- Email: [email protected]
Connecting low-income smokers to tobacco treatment services
Addictive Behaviors 52 (2016) 108–114
Contents lists available at ScienceDirect
Addictive Behaviors
Connecting low-income smokers to tobacco treatment services Jonathan S. Slater a,b, Christina L. Nelson a, Michael J. Parks a,⁎, Jon O. Ebbert c a b c
Minnesota Department of Health, 85 East 7th Place, Saint Paul, MN 55164, USA Masonic Cancer Center, 424 Harvard St SE, University of Minnesota, Minneapolis, MN, USA Division of Primary Care Internal Medicine, 221 Fourth Avenue SW, Mayo Clinic, Rochester, MN 55905, USA
H I G H L I G H T S • • • • •
We designed two strategies for connecting low-income smokers to quitline services. Strategies were direct mail and opportunistic referral; both employed incentives. Each strategy reached individuals at differential levels of readiness to quit. Smoking abstinence rates at follow-up indicated both strategies had high impact. Both strategies had strengths, and both can be used in population-based practice.
a r t i c l e
i n f o
Article history: Received 1 June 2015 Received in revised form 2 October 2015 Accepted 11 October 2015 Available online 18 October 2015 Keywords: Population-based programs Smoking cessation Underserved populations Telehealth Financial incentive
a b s t r a c t The Affordable Care Act calls for using population-level incentive-based interventions, and cigarette smoking is one of the most significant health behaviors driving costs and adverse health in low-income populations. Telehealth offers an opportunity to facilitate delivery of evidence-based smoking cessation services as well as incentive-based interventions to low-income populations. However, research is needed on effective strategies for linking smokers to services, how to couple financial incentives with telehealth, and on how to scale this to population-level practice. The current paper evaluates primary implementation and follow-up results of two strategies for connecting low-income, predominantly female smokers to a telephone tobacco quitline (QL). The population-based program consisted of participant-initiated phone contact and two recruitment strategies: (1) direct mail (DM) and (2) opportunistic telephone referrals with connection (ORC). Both strategies offered financial incentives for being connected to the QL, and all QL connections were made by trained patient navigators through a central call center. QL connections occurred for 97% of DM callers (N = 870) and 33% of ORC callers (N = 4550). Self-reported continuous smoking abstinence (i.e., 30 smoke-free days at seven-month followup) was 20% for the DM group and 16% for ORC. These differences between intervention groups remained in ordered logistic regressions adjusting for smoking history and demographic characteristics. Each recruitment strategy had distinct advantages; both successfully connected low-income smokers to cessation services and encouraged quit attempts and continuous smoking abstinence. Future research and population-based programs can utilize financial incentives and both recruitment strategies, building on their relative strengths. Published by Elsevier Ltd.
1. Introduction Smoking prevalence among U.S. adults is 17% for those who live at or above the poverty level but 28% for those below it (Centers for Disease Control and Prevention, 2014), underscoring how income underlies disparities in cigarette smoking and smoking-related health problems (Jha et al., 2006; Thomas et al., 2008). High smoking rates persist among low-income women (Stewart et al., 2010), and smoking can account ⁎ Corresponding author at: Minnesota Department of Health, 85 East 7th Street, Saint Paul, MN 55164, USA. E-mail address: [email protected] (M.J. Parks).
http://dx.doi.org/10.1016/j.addbeh.2015.10.013 0306-4603/Published by Elsevier Ltd.
for up to half of mortality disparities associated with socioeconomic status among males (Jha et al., 2006). Financial incentives are known to successfully promote smoking cessation, especially in low-income populations (Blumenthal et al., 2013; Bryant et al., 2011; Volpp et al., 2009; Sigmon & Patrick, 2012), yet a dearth of research exists on how such evidence-based interventions translate to population-level practice (Ammerman, Smith, & Calancie, 2014; Spoth et al., 2013; Lewis, 2010). Subsequently, a more translatable evidence base consisting of practice-based evaluations and not necessarily randomized trials of population-level interventions is needed (Green, 2008; Sanson-Fisher, Bonevski, Green & D'Este, 2007; Ammerman et al., 2014). Such evidence is critical for addressing public health priorities exemplified by the
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
Affordable Care Act such as scaling up incentive-based interventions in order to promote smoking cessation in low-income populations (Blumenthal et al., 2013: 497–498; Kassler, Tomoyasu, & Conway, 2015). Telehealth has potential for effectively delivering cessation services to large numbers in underserved populations (Bashshur et al., 2014; Wootton et al., 2005; Wootton, 2012). As a primary example, free state telephone tobacco quitlines (QLs) offer an evidence-based and population-level approach to increasing smoking abstinence rates (Stead et al., 2007; Fiore et al., 2008), and low-income and non-White populations are inclined to use free QLs (Burns et al., 2011; Zhu et al., 2011). However, QL utilization rates are markedly low across the U.S., with a state-level average of approximately 2% (Zhu et al., 2012). Consequently, more research is needed on strategies that successfully extend the reach of QLs, particularly to low-income populations (Zhu et al., 2012). Specifically, QL research is needed that (1) incorporates financial incentives, (2) targets individuals at various stages of motivation for quitting, and (3) focuses on “reactive” recruitment rather than “proactive” approaches (Stead et al., 2013; Asfar et al., 2011; Mathew et al., 2014). Few published QL interventions have utilized financial incentives (Stead et al., 2013), and more interventions need to reach adults not motivated to quit (Asfar et al., 2011). Although proactive telephone-based cessation interventions (i.e., calls initiated by counselors [Lichtenstein et al., 1996]) have been effective in low-income populations (e.g., Solomon et al., 2005), less is known about reactive strategies (i.e., cessation counseling provided on demand). Proactive QL recruitment is designed to contact potential participants directly with QL operators in order to connect smokers directly to the QL at time of initial contact. Proactive strategies may potentially discourage participation of individuals who are not ready to take immediate and direct behavioral steps towards quitting. A reactive strategy is designed to refer potential users to a QL, with the expectation that individuals who are willing to participate will contact the QL at an appropriate time for them after receipt of referral. Reactive strategies are important because they have potential to recruit those who are ready to take action-oriented steps towards quitting, and it gives individuals time to contemplate possible steps towards quitting without immediate pressure. Reactive strategies can also encourage indirect pathways to behavior change via naturally occurring psychosocial mechanisms within the environment that individuals receive their initial referral to the QL (see e.g., Parks et al., 2015).
1.1. The current study We report primary implementation and follow-up results from a population-based program that utilized financial incentives and two strategies designed to connect low-income smokers to Minnesota's QL, among a low-income sample primarily comprised of females. Mirroring patterns across the U.S., QL reach and utilization rates are low in Minnesota (see Patten et al., 2011). Following seminal research in implementation science (e.g., Fixsen et al., 2005; Glasgow, Vogt, & Boles, 1999), we examine participant responsiveness and retention as well as program fidelity measured via response rates and QL connection rates. We also examine primary follow-up and effectiveness outcomes measured as smoking status at time of follow-up with a focus on smoking abstinence rates. We examine these outcomes for two recruitment strategies separately, providing a basis for comparative effectiveness. Since direct mail is a cost-effective, population-level strategy for connecting individuals to preventive and telehealth services (Slater et al., 2005; Soet & Basch, 1997), one recruitment strategy was direct mail. The second was a centralized patient navigation system. Both strategies relied on individual-initiated phone contact (see Soet & Basch, 1997) and offered financial incentives to low-income smokers for being connected to the QL via three-way phone calls conducted by trained patient navigators (see Methods section for details).
