The influence of television advertisements on promoting calls to telephone quitlines

Abstract

The aim of the study was to assess the relative effectiveness of cessation, secondhand smoke and other tobacco control television advertisements in promoting quitlines in nine states from 2002 through 2005. Quarterly, the number of individuals who used quitlines per 10 000 adult smokers in a media market are measured. Negative binomial regression analysis was used to link caller rates to market-level exposure to tobacco control television advertisements overall and by message theme. The relationship between caller rates and advertising exposure was positive and statistically significant (P < 0.001). Advertisements that focus on promoting cessation (P < 0.001), highlighting the dangers of secondhand smoke (P = 0.037), and all other tobacco countermarketing advertisements (P = 0.027) were significantly associated with quitline caller rates. For every 10% increase in exposure to cessation, secondhand smoke and other tobacco countermarketing advertisements, caller rates increased by 1.1, 0.2 and 0.4%, respectively. Caller rates significantly increased in quarters when cigarette excise tax increased (P < 0.001) and when the percentage of the population covered by comprehensive smoke-free air laws increased (P = 0.022). Although advertisements promoting cessation are the most effective in driving quitline use, other topics, such as messages highlighting the dangers of secondhand smoke, also prompt their quitlines.

Introduction

In 2007, the Institute of Medicine (IOM) presented a blueprint for action to ‘reduce smoking so substantially that it is no longer a public health problem for our nation’ [1, p.1]. IOM called for a two-pronged strategy to include (i) strengthening and fully implementing traditional evidence-based tobacco control measures and (ii) changing the regulatory landscape to permit federal regulation of tobacco product content, design, marketing and distribution. Traditional evidence-based interventions that have had the most success include price increases, smoke-free policies that prohibit smoking in all indoor areas of workplaces and public places, aggressive media campaigns, access to cessation services, comprehensive advertising bans and graphic warning labels on tobacco products [2–6]. Along with price increases and smoke-free policies, interventions that have been shown to be effective in increasing cessation include smoking cessation telephone help lines (quitlines), reduced out-of-pocket costs for cessation therapies and mass media campaigns [1, 3–7].

Quitlines typically provide a range of services, including counseling, taped messages providing tips for smoking cessation, self-help materials and other cessation information [8]. Quitlines commonly have fax referral systems that allow health care providers to arrange for a quitline specialist to call a patient, and many quitlines provide nicotine replacement therapy (NRT) to callers for free on a limited basis or at a reduced price. The Task Force for Community Preventive Services systematic review found that quitlines are effective in promoting smoking cessation [3, 6].

A number of studies have also shown that the number of people who call quitlines is very responsive to mass media promotional efforts [4]. Specifically, increases in exposure to television advertisements are associated with increases in calls to quitlines [9–15]. This is consistent with the Task Force for Community Preventive Services review and a more recent review showing that mass media campaigns are effective in reducing smoking [4]. Research also indicates that publicizing the availability of NRT in the news media can increase quitline call volume [16] and awareness of cessation services [17]. Although studies have shown that media is effective at promoting quitline use, the relationship between tobacco control television advertising characteristics and adult cessation is unclear [9, 12, 18, 19]. A review of existing literature related to mass media campaigns and smoking cessation promotion concluded that policies should support consistent exposure to anti-tobacco media with a negative health effects focus [20]. In this study, we examine how television advertisements focusing on cessation, secondhand smoke and other themes influence quitline caller rates, controlling for the influence of cigarette excise taxes and smoke-free air laws using data from nine states between 2002 and 2005.

Materials and methods

The main outcome measure is the quitline caller rate, defined as the number of individuals who used quitline services per 10 000 adult smokers in the media market. The Centers for Disease Control and Prevention’s Office on Smoking and Health worked with state quitline providers to obtain caller data for nine state quitlines. The quitlines were chosen to represent a range of quitline services (e.g. provision of NRT), geographic location, and level of media promotion. The quitline caller data set consists of a single record for each caller who received any services from the quitline. Quitline caller rates are measured quarterly from the first quarter of 2002 through the fourth quarter of 2005 for 38 designated market areas located in the selected nine states. Quitline caller data were not available for all nine states from 2002 through 2005. Table I presents the state quitlines included in our analysis along with their associated media markets, market-level descriptive statistics and data availability.

Table I.

