NY TCP's antismoking media campaign resulted in significant increases in smokers' exposure to paid advertisements. Concurrent with these increases, both intentions to quit and cessation attempts increased significantly among smokers from 2003 through 2009, and current smoking prevalence significantly declined, consistent with similar findings in many other state campaign evaluations (11
). On the basis of New York State adult population estimates in 2003 and 2009, these results suggest that there were approximately 502,000 fewer current smokers and 204,000 more smokers trying to quit in 2009 than in 2003.
Although many outcomes changed favorably during this time, many of these changes appeared to plateau during 2008 and 2009. This plateau in outcome change corresponded with declines in advertising activity in 2008 and the latter half of 2009. Positive changes in cessation-related outcomes between 2007 and 2009 were minimal compared to prior yearly changes from 2003 to 2006. These patterns suggest that fluctuations in advertising efforts affect both individual-level exposure to the campaign and trends in key outcome indicators, underscoring that consistent implementation is necessary for the long-term success of health communication campaigns.
Several lessons can be drawn from NY TCP's implementation of health communication interventions that may be instructive for other states. First, increases in media expenditures and placements directly translate into increases in individual awareness of advertisements. In addition, heavy resources are required to attain measurable levels of exposure to campaign messages. Since 2003, NY TCP's budget for health communication increased dramatically, which translated into significant increases in smokers' awareness of the campaign. Maintaining high levels of exposure is necessary to effect change (25
) but is a challenge because of uncertainties about the availability of funds for paid advertising.
A second lesson is that NY TCP's media campaign used off-the-shelf advertisements that are evidence-based and readily available from CDC's Media Campaign Resource Center and other outlets. This choice allowed the campaign to avoid costly formative and creative content development and devote more resources to media placement and advertising purchases. The net result was that NY TCP was able to maximize smokers' awareness of the campaign and influence key outcomes with existing media resources.
This study had several limitations. First, NY TCP's media campaign occurred in the context of other broader program efforts including community partnerships, policy promotion, and cessation services such as the New York State Smokers' Quitline and cessation centers. In addition, antismoking policies changed during this period and included a dramatic increase in cigarette excise taxes in 2009, which may have helped to offset the effect of reductions in campaign GRPs that occurred between 2007 and 2009. These factors can make it difficult to identify the effects of any single program area, including media. A second limitation is that our data are limited to New York adult smokers and thus may not be generalized to the national population of smokers.
However, 2 aspects of the media campaign highlight its plausibility as a driver in the observed outcome changes. First, the NY TCP media campaign has consistently been one of the largest single interventions in the program, representing more than 25% of the overall NY TCP budget over the life of the program. Second, the media campaign also likely represents the program's most influential intervention in terms of the overall number of smokers reached, because of the wide distribution and visibility of television advertisements. To confirm the influence of the campaign, future studies using multivariate models to examine the direct relationship between exogenous market-level campaign GRPs and individual-level outcome indicators are needed. These future studies should undertake analyses to identify the specific effects of the media campaign, adjusting for other potentially confounding influences.