109
2. Methods 2.1. Participants and intervention 2.1.1. Overview and setting From September 2010 to September 2012, the program was implemented through “Sage Programs”: Sage, Minnesota's National Breast and Cervical Cancer Early Detection Program (NBCCEDP) and Sage Scopes, Minnesota's Colorectal Cancer Control Program (CRCCPs) at the Minnesota Department of Health (see Lee et al., 2014; Slater et al., 2005). The recruitment timeframe was contingent on funding; consequently, funding and recruitment ended concurrently. Sage provides free breast and cervical cancer screening services to inadequately insured women 40 years of age or older, with household incomes at or below 250% of the US federal poverty level. Sage Scopes provides free colorectal cancer screening to a much more limited number of men and women ages 50 and older but who otherwise meet the same eligibility criteria as Sage. Unique among NBCCEDPs and CRCCPs, Sage Programs (hereafter referred to as Sage) has a single call center staffed by “patient navigators” (see Freund et al., 2008) trained in motivational interviewing (see Rollnick, Miller, & Butler, 2008). More than 22% of Sage participants smoke cigarettes. Since NBCCEDPs target lowincome females, as previously noted, the current program focuses on a sample that is disproportionately female. 2.1.2. Intervention The current program offered a $20 incentive to callers for being connected to Minnesota's QL via a three-way call conducted by Sage patient navigators. Two recruitment strategies were used: (1) direct mail (DM) and (2) opportunistic referral with connection (ORC) through the Sage Call Center call transfer system. DM was designed to prompt cigarette smokers to call Sage's toll-free phone number rather than serve an educational function. Mailers consisted of a folded card with emotionally evocative messages and graphics as well as a small insert card advertising the financial incentive offer. We employed a loss-framed message (see Rothman & Salovey, 1997) coupled with a high-efficacy message (see Witte & Allen, 2000). These DM designs are based on health communication and health behavior theory. Specifically, the fear appeal message or loss-frame message is designed to inform an individual that a certain behavior will lead to an undesirable outcome, such as long-term smoking leading to physical disability or mortality (see Rothman and Salovey, 1997). Health communication research shows that these health messages based on fear appeal coupled with a clear articulation of achievable behavioral steps (i.e., high-efficacy message) produce the greatest behavior change (Rothman and Salovey, 1997; Witte and Allen, 2000; Slater et al., 2005). Theory suggests that responses to such health messages can be either a “danger control action” or a “defensive response.” Danger control actions are actions taken when individuals feel (1) susceptible to a health problem and (2) capable of completing behavioral steps necessary for reducing the risk of the presented health problem. Defensive responses are actions antithetical to the proposed behavior change (i.e., behavioral steps away from the protective response) that are based on feelings of high fear in addition to low self-efficacy (for a more thorough discussion see Witte et al., 2001). The purpose of the design was to promote risk susceptibility associated with cigarette smoking in addition to offering clear and achievable steps for action. An example of a mailer is presented in Fig. 1. Following past research, two rounds of mailings were employed (Slater et al., 2005). The incentive offer was presented via a small inserted card affixed to the inside of the mailer. This inserted card read: “Special offer: Call today and we'll pay you $20 plus give you the free tools to quit smoking,” and it included Sage's toll-free phone number with a tracking promotion code that patient navigators recorded during calls. The presentation of the incentive offer within the mailer was intended to influence the decisional balance by reducing perceived barriers, providing
110
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
and no further information was gathered if they refused to be connected. QL services included free smoking cessation telephone counseling with a maximum of five sessions within a six-month period. If desired, the QL provided self-help materials and free nicotine replacement therapy, and QL staff scheduled follow-up calls with patients. It is also important to highlight that the financial incentive was only offered to individuals who completed three-way QL connections and who did not disconnect prior to speaking with QL operators. The incentive was not linked to the extent to which they used available services. This was done for practical reasons (utilization data could not be directly obtained from the QL staff), but this has precedence in consumer marketing. It is analogous to incenting people via a coupon based on the expectation that this exposure is sufficient for some people to become customers (see e.g., Rothschild & Gaidis, 1981). The incentive was a $20 gift card, similar to a credit card that is almost universally accepted for goods and services. Once connection to the QL was confirmed, patient navigators used the participants' information to send gift cards via mail. Participants who completed the QL connection were interviewed via telephone by trained staff at least seven months after the initial QL connection. The average time between initial QL transfer and follow-up surveys was 8.8 months (SD = 2.84).1 There was no contact between Sage patient navigators and participants, or between surveyors and participants, during the timeframe after initial QL connection and before follow-up surveys because QL operators were responsible for cessation-oriented contact after three-way transfers from the Call Center. Surveys and survey timing were modeled after North American Quitline Consortium (NAQC) recommendations (see North American Quitline Consortium, 2011). Ten call attempts were made before participants were deemed unresponsive, and these contacts were made at various times and on various days of the week. Fig. 1. Direct mail mailer example with loss-frame and high-efficacy message.
cues to action, and making pros of behavior change more salient (see Becker, 1974; Prochaska et al., 2008; Weinstein et al., 2008). Recruitment to ORC consisted of Sage's patient navigators obtaining the smoking status of all individuals who called the Sage Call Center for breast, cervical, or colorectal cancer screening information or appointments and opportunistically presenting the QL referral offer (i.e., a $20 incentive for being connected to the QL) to self-reported smokers. Patient navigators handled phone calls for both intervention groups (six full-time navigators were involved). Navigators followed a script designed to: (1) identify whether the caller was responding to a mailing, (2) record the promotion code printed on the mailer and financial incentive card, (3) determine callers' decision about participation, and (4) make QL connections for willing participants. For willing participants, patient navigators put participants on hold and called QL agents. Patient navigators then confirmed with QL operators that a Sage patient would be connected to the QL via a three-way call. Once confirmed, patient navigators conferenced the Sage participant back into the call and remained on the line until communication between participants and QL operators was established. Thus, QL connection was defined as a confirmed connection between participants and QL operators via three-way calls conducted by patient navigators. Once participants were connected to the QL, trained QL operators were responsible for handling participants and administering cessation-related services. All individuals who agreed to be connected to the QL were deemed as “enrolled” in the program. Approximately 45% of callers enrolled in the program, and there were major differences in enrollment rates between recruitment strategies (DM = 97% enrollment vs. ORC = 35% enrollment). For enrolled individuals, demographic information was gathered at follow-up and promotion codes were recorded prior to QL connections. Callers were considered nonparticipants
2.1.3. Participants and sample size The DM intervention targeted 9922 low-income smokers who had previously been screened for cancer through Sage. Smoking status was determined by two means: (1) a prior record of being a smoker obtained during previous contact with Sage, or (2) Sage enrollment forms (e.g., from local health clinics) that identified the individual as a smoker. The ORC intervention was provided to smokers from among the approximately 23,000 calls fielded by Sage patient navigators through the Call Center during the intervention period. A total of 870 smokers responded to the DM intervention, and the ORC strategy was administered to 4550 self-identified smokers. After being presented with the script by patient navigators, 2456 smokers were connected to the QL: 844 from the DM group and 1612 from the ORC group. Because they are Sage's target population, the sample was predominantly (98%) female (1153 females and 26 males) across both interventions. The analytic sample did not exclude males for multiple reasons. Even though the primary target of the program was lowincome females, dramatic mortality disparities associated with smoking and income exist among males. The males in the sample were also found to be heavy smokers,2 and excluding males from the analysis had potential to bias program effects in terms of increasing quit rates and connection rates. The analytic sample is based on participants who were reached with the survey at time of follow-up after QL connections.3 The final analytic sample was comprised of 418 DM and 1 Supplementary analyses demonstrated that survey timing was not systematically related to smoking cessation outcomes or participant characteristics. 2 There were no differences between males and females in regards to demographic and smoking history variables, except for years smoked. One-way ANOVA and t-test with unequal variance showed males' average years smoked (36) was greater than females' years smoked (31) (p b .05). 3 Descriptive information was not gathered for the majority of participants who made QL connections, and therefore it was not possible to compare individuals who participated in the follow-up survey versus those who did not complete the survey.
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
761 ORC participants who had complete data; approximately 3% and 4% (DM and ORC, respectively) of the surveyed participants were removed because of incomplete data. For both intervention groups, no significant differences were found between the analytic sample and individuals eliminated because of missing data in terms of tobacco cessation and background characteristics. We use the total sample to examine implementation results, and summative results strictly rely on the final analytic sample. For implementation results we consider “enrollment” rates and response rates prior to follow-up surveys. We calculate enrollment rates based on connection rates subsequent to contact with patient navigators because patient navigation was a core component of the program and non-callers in the DM group did not receive this component. There is an implementation outcome of participant responsiveness that is unique to the DM group (i.e., call rate in response to the mailing) consisting of the response rate to the DM campaign measured as the total number of DM callers that called Sage (prior to QL connections) relative to the total number of DM individuals who were sent mailers. All summative results assessing quit status utilize follow-up survey results comprised of the final analytic sample of 418 DM and 761 ORC participants. 2.2. Outcomes 2.2.1. Implementation outcomes As noted, we examine primary implementation outcomes (see Durlak & DuPre, 2008) that assess participant responsiveness and retention, and implementation fidelity, for both recruitment groups. Specifically, these outcomes were: (1) QL connection rates (i.e., enrollment rates) for each intervention group which is defined as the proportion of individuals who called the Sage Call Center and who had confirmed QL connections relative to individuals who called Sage and did not accept the QL connection offered by patient navigators (i.e., excluding all DM individuals who did not call Sage); and (2) a self-reported measure for whether participants were offered QL services after connection with QL operators (in addition to a self-reported reason for not being offered services). 2.2.2. Primary summative outcomes We also examine primary follow-up outcomes assessed by selfreport survey measures with a focus on quit status. Specifically, for these primary summative outcomes we assess self-report smoking abstinence rates for both recruitment groups. Following past research (DiClement et al., 1991), smoking abstinence was defined as quit attempts lasting for 24 h, quit attempts lasting longer than 24 h but not longer than 30 days, or 30 consecutive smoke-free days at the time of the follow-up (i.e., continuous smoking cessation). Follow-up surveys also allowed us to examine other program components via self-report measures including the importance of incentives for QL connections (1 = important, 0 = not important), and the likelihood of contacting QL without recruitment (1 = unlikely, 0 = likely). We also were able to gather information on age, education level, race, ethnicity, and smoking history via follow-up surveys. 2.3. Analytic strategy Primary analyses consisted of descriptive statistics, separately conducted for each intervention group. Inferential statistical analyses were also conducted on both groups combined. Descriptive analyses consisted of cross-tabulations (with Chi-square test statistics) and mean comparisons with unequal variances in order to examine differences between DM and ORC. For examining differences in abstinence rates more systematically, ordered logistic regression models were employed using a three-category breakdown of tobacco cessation (0 = no quit attempt, 1 = quit attempt, 2 = continuous abstinence). These models adjusted for demographics, smoking characteristics, program components, and prior Sage participation. Regression results are
111
Table 1 Survey results and participant characteristics by DM and ORC groups (N = 1179). Variables
DM (N = 418)
ORC (N = 761)
p-Value
% (N) or mean (SD) Background characteristics Female Age Education White African American Native American Hispanic Asian Other races/ethnicities
98.3% 53.8 3.5 69.6% 16.8% 8.4% 2.2% 0.0% 3.1%
(411) (8.19) (0.96) (291) (70) (35) (9) (0) (13)
97.5% 51.8 3.5 77.5% 15.1% 3.0% 2.1% 0.1% 2.1%
(742) (7.40) (0.86) (590) (115) (23) (16) (1) (16)
N.S. b.001 N.S. b.001 b.001 b.001 b.001 b.001 b.001
Program outcomes Quit status No quit attempt Made quit attempt No smoking in past 30 daysa Incentive important for connection Offered quitline services
23.4% 56.5% 20.1% 71.1% 81.6%
(98) (236) (84) (297) (341)
30.9% 53.2% 15.9% 66.6% 82.7%
(235) (405) (121) (507) (629)
b.05 b.05 b.05 N.S. N.S.