Summary of quitline states, associated media markets, and descriptive statistics

			Distribution of TARPs by objective
Quitline state	Nielsen media markets	Average population-weighted quarterly TARPs (Min. to Max.)	Cess. (%)	SHS (%)	Oth. (%)	N (n)
Arizona	Phoenix (Prescott)	500 (27–1521)	84	2	14	16 (1)
	Tucson (Sierra Vista)	171 (0–779)	100	0	0	16 (0)
Georgia	Albany	108 (0–634)	82	18	0	16 (0)
	Atlanta	186 (0–693)	49	51	0	16 (0)
	Augusta	82 (0–383)	89	11	0	16 (0)
	Columbus	271 (0–774)	76	24	0	16 (0)
	Macon	289 (0–691)	87	13	0	16 (0)
	Savannah	196 (0–722)	66	34	0	16 (0)
Maine	Bangor	121 (0–950)	86	0	14	16 (0)
	Portland–Auburn	215 (0–1711)	89	0	11	16 (1)
	Presque Isle	211 (0–2403)	85	0	15	16 (1)
Minnesota	Duluth–Superior	1412 (198–3803)	63	35	2	16 (7)
	Mankato	273 (12–496)	84	12	4	16 (0)
	Minneapolis-St. Paul	862 (75–1624)	75	15	10	16 (4)
New York	Albany-Schenectady-Troy	1593 (955–2562)	63	34	3	4 (3)
	Binghamton	1229 (990–1365)	52	48	0	4 (3)
	Buffalo	1398 (571–2527)	71	22	7	4 (2)
	Elmira (Corning)	679 (367–884)	54	46	0	4 (0)
	New York	545 (238–1103)	75	25	0	4 (0)
	Rochester	1218 (505–2177)	71	24	5	4 (2)
	Syracuse	1445 (1338–1507)	45	48	7	4 (4)
	Utica	850 (539–1284)	34	65	0	4 (1)
	Watertown	804 (577–1175)	50	44	6	4 (0)
Ohio	Cleveland-Akron (Canton)	1110 (316–1566)	38	62	0	9 (6)
	Columbus	1278 (427–1808)	50	50	0	9 (5)
	Dayton	1121 (340–2033)	58	42	0	9 (5)
	Lima	991 (405–1792)	44	56	0	9 (3)
	Toledo	1016 (259–1788)	51	49	0	9 (2)
	Zanesville	751 (217–1439)	55	45	0	9 (1)
Pennsylvania	Harrisburg-Lancaster-Lebanon-York	328 (0–1239)	83	1	16	14 (1)
	Johnstown-Altoona	307 (0–1224)	80	0	20	14 (1)
	Philadelphia	396 (0–1322)	80	1	19	14 (1)
	Pittsburgh	440 (0–1427)	72	0	28	14 (2)
	Wilkes Barre-Scranton	357 (0–1326)	78	1	22	14 (1)
Utah	Salt Lake City	2619 (351–4737)	81	19	0	16 (13)
Washington	Seattle–Tacoma	820 (2–1572)	88	12	0	16 (3)
	Spokane	1186 (0–2539)	84	14	1	16 (7)
	Yakima–Pasco–Richland-Kennewick	381 (1–749)	86	14	0	16 (0)
	Yakima–Pasco–Richland-Kennewick	381 (1–749)	84	2	14	16 (1)

Min., minimum; Max., maximum; Cess., cessation; SHS, secondhand smoke; Oth., other; N, number of observations (quarters); n, number of quarters where TARPs met or exceeded 1200.

Although media markets do cross state borders, most are predominantly or completely contained within a single state. We refer to this as the market’s primary state. When a media market crossed state boundaries, we adjusted the caller rates to reflect the population of smokers within the media market in the primary state (because callers outside the state are not eligible to use the state’s quitline).

The principal independent variable is quarterly television target audience rating points (TARPs) for adults aged 18 or older. TARPs are defined as the product of the percentage of the audience exposed to a commercial (i.e. audience reach) and the frequency of exposure (i.e. frequency). If 50% of the population was exposed to a commercial three times in a quarter, TARPs would equal 150 (50 × 3). CDC recommends that campaigns should have an average of 1200 TARPs per quarter during the introduction of a campaign and 800 TARPs thereafter [2]. Of note, 18% of our data meet or exceed this recommended initial level (Table I). Television media data were licensed from Nielsen Media Research. Media markets consist of a collection of zip codes centered on a major city or metropolitan area. As stated previously, most of the media markets are located primarily within a single state. The Nielsen data contain an indicator for the year and quarter in which an advertisement aired, the sponsor of the advertisement (e.g. New York State Department of Health), a short description of the advertisement, and the TARPs for the advertisement.

Television advertisements were reviewed and coded using a training guide for tobacco control television advertisements developed by the authors. Because advertisement descriptions were very brief (e.g. ‘BOY/SOCCER/PACK CRUSHED BY SNEAKER’), Nielsen Media Research also provided digital media files for each of the advertisements. Using the media files, advertisements were coded for target audience and topic, which was used to define message objective. Target audiences include adults only, youth only, or both. Advertisements were also characterized by their objective—promoting cessation or highlighting the health consequences of smoking, highlighting the dangers of secondhand smoke, supporting clean indoor air laws, industry manipulation, smoking during pregnancy, prevention or other.

