether there were changes in quality between websites reviewed in the current study and those of our previous review, published in 2004 [5
]. Both studies used standardized procedures and assessment instruments when evaluating websites, and, in general, findings were similar for both reviews. For example, the percentage of websites meeting exclusion criteria was very similar between studies (77%, 156/202 in 2004 and 74%, 65/88 in 2007). In both cases, the most common reasons for exclusion were sites offering only product sales and undirected libraries of articles about smoking, smoking cessation, tobacco advocacy, and other tobacco-related topics. Results from both studies indicate that individuals searching for help with quitting smoking are most likely to encounter websites that do not offer smoking cessation treatment. Web-based interventions should include a clear organizational structure that actively guides users through the treatment process.
Results of this study indicate that while the majority of reviewed websites provide coverage for most key content areas identified as the core of smoking cessation treatment [8
], the depth of coverage for key topics was most often minimal. While there are no empirical studies demonstrating how much detail is needed to help smokers quit, it is likely that providing more than a brief mention of important content areas would result in better treatment outcomes. It may be particularly useful for websites to be designed in such a way as to allow users to drill down to their desired level of detail on any given topic. Sites that provide only minimal coverage of important information may do a disservice to smokers who are seeking to quit and need additional information to enhance the quit attempt. However, it is also likely that websites that present too much detailed information on each page may result in users missing important content. Thus, it is not sufficient to consider only general standards of usability [23
] when designing a behavior change website; it is also critical to understand the ways in which individuals use websites to make behavioral changes (ie, behavioral informatics). For example, some users may prefer to read science-based resources such as quitting guides or published manuscripts, while others may prefer to connect with other smokers in a community forum. Some may feel comfortable using interactive features that yield individually tailored information, while others may have concerns about privacy. Understanding the ways in which users interact with a cessation website and the relative contributions of various treatment components will help advance the science of Web-based behavior change.
Providing practical counseling was the key content area that received the most coverage: all websites provided at least minimal coverage of this topic. This is not surprising given that providing practical counseling and information could be described as the core content area of tobacco dependence treatment. However, within this topic, few websites prompted users to reflect back on prior quit attempts. Such reflection is important as it helps smokers to identify triggers, situations that are high risk for relapse, and techniques they found useful and could employ again [8
]. Likewise, while most sites recommended the use of pharmacotherapy, most often NRT, the information provided tended to be superficial and was limited by a lack of explanation regarding instructions for use, contraindications, and potential side effects. This problem was noted in our earlier study and has been discussed in other reviews [6
]. Content relevant to medication use was also the most likely of all content areas to contain serious errors. In some cases, the error was implicit in that NRT was included in a list of other unproven or unsafe alternatives, which lends a halo of legitimacy to those alternative treatments while also reducing the relative strength of the recommendation for established efficacious treatments such as NRT. In other cases, errors were more explicit, such as stating that using NRT is dangerous and should be avoided.
There was a notable lack in using the capacity of the Internet for personalization of treatment. An important part of motivating smokers to quit is to personalize information relevant to quitting. Reasons for quitting, perceived risks of continued smoking, as well as perceived benefits and barriers to quitting should all be identified by the individual to have maximal impact [25
]. The PHS guideline specifies that smoking cessation interventions should encourage smokers to discuss their reasons for wanting to quit and should provide personalized information about the risks of smoking and rewards of quitting. While the majority of websites provided information about the risks and rewards of quitting smoking, these key components were usually treated only with generic lists of benefits and health risks rather than personally relevant messages as specified by the PHS guideline. Approximately one-quarter of websites asked users to identify consequences of tobacco use and benefits of quitting that were personally relevant. Few websites took advantage of the Internet’s unique ability to provide individually tailored feedback. In the current study, benefits of quitting and risks of continued smoking were most often presented as generic lists, with no attempt at personalization.
Related to this issue, the interactive capability of WATIs was generally underutilized. Across all of the key treatment components, only a minority of websites provided interactive features. The most common use of interactivity was in the area of providing intra-treatment social support, frequently in the form of chat rooms, buddy lists, and emailed support. Recommendations for pharmacotherapy were frequently interactive in nature, although limited to the administration and scoring of assessments of nicotine dependence. Perhaps the most glaring failure to leverage the capabilities of the Internet was in providing follow-up contact. Follow-up contacts can be used to motivate smokers to make a quit attempt or to reconsider cessation following slip/relapse and to provide support during difficult times while quitting [8
]. Providing follow-up was one of the least used key treatment components observed in WATI sites. In the present study, just over one-fifth of websites provided any sort of treatment follow-up.
Compared with our earlier review, the current results indicate modest improvement in the quality of coverage in key content areas. Areas showing the most improvement were giving personalized advice to quit, providing practical counseling, recommending medications to aid quitting, and enhancing motivation (discussing personal relevance of quitting, perceived risks, and roadblocks to quitting). In no case did we observe a significant decline in the quality of website content. It is encouraging that the quality of some content areas may be improving. However, there remains substantial room for further improvements. Most often, the reviewed websites provided only minimal coverage of key component areas. Only in discussions of risks and roadblocks did most websites provide more than adequate or extensive coverage, and even in these areas, few websites took advantage of the interactive capacity of the Internet to truly personalize treatment.
Results of this study should be considered in the context of several limitations. First, this review should not be considered an exhaustive analysis. In the present study, we reviewed only English-language websites. Thus, the quality of websites available in other languages remains unknown. Given that much of the world is non-English-speaking, we encourage researchers with fluency in other languages to conduct similar reviews of non-English websites. Second, comparisons between the 2004 and current review are made with a notable caveat. The search procedure used in the 2004 paper was comprehensive, including all websites retrieved in the first 10 pages of search returns. The current review used a search strategy that was designed to mimic the search pattern of most Internet users. That is, we included only those websites retrieved on the first page of search returns. It is possible that higher quality websites are more likely to be retrieved in the first page of search returns. If true, the increases in quality observed between the 2004 and current dataset may be an artifact of the search procedures rather than a reflection of a real improvement in the quality of smoking cessation websites. In a review of popular smoking cessation websites identified by survey respondents, Etter [7
] concluded that users had difficulty finding the highest quality websites: the three highest rated websites in that review attracted only 7% of visitors. The sites that were most commonly used (ie, the most popular) were not the highest quality. Third, the present study was designed to address the quality of content presented in the websites. Elements of usability such as navigation, layout, and accessibility are also important to a user’s experience and likely play an important role in the effectiveness of a behavior change website. These elements should be examined in future studies.
Results of this study indicate that the content and quality of information contained in smoking cessation websites may be improving. However, more often that not, smokers looking for assistance online will find websites that do not provide evidence-based guidance and assistance. Moreover, numerous questions remain to be answered about WATIs. Research needs to move beyond quantitative assessments of the amount and accuracy of information provided via the Web and begin to examine the qualitative nature of WATI sites and the relationships that exist between these sites and their users. For example, research is needed to determine whether there are intrapersonal (eg, age, education, health literacy, need/preference for social support) or environmental characteristics (eg, support/incentives from a health maintenance organization or employer, presence of smoking policy restrictions) that predict better or worse outcomes when using Internet-delivered cessation interventions. The Internet holds great potential to reach millions of smokers who may not otherwise seek cessation treatment. Efforts are needed to ensure that the content of Internet interventions is sound so that we can begin to understand how, for whom, and by what mechanism(s) WATIs may be effective.