Smoking characteristics Years smoked Smoked every day Quit attempt with meds in past Lives with smoker Unlikely to contact quitline
32.4 93.3% 67.0% 37.1% 67.0%
(10.45) (390) (280) (155) (280)
30.5 94.0% 61.6% 38.6% 73.9%
(11.39) (715) (469) (294) (562)
b.01 N.S. N.S. N.S. b.05
Notes: Percentages indicate the percentage of the DM or ORC groups. Statistical significance tests were conducted to compare the two groups; binary variable significance tests were based on chi-square statistic; continuous variable significance tests were based on t-test with unequal variances of mean differences. DM = direct mail. ORC = opportunistic referral with connection. The range for years smoked is 1–62; for age, 18–75; and education ranges from 1 (completed 8th grade or less) to 6 (graduate school). a This is a measure of 30-day point prevalence abstinence after seven months from beginning of intervention, which is the recommended measure of abstinence by the North American Quitline Consortium.
reported within the text (not in the tables). Since patient navigators had different rates of recruitment and connections made, clusteradjusted robust standard errors were used in order to account for clustering within patient navigators and heteroscedasticity. All analyses were run in Stata, version 12. 3. Results 3.1. DM intervention group The DM intervention elicited an 8.5% response rate, defined as the number of individuals who called Sage relative to the number of individuals sent DM. Of the individuals who called Sage within the DM group, there was a QL connection or enrollment rate of 97% for DM individuals. As shown in Table 1, 57% of the DM group made at least one quit attempt after QL connections but relapsed before follow-up, and 20.1% reported at least 30 days of continuous smoking abstinence. For those who relapsed, the average number of reported quit attempts was 4.75 (SD = 6.94). The incentive was an important motivator for QL connections for the majority of the DM group (71%), and over 80% were offered QL services after being connected. Of those not offered services, about 40% reported insurance coverage as the reason,4 and 60% reported 4 Idiosyncrasies exist among state QLs. A conglomerate of QLs exists across Minnesota, but a central QL serves uninsured or inadequately insured individuals. This central Minnesota QL phone line was the only QL used in the current program. Insurance coverage was determined by QL staff after three-way connections by Sage patient navigators. Participants reported post hoc whether their respective health insurance plans would cover QL services; if insurance coverage was available, participants were informed they would be connected to a separate QL by QL operators. All information about whether participants received QL services were gathered via self-report surveys at seven-month follow-up.
112
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
miscellaneous reasons associated with implementation fidelity and quality (e.g., phone connection did not work, accidental disconnection, reported not receiving a call back from QL). Individuals not offered services were included in connection rate and survey analyses (i.e., Intention-to-Treat analysis). No significant differences were found between individuals offered and not offered services in terms of demographic and smoking characteristics. The DM group smoked 32.4 years on average, and the vast majority was daily smokers (see Table 1). A high percentage had attempted to quit in the past with medication, and about 37% lived with a smoker. Most participants indicated they were unlikely to have contacted a QL without the DM recruitment. As noted, because Sage's cancer screening program was the source of the mailing list, this group was comprised almost exclusively of females, with an average age of 53.8 years. The majority of participants had completed high school (or equivalent) or some technical school training and were predominantly non-Hispanic White. Of the non-White population, 17% were African American and 8% were Native American. 3.2. ORC intervention group The QL connection rate for the ORC group was 35%. Displayed in the middle column of Table 1, the majority of the ORC group reported making a quit attempt but relapsed before seven months (53.2%); almost 16% had not smoked in the past 30 days seven months after QL connections. For individuals who relapsed, the average number of quit attempts was 3.92 (SD = 4.91). Almost 70% found the incentive to be an important factor for their QL connection, and approximately 83% was offered QL services after being connected. No significant differences were found between individuals offered and not offered services in terms of demographic or smoking characteristics, and the majority reported miscellaneous reasons for not receiving services. Almost three quarters reported that they were unlikely to have contacted a QL in the absence of ORC. On average, participants had smoked for approximately 31 years. Over 90% reported smoking on a daily basis, fewer than half of participants lived with a smoker, and the majority of the ORC group reported a past quit attempt (with medication). Again, the sample was mostly female, with an average level of education of a high school diploma or equivalent and primarily nonHispanic White. Of the non-White participants in ORC, 15% were African Americans; the remaining 7% were Native American, Hispanic, and other races or ethnicities. 3.3. Ordered logistic regression results The DM group was more likely than the ORC group to make a quit attempt and achieve continuous smoking abstinence, adjusting for all variables in Table 1 (OR = 1.45; 95% CI = 1.09, 1.91). When participants not offered QL services were excluded, a more robust difference between DM and ORC was found, adjusting for all variables in previous models (OR = 1.71; 95% CI = 1.16, 2.52). Additional analyses showed that being offered QL services was strongly related to making a quit attempt and achieving continuous smoking abstinence, adjusting for intervention group membership (DM vs. ORC) and all control variables in previous models (OR = 2.42; 95% CI = 1.90, 3.08). 4. Discussion This study offers implementation and summative findings from an incentive-based, population-level program that utilized two recruitment strategies for connecting low-income, underinsured individuals to tobacco cessation services. Relying on individual-initiated phone contact made with trained patient navigators (i.e., forms of reactive recruitment), the two strategies consisted of (1) DM outreach and (2) opportunistic QL referral.
Both strategies successfully connected low-income smokers to cessation services as well as encouraged the cessation process. Moreover, the interventions disproportionately reached low-socioeconomic individuals as well as racial and ethnic minorities in Minnesota. As an example, Native Americans represented 8.4% of the DM group and 3% of the ORC group whereas they only comprise 2% of the general Minnesota population. Even more strikingly, African Americans comprised 17% of the DM group and 15% of the ORC group but they represent only 6% of the general Minnesota population (see U.S. Census Bureau, 2012). Even though we could find no studies to which our program results could be directly compared, past research using DM and QLs can work as a reference point. For instance, DM elicited a very high response rate (8.5%), almost double the rate of similar DM interventions (see e.g., Schuck et al., 2014). Past research on referring individuals to QLs through primary care settings have demonstrated enrollment rates of 1% to 7% (see Vidrine et al., 2013), indicating our DM response rate as well as connection rates for both DM and ORC (97% and 35%, respectively) are potentially more effective than other QL connection strategies that did not utilize financial incentives. Finally, the results for cessation rates are substantially higher than the annual sustained cessation rate of 5% in the US (see McAfee et al., 2013). DM provided an avenue for reaching a large number of individuals with cessation-related services. DM successfully reached a higher proportion of non-White populations relative to the ORC approach, particularly African American and Native American individuals. Conversely, ORC provided access to a substantial number of individuals seeking health services or information unrelated to tobacco cessation, and it successfully reached more low-income individuals with direct one-on-one communication regarding tobacco cessation. However, individuals in the ORC group were less likely to make QL connections and less likely to quit relative to DM individuals. People in ORC were also less prepared for cessation as indicated by self-reported likelihood of contacting a QL without the intervention. A major strength of DM is that it reaches more individuals with tobacco cessation intervention materials, and it recruits those who are more motivated to quit. Yet since this approach relies on individuals to initiate contact, fewer individuals make it to the stage of phone contact. A major strength of ORC's first phase is that it provides access to a relatively large number of individuals at various stages of motivation to quit. However, ORC leads to lower rates of QL connections and continuous smoking cessation compared to the DM strategy. ORC may reach people who are not yet seriously contemplating quitting, but it may also provide pre-contemplators an avenue to initiate, and potentially succeed at, smoking cessation. In short, ORC is an effective strategy for recruiting less motivated smokers to cessation services (see Asfar et al., 2011). Results highlight that being successfully offered QL services was robustly related to participants' quit status. Whereas the implementation quality associated with the current population-based program is important to note (see Fixsen et al., 2005; Durlak & DuPre, 2008), issues associated with implementation could potentially be bypassed if, for example, DM or ORC strategies were implemented directly through a state's free QL. Other state QLs could feasibly obtain a highly targeted mailing list (e.g., Medicaid, Medicare) and administer the DM strategy through their ongoing operations. Additionally, ORC could potentially be disseminated to health care settings as an alternative to fax-referral strategies (see Willett et al., 2009). Finally, financial incentives were important for both ORC and DM, and future dissemination would require this core component. However, the current analysis was not a randomized controlled trial designed to test intervention components. Components of the current program were demonstrated to be evidence-based approaches for affecting smoking cessation, and the purpose of the current project was to assess the implementation and effectiveness of two incentive-based strategies for recruiting low-income individuals to smoking cessation services in order to address the Affordable Care Act's dictum to scale evidence-
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
based interventions to population-level practice and increase smoking abstinence rates in low-income populations (see Blumenthal et al., 2013; Spoth et al., 2013). Future practice-based research should consider other experimental and quasi-experimental designs, such as a factorial research design (see Collins et al., 2014), to test how each component of the current intervention contributed to its success. Another potential limitation of the study was that the sample was disproportionately female as a result of the overarching programmatic goals of the NBCCEDP. Even though it is important to target low-income female smokers because their smoking rates are persistently high (Stewart et al., 2010), the current results could change if the program were implemented within an all-male sample or a sample with a more equal distribution of males and females. 5. Conclusion Disadvantaged populations have limited access to, and tend to underutilize, quality health care and preventive services (Kassler et al., 2015; Adler & Newman, 2002). Preventive interventions and population-based programs that extend preventive care and tobacco cessation services to underserved populations are pressing public health priorities (Kassler et al., 2015; Blumenthal et al., 2013). The near ubiquity of telephone technology provides an avenue for reaching underserved individuals. Moreover, incentive-based interventions can be successfully scaled to population-level practice, and the two recruitment strategies employed in the current program can successfully target populations disproportionately affected by the tobacco epidemic by extending the reach of state QLs. Potentially other programs, such as other states' NBCCEDPs, Medicaid, and QLs, can target low-income and inadequately insured individuals for referral to smoking cessation services through the use of financial incentives and these recruitment strategies, considering the strengths and weaknesses of both approaches. Disclosures Project funded through Centers for Disease Control and Prevention (American Recovery and Reinvestment Act; Patient Protection and Affordable Care Act); grant FOA DP09-90101SUPP10. The Centers for Disease Control and Prevention had no role in the design, collection, analysis, and interpretation of the data, or the writing of the manuscript and decision to submit for publication. Jon O. Ebbert has received funding from Pfizer and Orexigen and personal fees from GlaxoSmithKline outside of the current research. The authors declare that they have no conflict of interest, and all authors contributed to and have approved the final manuscript. Acknowledgments We thank QUITPLAN® Helpline staff, Shelly Madigan, Sage patient navigators, Janis Taramelli, and Michelle Waste for their efforts.