Advertisements were reviewed and coded by at least two coders in multiple rounds to allow for resolution of differing characterizations across coders. Before reconciliation, Cohen’s Kappa scores ranged from 0.71 to 1.000 with a mean score of 0.81. Because quitlines only provide services to adults, media data were limited to advertisements targeting adults aged 18 or older. Media data were also limited to advertisements with confirmed tobacco countermarking relevance.

In addition to a measure of total TARPs, we created three theme-specific TARP measures relating to (i) cessation, (ii) dangers of secondhand smoke exposure and (iii) all other tobacco-related topics. The cessation measure includes TARPs for advertisements characterized as addressing the health consequences of smoking. A small number of advertisements contained both cessation and secondhand smoke messages. For this analysis, we included those TARPs in the cessation measure. The secondhand smoke measure includes TARPs for advertisements that highlight the dangers of secondhand smoke. The other measure includes TARPs for advertisements for a diverse set of themes, such as youth prevention, industry manipulation, support for clean indoor air laws and the risks of smoking while pregnant.

Quitline callers were assigned to a Nielsen media market based on their zip code of residence. The total number of callers in each media market was then matched with media market-level TARPs.

Cigarette excise tax rates come from the CDC State Tobacco Activities Tracking and Evaluation system. Data on smoke-free air laws come from the Americans for Nonsmokers’ Rights [21], which identifies enforcement dates for laws and/or ordinances imposing 100% smoke-free workplaces, restaurants or freestanding-bars across all U.S. municipalities (e.g. cities, counties and states). We combined data on smoke-free air laws and ordinances with annual Census population estimates to construct quarterly measures of the percentage of the media market population covered by all three types of smoke-free air laws (i.e. workplace, restaurant and free-standing bars). We created event indicators for increases in cigarettes excise taxes and 100% smoke-free air law coverage. Specifically, within each state and by quarter, an event indicator is equal to one when the cigarette excise tax increased from the previous quarter and zero otherwise. Similarly, an event indicator for smoke-free air laws equaled one when the percentage of the population in the market’s primary state covered by 100% smoke-free air laws increased by more than 1%.

Analysis

We began our analysis of the relationship between tobacco countermarketing television TARPs and quitline caller rates by looking at simple trends. First, we examined trends in total annual TARPs by state to see whether there were similar patterns across states. Next, we compared trends in average quarterly TARPs for all 38 media markets to average quitline caller rates to examine the simple correlation between television TARPs and quitline caller rates.

We then examined the relationship between overall exposure to tobacco countermarketing television advertisements by conducting a negative binomial regression modeling quarterly media market–level quitline callers on total TARPs using the number of adult smokers in the media market as the model exposure variable. We then looked at how the content of television advertisements influences quitline caller rates by conducting a negative binomial regression regressing quitline caller rates on the three separate thematic TARP measures: cessation, secondhand smoke and other. Each model also controls for cigarette excise tax increase events, 100% smoke-free air law coverage increase events, media market, year and quarter.

The model coefficient for logged TARPs in our negative binomial regressions is the TARPs elasticity of caller rates. We compared elasticities to test the relative effectiveness of cessation, secondhand smoke, and other tobacco countermarketing TARPs. Elasticities represent the percentage change in the outcome variable for a given percentage change in the independent variable. For example, if the elasticity for cessation TARPs from the analyses described earlier is 0.4, this implies that increasing cessation TARPs by 10% would lead to a 4% increase in caller rates.

Results

Descriptive analyses

The results of our simple trend analysis show that total annual TARPs varied greatly across states (Fig. 1). Moreover, changes over time are not consistent across states (e.g. Ohio’s highest levels are in 2005, whereas Minnesota’s, Utah’s, and Washington’s occur in 2003). Figure 2 presents trends in average quarterly TARPs and quitline caller rates. The correlation coefficient between average total quarterly TARPs and average quitline caller rates is 0.30 (P < 0.001).

Fig. 1.

Annual population-weighted television TARPs by quitline state, 2002–5. Estimates were obtained by first calculating quarterly state-level average TARPs across media markets within the state using market population as a weight. Annual state-level TARPs were then estimated by summing the market-population weighted average quarterly TARPs.

Fig. 2.

Average population-weighted quarterly television TARPs and quitline caller rates.