References Adler, N.E., & Newman, K. (2002). Socioeconomic disparities in health: Pathways and policies. Health Affairs, 21, 60–76. Ammerman, A., Smith, T., & Calancie, L. (2014). Practice-based evidence in public health: Improving reach, relevance, and results. Annual Review of Public Health, 35, 47–63. Asfar, T., Ebbert, J.O., Klesges, R.C., & Relyea, G.E. (2011). Do smoking reduction interventions promote cessation in smokers not ready to quit? Addictive Behaviors, 36, 764–768. Bashshur, R.L., Shannon, G.W., Smith, B.R., Alverson, D.C., Antoniotti, N., ... Yellowlees, P. (2014). The empirical foundations of telemedicine interventions for chronic disease management. Telemedicine and e-Health, 20, 769–800. Becker, M.H. (1974). The health belief model and personal health behavior. Health Education Monographs, 2, 324–473. Blumenthal, K.J., Saulsgiver, K.A., Norton, L., Toxel, A.B., Anarell, J.P., ... Volpp, K.G. (2013). Medicaid incentive programs to encourage healthy behavior show mixed results to date and should be studied and improved. Health Affairs, 32(3), 497–507. Bryant, J., Bonevski, B., Paul, C., McElduff, P., & Attia, J. (2011). A systematic review and meta-analysis of the effectiveness of behavioural smoking cessation interventions in selected disadvantaged groups. Addiction, 106, 1568–1585.
113
Burns, E.K., Deaton, E.A., & Levinson, A.H. (2011). Rates and reasons: Disparities in low intentions to use a state smoking cessation quitline. American Journal of Health Promotion, 25, S59–S65. Centers for Disease Control and Prevention (2014). Current cigarette smoking among adults—United States, 2005–2012. MMWR, 63(2), 29–34. Collins, L.M., Dziak, J.J., Kugler, K.C., & Trail, J.B. (2014). Factorial experiments: Efficient tools for evaluation of intervention components. American Journal of Preventive Medicine, 47, 498–504. DiClement, C.C., Prochaska, J.O., Fairhurst, S.K., Velicer, W.F., Velasquez, M.M., & Rossi, J.S. (1991). The process of smoking cessation: An analysis of precontemplation, contemplation, and preparation of stages of change. Journal of Consulting and Clinical Psychology, 59, 295–304. Durlak, J.A., & DuPre, E.P. (2008). Implementation matters: A review of research on the influence of implementation on program outcomes and the factors affecting implementation. American Journal of Community Psychology, 41, 327–350. Fiore, M.C., et al. (2008). Treating tobacco use and dependence: Clinical practice guidelines 2008 update. Washington, DC: US Department of Health and Human Services. Fixsen, D.L., Naoom, S.F., Blasé, K.A., Friedman, R.M., & Wallace, F. (2005). Implementation research: A synthesis of the literature. Tampa, FL: University of South Florida, Louis de la Parte Florida Mental Health Institute, The National Implementation Research Network (FMHI Publication #231). Freund, K.M., Battaglia, T.A., Calhoun, E., Dudley, D.J., Fiscella, K., ... The Patient Navigation Research Program Group (2008). National Cancer Institute Patient Navigation Research Program: Methods, protocol, and measures. Cancer, 113(12), 3391–3399. Glasgow, R.E., Vogt, T.M., & Boles, S.M. (1999). Evaluating the public health impact of health promotion interventions: The RE-AIM framework. American Journal of Public Health, 89, 1322–1327. Green, L.W. (2008). Making research relevant: If it is an evidence-based practice, where's the practice-based evidence? Family Practice, 25(S1), i20–i24. Jha, P., Peto, R., Zatonski, W., Boreham, J., Jarvis, M.J., & Lopez, A.D. (2006). Social inequalities in male mortality, and in male mortality from smoking: Indirect estimation from national death rates in England and Wales, Poland, and North America. Lancet, 368, 367–370. Kassler, W.J., Tomoyasu, N., & Conway, P.H. (2015). Beyond a traditional payer—CMS's role in improving population health. New England Journal of Medicine, 372, 109–111. Lee, N.C., Wong, F.L., Jamison, P.M., Jones, S.F., Galaska, L., ... Stokes-Townsend, G. -A. (2014). Implementation of the National Breast and Cervical Cancer Early Detection Program: The beginning. Cancer, 120(16), 2540–2548. Lewis, S. (2010). Creating incentives to improve population health. Preventing Chronic Disease, 7, A94. Lichtenstein, E., Glasglow, R.E., Lando, H.A., Ossip-Klein, D.J., & Boles, S.M. (1996). Telephone counseling for smoking cessation: Rationales and meta-analytic review of evidence. Health Education Research, 11, 243–257. Mathew, A.R., Burris, J.L., Alberg, A.J., Cummings, K.M., & Carpenter, M.J. (2014). Impact of a brief telephone referral on quitline use, quit attempts and abstinence. Health Education Research, Advance Online Publication. McAfee, T., Davis, K.C., Alexander, R.L., Pechacek, T.F., & Bunnell, R. (2013). Effect of the first federally funded US antismoking national media campaign. Lancet, 382, 14–20. Consortium, N.A.Q. (2011). Guides on quitlines and research. Retrieved from http://c. ymcdn.com/sites/www.naquitline.org/resource/resmgr/research/ researchguideapril2013.pdf. Parks, M.J., Slater, J.S., Rothman, A.J., & Nelson, C.L. (2015). Interpersonal communication and smoking cessation in the context of an incentive-based program: Survey evidence from a telehealth intervention in a low-income population. Journal of Health Communication, Advance Online Publication. http://dx.doi.org/10.1080/10810730. 2015.1039677. Patten, C.A., Smith, C.M., Brockman, T.A., Decker, P.A., Hughes, C.A., ... Zhu, S. -H. (2011). Support-person promotion of a smoking quitline: A randomized control trial. American Journal of Preventive Medicine, 41(1), 17–23. Prochaska, J.O., Redding, C.A., & Evers, K.E. (2008). The transtheoretical model and stages of change. In K. Glanz, B.K. Rimer, & K. Viswanath (Eds.), Health behavior and health education: Theory, research, and practice (pp. 97–121) (4th ed.). San Francisco, CA: Jossey-Bass. Rollnick, S., Miller, W.R., & Butler, C.C. (2008). Motivational interviewing in health care: Helping patients change behavior. New York, NY: The Guilford Press. Rothman, A.J., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin, 121(1), 3–19. Rothschild, M.L., & Gaidis, W.C. (1981). Behavioral learning theory: Its relevance to marketing and promotion. Journal of Marketing, 13, 263–269. Sanson-Fisher, R.W., Bonevski, B., Green, L.W., & D'Este, C. (2007). Limitations of the randomized controlled trial in evaluating population-based health interventions. American Journal of Preventive Medicine, 33(2), 155–161. Schuck, K., Bricker, J.B., Otten, R., Kleinjan, M., Brandon, T.H., & Engels, R.C.M.E. (2014). Effectiveness of proactive quitline counselling for smoking parents recruited through primary schools: Results of a randomized controlled trial. Addiction, 109, 830–841. Sigmon, S.C., & Patrick, M.E. (2012). The use of financial incentives in promoting smoking cessation. Preventive Medicine, 55, S24–S32. Slater, J.S., Henly, G.A., Ha, C.N., Malone, M.E., Nyman, J.A., ... McGovern, P.G. (2005). Effect of direct mail as a population-based strategy to increase mammography use among low-income under-insured women ages 40 to 64 years. Cancer Epidemiology, Biomarkers & Prevention, 14(10), 2346–2352. Soet, J.E., & Basch, C.E. (1997). The telephone as a communication medium for health education. Health Education & Behavior, 24, 759–772. Solomon, L.J., Marcy, T.W., Howe, K.D., Skelly, J.M., Reinier, K., & Flynn, B.S. (2005). Does extended proactive telephone support increase smoking cessation among lowincome women using nicotine patches? Preventive Medicine, 40, 306–313.