Multivariate regressions

The results show that increases in total exposure to television tobacco countermarketing advertisements are associated with increases in quitline caller rates (P < 0.001) (Table II). The elasticity implies that a 100% increase in total tobacco countermarketing television TARPs would lead to an 11% increase in quitline caller rates. Cigarette excise tax increase events (P < 0.001) and 100% smoke-free air law coverage increase events (P = 0.022) are also associated with an increase in quitline caller rates. The results from the second model show that cessation (P < 0.001), secondhand smoke (P = 0.037) and other tobacco countermarketing advertisements (P = 0.027) were all positively associated with quitline caller rates, adjusted for the effect of cigarette excise tax increase events and 100% smoke-free air law coverage increase events (Table III). The elasticities suggest that cessation TARPs (0.11) were relatively more effective than secondhand smoke (0.02) or other tobacco countermarketing TARPs (0.04). Comparing elasticities indicates that a 100% increase in cessation TARPs would lead to an 11% increase in caller rates, whereas the same size increase in secondhand smoke and other tobacco countermarketing TARPs would only lead to a 2 and a 4% increase in caller rates, respectively.

Table II.

Negative binomial regression of quitline call rates on total television TARPs

Quitline caller rate	Regression coefficient (P value) [95% CI]	Mean quarterly value
Logged TARPs	0.12 (0.000) [0.10, 0.14]	4.82
Indicator for the quarter in which an increase in cigarette excise taxes occurred	0.47 (0.000) [0.29, 0.65]	0.08
Indicator for the quarter in which an increase in statewide smoke-free air law coverage occurred	0.27 (0.022) [0.04, 0.50]	0.03
Observations	448

Table III.

Negative binomial regression of quitline caller rates on television TARPs by advertisement objective

Quitline call rate	Regression coefficient (P value) [95% CI]	Mean quarterly value
Logged cessation TARPs	0.11 (0.000) [0.09, 0.13]	4.23
Logged secondhand smoke TARPs	0.02 (0.037) [0.00, 0.05]	1.86
Logged other tobacco countermarketing TARPs	0.04 (0.027) [0.00, 0.07]	0.65
Indicator for the quarter in which an increase in cigarette excise taxes occurred	0.49 (0.000) [0.32, 0.66]	0.08
Indicator for the quarter in which an increase in statewide smoke-free air law coverage occurred	0.27 (0.016) [0.05, 0.48]	0.02
Observations	448

Discussion

Consistent with previous research and documented effectiveness of mass media efforts to promote quitline use, this study provides strong evidence that television advertisements are effective. Although cessation advertisements have the highest relative effectiveness on increasing quitline caller rates, advertisements highlighting the dangers of secondhand smoke or other issues in tobacco control can also drive smokers to use quitlines. These results can help program planners to better understand how different types of media influence quitline caller rates. Over the study period, increases in quitline caller rates were associated with quarters when cigarette excise taxes increased and when statewide smoke-free air law coverage increased. This finding is consistent with other studies. For example, in 2003, when New York City’s smoke-free air law was implemented, free NRT was offered to smokers who called the quitline and call volume increased dramatically [22]. Similarly, in 2008, Wisconsin increased the state cigarette excise tax rate by $1.00 (from $0.77 to $1.77 per pack) and preceded this price increase with efforts to promote the state’s quitline service and offers of free NRT [23]. As a result of these coordinated efforts, the demand for state quitline services increased nearly threefold (from 10 000 to 27 000) from previous years.

Although the findings are encouraging, the study has several limitations. Media exposure, as measured by TARPs, may not have captured all of the tobacco countermarketing television advertisements. For example, the current analyses did not measure other quitline promotions, such as radio and Internet advertisements and efforts to gain ‘earned’ media (e.g., press releases). Unfortunately, we are not aware of any system that exists across all of the states in the study to capture this information. However, the indicators of the periods when cigarette taxes and smoke-free air law coverage increased may also capture any related publicity related to these policy changes. Advertising in one media market likely spills over into another media market. As a result, our findings may have underestimated the strength of the relationship between exposure to tobacco countermarketing media and quitline caller rates. We could not confirm for all advertisements that the quitline phone number was included. When we were able to obtain the specific advertisements, nearly all had the number included. Additionally, our measures of cigarette excise tax rates and 100% smoke-free air law coverage were based on the primary state where each media market is located; however, many media markets cross state borders. Finally, this study cannot say whether increased callers to quitlines lead to more long-term behavioral changes, such as sustained cessation. It is possible that increased levels of advertising provoke more smokers to call a quitline, but it is also possible that many of these smokers are not prepared to quit.

Funding for state tobacco control programs on average continues to be far below the CDC recommended funding levels [24]. The importance of adequate funding is most evident in the effective implementation of countermarketing media campaigns [24]. Evidence shows that to be effective, these types of campaigns must have sufficient reach, frequency and duration [25, 26].

Funding

Support for this research was provided by the Office on Smoking and Health, Centers for Disease Control and Prevention.

Conflict of interest statement

None declared.

Acknowledgements

The findings and conclusions are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. The responsibility for all of the presented material rests exclusively with the authors.