114
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
Spoth, R., Rohrbach, L.A., Greenberg, M., Leaf, P., Brown, H., ... Society for Prevention Research Type 2 Translational Task Force Members and Contributing Authors (2013). Addressing core challenges for the next generation of type 2 translational research and systems: The Translation Science to Population Impact (TSci Impact) framework. Prevention Science, 14, 319–351. Stead, L.F., Hartmann-Boyce, J., Perera, R., & Lancaster, T. (2013). Telephone counselling for smoking cessation. Cochrane Database of Systematic Reviews, 8, CD002850. Stead, L.F., Perera, R., & Lancaster, T. (2007). A systematic review of interventions for smokers who contact quitlines. Tobacco Control, 16, i3–i8. Stewart, M.J., Kushner, K.E., Greaves, L., Letourneau, N., Spitzer, D., & Boscoe, M. (2010). Impacts of a support intervention for low-income women who smoke. Social Science & Medicine, 71, 1901–1909. Thomas, S., Fayter, D., Misso, K., Ogilvie, D., Petticrew, M., ... Worthy, G. (2008). Population tobacco control interventions and their effects on social inequalities in smoking: A systematic review. Tobacco Control, 17, 230–237. U.S. Census Bureau (2012). State and county quickfacts: Minnesota. http://quickfacts. census.gov/qfd/states/27000.html Vidrine, J.I., Shete, S., Cao, Y., Greisinger, A., Harmonson, P., ... Wetter, D.W. (2013). Ask-advise-connect: A new approach to smoking treatment delivery in health care settings. Journal of the American Medical Association Internal Medicine, 173, 458–464. Volpp, K.G., Troxel, A.B., Pauly, M.V., Glick, H.A., Puig, A., ... Audrain-McGovern, J. (2009). Randomized, controlled trial of financial incentives for smoking cessation. New England Journal of Medicine, 360, 699–709.
Weinstein, N.D., Sandman, P.M., & Blalock, S.J. (2008). The precaution adoption process model. In K. Glanz, B.K. Rimer, & K. Viswanath (Eds.), Health behavior and health education: Theory, research, and practice (pp. 123–147) (4th ed.). San Francisco, CA: Jossey-Bass. Willett, J.G., Hood, N.E., Burns, E.K., Swetlick, J.L., Wilson, S.M., ... Levinson, A.H. (2009). Clinical faxed referrals to a tobacco quitline: Reach, enrollment, and participant characteristics. American Journal of Preventive Medicine, 36, 337–340. Witte, K., & Allen, M. (2000). A meta-analysis of fear appeals: Implications for effective public health campaigns. Health Education & Behavior, 27, 591–615. Witte, K., Meyer, G., & Martell, D. (2001). Effective health risk messages: A step-by-step guide. Thousand Oaks, CA: Sage Publications, Inc. Wootton, R. (2012). Twenty years of telemedicine in chronic disease management—An evidence synthesis. Journal of Telemedicine and Telecare, 18, 211–220. Wootton, R., Jebamani, L.S., & Dow, S.A. (2005). E-health and the universitas 21 organization: 2. Telemedicine and underserved populations. Journal of Telemedicine and Telecare, 11, 221–224. Zhu, S. -H., Gardiner, P., Cummins, S., Anderson, C., Wong, S., ... Gamst, A. (2011). Quitline utilization rates of African-American and white smokers: The California experience. American Journal of Health Promotion, 25, S51–S58. Zhu, S. -H., Lee, M., Zhuang, Y. -L., Gamst, A., & Wolfson, T. (2012). Interventions to increase smoking cessation at the population level: How much progress has been made in the last two decades? Tobacco Control, 21, 110–118.
Contents lists available at ScienceDirect
Addictive Behaviors
Connecting low-income smokers to tobacco treatment services Jonathan S. Slater a,b, Christina L. Nelson a, Michael J. Parks a,⁎, Jon O. Ebbert c a b c
Minnesota Department of Health, 85 East 7th Place, Saint Paul, MN 55164, USA Masonic Cancer Center, 424 Harvard St SE, University of Minnesota, Minneapolis, MN, USA Division of Primary Care Internal Medicine, 221 Fourth Avenue SW, Mayo Clinic, Rochester, MN 55905, USA
H I G H L I G H T S • • • • •
We designed two strategies for connecting low-income smokers to quitline services. Strategies were direct mail and opportunistic referral; both employed incentives. Each strategy reached individuals at differential levels of readiness to quit. Smoking abstinence rates at follow-up indicated both strategies had high impact. Both strategies had strengths, and both can be used in population-based practice.
a r t i c l e
i n f o
Article history: Received 1 June 2015 Received in revised form 2 October 2015 Accepted 11 October 2015 Available online 18 October 2015 Keywords: Population-based programs Smoking cessation Underserved populations Telehealth Financial incentive
a b s t r a c t The Affordable Care Act calls for using population-level incentive-based interventions, and cigarette smoking is one of the most significant health behaviors driving costs and adverse health in low-income populations. Telehealth offers an opportunity to facilitate delivery of evidence-based smoking cessation services as well as incentive-based interventions to low-income populations. However, research is needed on effective strategies for linking smokers to services, how to couple financial incentives with telehealth, and on how to scale this to population-level practice. The current paper evaluates primary implementation and follow-up results of two strategies for connecting low-income, predominantly female smokers to a telephone tobacco quitline (QL). The population-based program consisted of participant-initiated phone contact and two recruitment strategies: (1) direct mail (DM) and (2) opportunistic telephone referrals with connection (ORC). Both strategies offered financial incentives for being connected to the QL, and all QL connections were made by trained patient navigators through a central call center. QL connections occurred for 97% of DM callers (N = 870) and 33% of ORC callers (N = 4550). Self-reported continuous smoking abstinence (i.e., 30 smoke-free days at seven-month followup) was 20% for the DM group and 16% for ORC. These differences between intervention groups remained in ordered logistic regressions adjusting for smoking history and demographic characteristics. Each recruitment strategy had distinct advantages; both successfully connected low-income smokers to cessation services and encouraged quit attempts and continuous smoking abstinence. Future research and population-based programs can utilize financial incentives and both recruitment strategies, building on their relative strengths. Published by Elsevier Ltd.
1. Introduction Smoking prevalence among U.S. adults is 17% for those who live at or above the poverty level but 28% for those below it (Centers for Disease Control and Prevention, 2014), underscoring how income underlies disparities in cigarette smoking and smoking-related health problems (Jha et al., 2006; Thomas et al., 2008). High smoking rates persist among low-income women (Stewart et al., 2010), and smoking can account ⁎ Corresponding author at: Minnesota Department of Health, 85 East 7th Street, Saint Paul, MN 55164, USA. E-mail address: [email protected] (M.J. Parks).
http://dx.doi.org/10.1016/j.addbeh.2015.10.013 0306-4603/Published by Elsevier Ltd.
for up to half of mortality disparities associated with socioeconomic status among males (Jha et al., 2006). Financial incentives are known to successfully promote smoking cessation, especially in low-income populations (Blumenthal et al., 2013; Bryant et al., 2011; Volpp et al., 2009; Sigmon & Patrick, 2012), yet a dearth of research exists on how such evidence-based interventions translate to population-level practice (Ammerman, Smith, & Calancie, 2014; Spoth et al., 2013; Lewis, 2010). Subsequently, a more translatable evidence base consisting of practice-based evaluations and not necessarily randomized trials of population-level interventions is needed (Green, 2008; Sanson-Fisher, Bonevski, Green & D'Este, 2007; Ammerman et al., 2014). Such evidence is critical for addressing public health priorities exemplified by the
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
Affordable Care Act such as scaling up incentive-based interventions in order to promote smoking cessation in low-income populations (Blumenthal et al., 2013: 497–498; Kassler, Tomoyasu, & Conway, 2015). Telehealth has potential for effectively delivering cessation services to large numbers in underserved populations (Bashshur et al., 2014; Wootton et al., 2005; Wootton, 2012). As a primary example, free state telephone tobacco quitlines (QLs) offer an evidence-based and population-level approach to increasing smoking abstinence rates (Stead et al., 2007; Fiore et al., 2008), and low-income and non-White populations are inclined to use free QLs (Burns et al., 2011; Zhu et al., 2011). However, QL utilization rates are markedly low across the U.S., with a state-level average of approximately 2% (Zhu et al., 2012). Consequently, more research is needed on strategies that successfully extend the reach of QLs, particularly to low-income populations (Zhu et al., 2012). Specifically, QL research is needed that (1) incorporates financial incentives, (2) targets individuals at various stages of motivation for quitting, and (3) focuses on “reactive” recruitment rather than “proactive” approaches (Stead et al., 2013; Asfar et al., 2011; Mathew et al., 2014). Few published QL interventions have utilized financial incentives (Stead et al., 2013), and more interventions need to reach adults not motivated to quit (Asfar et al., 2011). Although proactive telephone-based cessation interventions (i.e., calls initiated by counselors [Lichtenstein et al., 1996]) have been effective in low-income populations (e.g., Solomon et al., 2005), less is known about reactive strategies (i.e., cessation counseling provided on demand). Proactive QL recruitment is designed to contact potential participants directly with QL operators in order to connect smokers directly to the QL at time of initial contact. Proactive strategies may potentially discourage participation of individuals who are not ready to take immediate and direct behavioral steps towards quitting. A reactive strategy is designed to refer potential users to a QL, with the expectation that individuals who are willing to participate will contact the QL at an appropriate time for them after receipt of referral. Reactive strategies are important because they have potential to recruit those who are ready to take action-oriented steps towards quitting, and it gives individuals time to contemplate possible steps towards quitting without immediate pressure. Reactive strategies can also encourage indirect pathways to behavior change via naturally occurring psychosocial mechanisms within the environment that individuals receive their initial referral to the QL (see e.g., Parks et al., 2015).
1.1. The current study We report primary implementation and follow-up results from a population-based program that utilized financial incentives and two strategies designed to connect low-income smokers to Minnesota's QL, among a low-income sample primarily comprised of females. Mirroring patterns across the U.S., QL reach and utilization rates are low in Minnesota (see Patten et al., 2011). Following seminal research in implementation science (e.g., Fixsen et al., 2005; Glasgow, Vogt, & Boles, 1999), we examine participant responsiveness and retention as well as program fidelity measured via response rates and QL connection rates. We also examine primary follow-up and effectiveness outcomes measured as smoking status at time of follow-up with a focus on smoking abstinence rates. We examine these outcomes for two recruitment strategies separately, providing a basis for comparative effectiveness. Since direct mail is a cost-effective, population-level strategy for connecting individuals to preventive and telehealth services (Slater et al., 2005; Soet & Basch, 1997), one recruitment strategy was direct mail. The second was a centralized patient navigation system. Both strategies relied on individual-initiated phone contact (see Soet & Basch, 1997) and offered financial incentives to low-income smokers for being connected to the QL via three-way phone calls conducted by trained patient navigators (see Methods section for details).
109
2. Methods 2.1. Participants and intervention 2.1.1. Overview and setting From September 2010 to September 2012, the program was implemented through “Sage Programs”: Sage, Minnesota's National Breast and Cervical Cancer Early Detection Program (NBCCEDP) and Sage Scopes, Minnesota's Colorectal Cancer Control Program (CRCCPs) at the Minnesota Department of Health (see Lee et al., 2014; Slater et al., 2005). The recruitment timeframe was contingent on funding; consequently, funding and recruitment ended concurrently. Sage provides free breast and cervical cancer screening services to inadequately insured women 40 years of age or older, with household incomes at or below 250% of the US federal poverty level. Sage Scopes provides free colorectal cancer screening to a much more limited number of men and women ages 50 and older but who otherwise meet the same eligibility criteria as Sage. Unique among NBCCEDPs and CRCCPs, Sage Programs (hereafter referred to as Sage) has a single call center staffed by “patient navigators” (see Freund et al., 2008) trained in motivational interviewing (see Rollnick, Miller, & Butler, 2008). More than 22% of Sage participants smoke cigarettes. Since NBCCEDPs target lowincome females, as previously noted, the current program focuses on a sample that is disproportionately female. 2.1.2. Intervention The current program offered a $20 incentive to callers for being connected to Minnesota's QL via a three-way call conducted by Sage patient navigators. Two recruitment strategies were used: (1) direct mail (DM) and (2) opportunistic referral with connection (ORC) through the Sage Call Center call transfer system. DM was designed to prompt cigarette smokers to call Sage's toll-free phone number rather than serve an educational function. Mailers consisted of a folded card with emotionally evocative messages and graphics as well as a small insert card advertising the financial incentive offer. We employed a loss-framed message (see Rothman & Salovey, 1997) coupled with a high-efficacy message (see Witte & Allen, 2000). These DM designs are based on health communication and health behavior theory. Specifically, the fear appeal message or loss-frame message is designed to inform an individual that a certain behavior will lead to an undesirable outcome, such as long-term smoking leading to physical disability or mortality (see Rothman and Salovey, 1997). Health communication research shows that these health messages based on fear appeal coupled with a clear articulation of achievable behavioral steps (i.e., high-efficacy message) produce the greatest behavior change (Rothman and Salovey, 1997; Witte and Allen, 2000; Slater et al., 2005). Theory suggests that responses to such health messages can be either a “danger control action” or a “defensive response.” Danger control actions are actions taken when individuals feel (1) susceptible to a health problem and (2) capable of completing behavioral steps necessary for reducing the risk of the presented health problem. Defensive responses are actions antithetical to the proposed behavior change (i.e., behavioral steps away from the protective response) that are based on feelings of high fear in addition to low self-efficacy (for a more thorough discussion see Witte et al., 2001). The purpose of the design was to promote risk susceptibility associated with cigarette smoking in addition to offering clear and achievable steps for action. An example of a mailer is presented in Fig. 1. Following past research, two rounds of mailings were employed (Slater et al., 2005). The incentive offer was presented via a small inserted card affixed to the inside of the mailer. This inserted card read: “Special offer: Call today and we'll pay you $20 plus give you the free tools to quit smoking,” and it included Sage's toll-free phone number with a tracking promotion code that patient navigators recorded during calls. The presentation of the incentive offer within the mailer was intended to influence the decisional balance by reducing perceived barriers, providing
110
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
and no further information was gathered if they refused to be connected. QL services included free smoking cessation telephone counseling with a maximum of five sessions within a six-month period. If desired, the QL provided self-help materials and free nicotine replacement therapy, and QL staff scheduled follow-up calls with patients. It is also important to highlight that the financial incentive was only offered to individuals who completed three-way QL connections and who did not disconnect prior to speaking with QL operators. The incentive was not linked to the extent to which they used available services. This was done for practical reasons (utilization data could not be directly obtained from the QL staff), but this has precedence in consumer marketing. It is analogous to incenting people via a coupon based on the expectation that this exposure is sufficient for some people to become customers (see e.g., Rothschild & Gaidis, 1981). The incentive was a $20 gift card, similar to a credit card that is almost universally accepted for goods and services. Once connection to the QL was confirmed, patient navigators used the participants' information to send gift cards via mail. Participants who completed the QL connection were interviewed via telephone by trained staff at least seven months after the initial QL connection. The average time between initial QL transfer and follow-up surveys was 8.8 months (SD = 2.84).1 There was no contact between Sage patient navigators and participants, or between surveyors and participants, during the timeframe after initial QL connection and before follow-up surveys because QL operators were responsible for cessation-oriented contact after three-way transfers from the Call Center. Surveys and survey timing were modeled after North American Quitline Consortium (NAQC) recommendations (see North American Quitline Consortium, 2011). Ten call attempts were made before participants were deemed unresponsive, and these contacts were made at various times and on various days of the week. Fig. 1. Direct mail mailer example with loss-frame and high-efficacy message.
cues to action, and making pros of behavior change more salient (see Becker, 1974; Prochaska et al., 2008; Weinstein et al., 2008). Recruitment to ORC consisted of Sage's patient navigators obtaining the smoking status of all individuals who called the Sage Call Center for breast, cervical, or colorectal cancer screening information or appointments and opportunistically presenting the QL referral offer (i.e., a $20 incentive for being connected to the QL) to self-reported smokers. Patient navigators handled phone calls for both intervention groups (six full-time navigators were involved). Navigators followed a script designed to: (1) identify whether the caller was responding to a mailing, (2) record the promotion code printed on the mailer and financial incentive card, (3) determine callers' decision about participation, and (4) make QL connections for willing participants. For willing participants, patient navigators put participants on hold and called QL agents. Patient navigators then confirmed with QL operators that a Sage patient would be connected to the QL via a three-way call. Once confirmed, patient navigators conferenced the Sage participant back into the call and remained on the line until communication between participants and QL operators was established. Thus, QL connection was defined as a confirmed connection between participants and QL operators via three-way calls conducted by patient navigators. Once participants were connected to the QL, trained QL operators were responsible for handling participants and administering cessation-related services. All individuals who agreed to be connected to the QL were deemed as “enrolled” in the program. Approximately 45% of callers enrolled in the program, and there were major differences in enrollment rates between recruitment strategies (DM = 97% enrollment vs. ORC = 35% enrollment). For enrolled individuals, demographic information was gathered at follow-up and promotion codes were recorded prior to QL connections. Callers were considered nonparticipants
2.1.3. Participants and sample size The DM intervention targeted 9922 low-income smokers who had previously been screened for cancer through Sage. Smoking status was determined by two means: (1) a prior record of being a smoker obtained during previous contact with Sage, or (2) Sage enrollment forms (e.g., from local health clinics) that identified the individual as a smoker. The ORC intervention was provided to smokers from among the approximately 23,000 calls fielded by Sage patient navigators through the Call Center during the intervention period. A total of 870 smokers responded to the DM intervention, and the ORC strategy was administered to 4550 self-identified smokers. After being presented with the script by patient navigators, 2456 smokers were connected to the QL: 844 from the DM group and 1612 from the ORC group. Because they are Sage's target population, the sample was predominantly (98%) female (1153 females and 26 males) across both interventions. The analytic sample did not exclude males for multiple reasons. Even though the primary target of the program was lowincome females, dramatic mortality disparities associated with smoking and income exist among males. The males in the sample were also found to be heavy smokers,2 and excluding males from the analysis had potential to bias program effects in terms of increasing quit rates and connection rates. The analytic sample is based on participants who were reached with the survey at time of follow-up after QL connections.3 The final analytic sample was comprised of 418 DM and 1 Supplementary analyses demonstrated that survey timing was not systematically related to smoking cessation outcomes or participant characteristics. 2 There were no differences between males and females in regards to demographic and smoking history variables, except for years smoked. One-way ANOVA and t-test with unequal variance showed males' average years smoked (36) was greater than females' years smoked (31) (p b .05). 3 Descriptive information was not gathered for the majority of participants who made QL connections, and therefore it was not possible to compare individuals who participated in the follow-up survey versus those who did not complete the survey.
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
761 ORC participants who had complete data; approximately 3% and 4% (DM and ORC, respectively) of the surveyed participants were removed because of incomplete data. For both intervention groups, no significant differences were found between the analytic sample and individuals eliminated because of missing data in terms of tobacco cessation and background characteristics. We use the total sample to examine implementation results, and summative results strictly rely on the final analytic sample. For implementation results we consider “enrollment” rates and response rates prior to follow-up surveys. We calculate enrollment rates based on connection rates subsequent to contact with patient navigators because patient navigation was a core component of the program and non-callers in the DM group did not receive this component. There is an implementation outcome of participant responsiveness that is unique to the DM group (i.e., call rate in response to the mailing) consisting of the response rate to the DM campaign measured as the total number of DM callers that called Sage (prior to QL connections) relative to the total number of DM individuals who were sent mailers. All summative results assessing quit status utilize follow-up survey results comprised of the final analytic sample of 418 DM and 761 ORC participants. 2.2. Outcomes 2.2.1. Implementation outcomes As noted, we examine primary implementation outcomes (see Durlak & DuPre, 2008) that assess participant responsiveness and retention, and implementation fidelity, for both recruitment groups. Specifically, these outcomes were: (1) QL connection rates (i.e., enrollment rates) for each intervention group which is defined as the proportion of individuals who called the Sage Call Center and who had confirmed QL connections relative to individuals who called Sage and did not accept the QL connection offered by patient navigators (i.e., excluding all DM individuals who did not call Sage); and (2) a self-reported measure for whether participants were offered QL services after connection with QL operators (in addition to a self-reported reason for not being offered services). 2.2.2. Primary summative outcomes We also examine primary follow-up outcomes assessed by selfreport survey measures with a focus on quit status. Specifically, for these primary summative outcomes we assess self-report smoking abstinence rates for both recruitment groups. Following past research (DiClement et al., 1991), smoking abstinence was defined as quit attempts lasting for 24 h, quit attempts lasting longer than 24 h but not longer than 30 days, or 30 consecutive smoke-free days at the time of the follow-up (i.e., continuous smoking cessation). Follow-up surveys also allowed us to examine other program components via self-report measures including the importance of incentives for QL connections (1 = important, 0 = not important), and the likelihood of contacting QL without recruitment (1 = unlikely, 0 = likely). We also were able to gather information on age, education level, race, ethnicity, and smoking history via follow-up surveys. 2.3. Analytic strategy Primary analyses consisted of descriptive statistics, separately conducted for each intervention group. Inferential statistical analyses were also conducted on both groups combined. Descriptive analyses consisted of cross-tabulations (with Chi-square test statistics) and mean comparisons with unequal variances in order to examine differences between DM and ORC. For examining differences in abstinence rates more systematically, ordered logistic regression models were employed using a three-category breakdown of tobacco cessation (0 = no quit attempt, 1 = quit attempt, 2 = continuous abstinence). These models adjusted for demographics, smoking characteristics, program components, and prior Sage participation. Regression results are
111
Table 1 Survey results and participant characteristics by DM and ORC groups (N = 1179). Variables
DM (N = 418)
ORC (N = 761)
p-Value
% (N) or mean (SD) Background characteristics Female Age Education White African American Native American Hispanic Asian Other races/ethnicities
98.3% 53.8 3.5 69.6% 16.8% 8.4% 2.2% 0.0% 3.1%
(411) (8.19) (0.96) (291) (70) (35) (9) (0) (13)
97.5% 51.8 3.5 77.5% 15.1% 3.0% 2.1% 0.1% 2.1%
(742) (7.40) (0.86) (590) (115) (23) (16) (1) (16)
N.S. b.001 N.S. b.001 b.001 b.001 b.001 b.001 b.001
Program outcomes Quit status No quit attempt Made quit attempt No smoking in past 30 daysa Incentive important for connection Offered quitline services
23.4% 56.5% 20.1% 71.1% 81.6%
(98) (236) (84) (297) (341)
30.9% 53.2% 15.9% 66.6% 82.7%
(235) (405) (121) (507) (629)
b.05 b.05 b.05 N.S. N.S.
Smoking characteristics Years smoked Smoked every day Quit attempt with meds in past Lives with smoker Unlikely to contact quitline
32.4 93.3% 67.0% 37.1% 67.0%
(10.45) (390) (280) (155) (280)
30.5 94.0% 61.6% 38.6% 73.9%
(11.39) (715) (469) (294) (562)
b.01 N.S. N.S. N.S. b.05
Notes: Percentages indicate the percentage of the DM or ORC groups. Statistical significance tests were conducted to compare the two groups; binary variable significance tests were based on chi-square statistic; continuous variable significance tests were based on t-test with unequal variances of mean differences. DM = direct mail. ORC = opportunistic referral with connection. The range for years smoked is 1–62; for age, 18–75; and education ranges from 1 (completed 8th grade or less) to 6 (graduate school). a This is a measure of 30-day point prevalence abstinence after seven months from beginning of intervention, which is the recommended measure of abstinence by the North American Quitline Consortium.
reported within the text (not in the tables). Since patient navigators had different rates of recruitment and connections made, clusteradjusted robust standard errors were used in order to account for clustering within patient navigators and heteroscedasticity. All analyses were run in Stata, version 12. 3. Results 3.1. DM intervention group The DM intervention elicited an 8.5% response rate, defined as the number of individuals who called Sage relative to the number of individuals sent DM. Of the individuals who called Sage within the DM group, there was a QL connection or enrollment rate of 97% for DM individuals. As shown in Table 1, 57% of the DM group made at least one quit attempt after QL connections but relapsed before follow-up, and 20.1% reported at least 30 days of continuous smoking abstinence. For those who relapsed, the average number of reported quit attempts was 4.75 (SD = 6.94). The incentive was an important motivator for QL connections for the majority of the DM group (71%), and over 80% were offered QL services after being connected. Of those not offered services, about 40% reported insurance coverage as the reason,4 and 60% reported 4 Idiosyncrasies exist among state QLs. A conglomerate of QLs exists across Minnesota, but a central QL serves uninsured or inadequately insured individuals. This central Minnesota QL phone line was the only QL used in the current program. Insurance coverage was determined by QL staff after three-way connections by Sage patient navigators. Participants reported post hoc whether their respective health insurance plans would cover QL services; if insurance coverage was available, participants were informed they would be connected to a separate QL by QL operators. All information about whether participants received QL services were gathered via self-report surveys at seven-month follow-up.
112
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
miscellaneous reasons associated with implementation fidelity and quality (e.g., phone connection did not work, accidental disconnection, reported not receiving a call back from QL). Individuals not offered services were included in connection rate and survey analyses (i.e., Intention-to-Treat analysis). No significant differences were found between individuals offered and not offered services in terms of demographic and smoking characteristics. The DM group smoked 32.4 years on average, and the vast majority was daily smokers (see Table 1). A high percentage had attempted to quit in the past with medication, and about 37% lived with a smoker. Most participants indicated they were unlikely to have contacted a QL without the DM recruitment. As noted, because Sage's cancer screening program was the source of the mailing list, this group was comprised almost exclusively of females, with an average age of 53.8 years. The majority of participants had completed high school (or equivalent) or some technical school training and were predominantly non-Hispanic White. Of the non-White population, 17% were African American and 8% were Native American. 3.2. ORC intervention group The QL connection rate for the ORC group was 35%. Displayed in the middle column of Table 1, the majority of the ORC group reported making a quit attempt but relapsed before seven months (53.2%); almost 16% had not smoked in the past 30 days seven months after QL connections. For individuals who relapsed, the average number of quit attempts was 3.92 (SD = 4.91). Almost 70% found the incentive to be an important factor for their QL connection, and approximately 83% was offered QL services after being connected. No significant differences were found between individuals offered and not offered services in terms of demographic or smoking characteristics, and the majority reported miscellaneous reasons for not receiving services. Almost three quarters reported that they were unlikely to have contacted a QL in the absence of ORC. On average, participants had smoked for approximately 31 years. Over 90% reported smoking on a daily basis, fewer than half of participants lived with a smoker, and the majority of the ORC group reported a past quit attempt (with medication). Again, the sample was mostly female, with an average level of education of a high school diploma or equivalent and primarily nonHispanic White. Of the non-White participants in ORC, 15% were African Americans; the remaining 7% were Native American, Hispanic, and other races or ethnicities. 3.3. Ordered logistic regression results The DM group was more likely than the ORC group to make a quit attempt and achieve continuous smoking abstinence, adjusting for all variables in Table 1 (OR = 1.45; 95% CI = 1.09, 1.91). When participants not offered QL services were excluded, a more robust difference between DM and ORC was found, adjusting for all variables in previous models (OR = 1.71; 95% CI = 1.16, 2.52). Additional analyses showed that being offered QL services was strongly related to making a quit attempt and achieving continuous smoking abstinence, adjusting for intervention group membership (DM vs. ORC) and all control variables in previous models (OR = 2.42; 95% CI = 1.90, 3.08). 4. Discussion This study offers implementation and summative findings from an incentive-based, population-level program that utilized two recruitment strategies for connecting low-income, underinsured individuals to tobacco cessation services. Relying on individual-initiated phone contact made with trained patient navigators (i.e., forms of reactive recruitment), the two strategies consisted of (1) DM outreach and (2) opportunistic QL referral.
Both strategies successfully connected low-income smokers to cessation services as well as encouraged the cessation process. Moreover, the interventions disproportionately reached low-socioeconomic individuals as well as racial and ethnic minorities in Minnesota. As an example, Native Americans represented 8.4% of the DM group and 3% of the ORC group whereas they only comprise 2% of the general Minnesota population. Even more strikingly, African Americans comprised 17% of the DM group and 15% of the ORC group but they represent only 6% of the general Minnesota population (see U.S. Census Bureau, 2012). Even though we could find no studies to which our program results could be directly compared, past research using DM and QLs can work as a reference point. For instance, DM elicited a very high response rate (8.5%), almost double the rate of similar DM interventions (see e.g., Schuck et al., 2014). Past research on referring individuals to QLs through primary care settings have demonstrated enrollment rates of 1% to 7% (see Vidrine et al., 2013), indicating our DM response rate as well as connection rates for both DM and ORC (97% and 35%, respectively) are potentially more effective than other QL connection strategies that did not utilize financial incentives. Finally, the results for cessation rates are substantially higher than the annual sustained cessation rate of 5% in the US (see McAfee et al., 2013). DM provided an avenue for reaching a large number of individuals with cessation-related services. DM successfully reached a higher proportion of non-White populations relative to the ORC approach, particularly African American and Native American individuals. Conversely, ORC provided access to a substantial number of individuals seeking health services or information unrelated to tobacco cessation, and it successfully reached more low-income individuals with direct one-on-one communication regarding tobacco cessation. However, individuals in the ORC group were less likely to make QL connections and less likely to quit relative to DM individuals. People in ORC were also less prepared for cessation as indicated by self-reported likelihood of contacting a QL without the intervention. A major strength of DM is that it reaches more individuals with tobacco cessation intervention materials, and it recruits those who are more motivated to quit. Yet since this approach relies on individuals to initiate contact, fewer individuals make it to the stage of phone contact. A major strength of ORC's first phase is that it provides access to a relatively large number of individuals at various stages of motivation to quit. However, ORC leads to lower rates of QL connections and continuous smoking cessation compared to the DM strategy. ORC may reach people who are not yet seriously contemplating quitting, but it may also provide pre-contemplators an avenue to initiate, and potentially succeed at, smoking cessation. In short, ORC is an effective strategy for recruiting less motivated smokers to cessation services (see Asfar et al., 2011). Results highlight that being successfully offered QL services was robustly related to participants' quit status. Whereas the implementation quality associated with the current population-based program is important to note (see Fixsen et al., 2005; Durlak & DuPre, 2008), issues associated with implementation could potentially be bypassed if, for example, DM or ORC strategies were implemented directly through a state's free QL. Other state QLs could feasibly obtain a highly targeted mailing list (e.g., Medicaid, Medicare) and administer the DM strategy through their ongoing operations. Additionally, ORC could potentially be disseminated to health care settings as an alternative to fax-referral strategies (see Willett et al., 2009). Finally, financial incentives were important for both ORC and DM, and future dissemination would require this core component. However, the current analysis was not a randomized controlled trial designed to test intervention components. Components of the current program were demonstrated to be evidence-based approaches for affecting smoking cessation, and the purpose of the current project was to assess the implementation and effectiveness of two incentive-based strategies for recruiting low-income individuals to smoking cessation services in order to address the Affordable Care Act's dictum to scale evidence-
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
based interventions to population-level practice and increase smoking abstinence rates in low-income populations (see Blumenthal et al., 2013; Spoth et al., 2013). Future practice-based research should consider other experimental and quasi-experimental designs, such as a factorial research design (see Collins et al., 2014), to test how each component of the current intervention contributed to its success. Another potential limitation of the study was that the sample was disproportionately female as a result of the overarching programmatic goals of the NBCCEDP. Even though it is important to target low-income female smokers because their smoking rates are persistently high (Stewart et al., 2010), the current results could change if the program were implemented within an all-male sample or a sample with a more equal distribution of males and females. 5. Conclusion Disadvantaged populations have limited access to, and tend to underutilize, quality health care and preventive services (Kassler et al., 2015; Adler & Newman, 2002). Preventive interventions and population-based programs that extend preventive care and tobacco cessation services to underserved populations are pressing public health priorities (Kassler et al., 2015; Blumenthal et al., 2013). The near ubiquity of telephone technology provides an avenue for reaching underserved individuals. Moreover, incentive-based interventions can be successfully scaled to population-level practice, and the two recruitment strategies employed in the current program can successfully target populations disproportionately affected by the tobacco epidemic by extending the reach of state QLs. Potentially other programs, such as other states' NBCCEDPs, Medicaid, and QLs, can target low-income and inadequately insured individuals for referral to smoking cessation services through the use of financial incentives and these recruitment strategies, considering the strengths and weaknesses of both approaches. Disclosures Project funded through Centers for Disease Control and Prevention (American Recovery and Reinvestment Act; Patient Protection and Affordable Care Act); grant FOA DP09-90101SUPP10. The Centers for Disease Control and Prevention had no role in the design, collection, analysis, and interpretation of the data, or the writing of the manuscript and decision to submit for publication. Jon O. Ebbert has received funding from Pfizer and Orexigen and personal fees from GlaxoSmithKline outside of the current research. The authors declare that they have no conflict of interest, and all authors contributed to and have approved the final manuscript. Acknowledgments We thank QUITPLAN® Helpline staff, Shelly Madigan, Sage patient navigators, Janis Taramelli, and Michelle Waste for their efforts.
References Adler, N.E., & Newman, K. (2002). Socioeconomic disparities in health: Pathways and policies. Health Affairs, 21, 60–76. Ammerman, A., Smith, T., & Calancie, L. (2014). Practice-based evidence in public health: Improving reach, relevance, and results. Annual Review of Public Health, 35, 47–63. Asfar, T., Ebbert, J.O., Klesges, R.C., & Relyea, G.E. (2011). Do smoking reduction interventions promote cessation in smokers not ready to quit? Addictive Behaviors, 36, 764–768. Bashshur, R.L., Shannon, G.W., Smith, B.R., Alverson, D.C., Antoniotti, N., ... Yellowlees, P. (2014). The empirical foundations of telemedicine interventions for chronic disease management. Telemedicine and e-Health, 20, 769–800. Becker, M.H. (1974). The health belief model and personal health behavior. Health Education Monographs, 2, 324–473. Blumenthal, K.J., Saulsgiver, K.A., Norton, L., Toxel, A.B., Anarell, J.P., ... Volpp, K.G. (2013). Medicaid incentive programs to encourage healthy behavior show mixed results to date and should be studied and improved. Health Affairs, 32(3), 497–507. Bryant, J., Bonevski, B., Paul, C., McElduff, P., & Attia, J. (2011). A systematic review and meta-analysis of the effectiveness of behavioural smoking cessation interventions in selected disadvantaged groups. Addiction, 106, 1568–1585.
113
Burns, E.K., Deaton, E.A., & Levinson, A.H. (2011). Rates and reasons: Disparities in low intentions to use a state smoking cessation quitline. American Journal of Health Promotion, 25, S59–S65. Centers for Disease Control and Prevention (2014). Current cigarette smoking among adults—United States, 2005–2012. MMWR, 63(2), 29–34. Collins, L.M., Dziak, J.J., Kugler, K.C., & Trail, J.B. (2014). Factorial experiments: Efficient tools for evaluation of intervention components. American Journal of Preventive Medicine, 47, 498–504. DiClement, C.C., Prochaska, J.O., Fairhurst, S.K., Velicer, W.F., Velasquez, M.M., & Rossi, J.S. (1991). The process of smoking cessation: An analysis of precontemplation, contemplation, and preparation of stages of change. Journal of Consulting and Clinical Psychology, 59, 295–304. Durlak, J.A., & DuPre, E.P. (2008). Implementation matters: A review of research on the influence of implementation on program outcomes and the factors affecting implementation. American Journal of Community Psychology, 41, 327–350. Fiore, M.C., et al. (2008). Treating tobacco use and dependence: Clinical practice guidelines 2008 update. Washington, DC: US Department of Health and Human Services. Fixsen, D.L., Naoom, S.F., Blasé, K.A., Friedman, R.M., & Wallace, F. (2005). Implementation research: A synthesis of the literature. Tampa, FL: University of South Florida, Louis de la Parte Florida Mental Health Institute, The National Implementation Research Network (FMHI Publication #231). Freund, K.M., Battaglia, T.A., Calhoun, E., Dudley, D.J., Fiscella, K., ... The Patient Navigation Research Program Group (2008). National Cancer Institute Patient Navigation Research Program: Methods, protocol, and measures. Cancer, 113(12), 3391–3399. Glasgow, R.E., Vogt, T.M., & Boles, S.M. (1999). Evaluating the public health impact of health promotion interventions: The RE-AIM framework. American Journal of Public Health, 89, 1322–1327. Green, L.W. (2008). Making research relevant: If it is an evidence-based practice, where's the practice-based evidence? Family Practice, 25(S1), i20–i24. Jha, P., Peto, R., Zatonski, W., Boreham, J., Jarvis, M.J., & Lopez, A.D. (2006). Social inequalities in male mortality, and in male mortality from smoking: Indirect estimation from national death rates in England and Wales, Poland, and North America. Lancet, 368, 367–370. Kassler, W.J., Tomoyasu, N., & Conway, P.H. (2015). Beyond a traditional payer—CMS's role in improving population health. New England Journal of Medicine, 372, 109–111. Lee, N.C., Wong, F.L., Jamison, P.M., Jones, S.F., Galaska, L., ... Stokes-Townsend, G. -A. (2014). Implementation of the National Breast and Cervical Cancer Early Detection Program: The beginning. Cancer, 120(16), 2540–2548. Lewis, S. (2010). Creating incentives to improve population health. Preventing Chronic Disease, 7, A94. Lichtenstein, E., Glasglow, R.E., Lando, H.A., Ossip-Klein, D.J., & Boles, S.M. (1996). Telephone counseling for smoking cessation: Rationales and meta-analytic review of evidence. Health Education Research, 11, 243–257. Mathew, A.R., Burris, J.L., Alberg, A.J., Cummings, K.M., & Carpenter, M.J. (2014). Impact of a brief telephone referral on quitline use, quit attempts and abstinence. Health Education Research, Advance Online Publication. McAfee, T., Davis, K.C., Alexander, R.L., Pechacek, T.F., & Bunnell, R. (2013). Effect of the first federally funded US antismoking national media campaign. Lancet, 382, 14–20. Consortium, N.A.Q. (2011). Guides on quitlines and research. Retrieved from http://c. ymcdn.com/sites/www.naquitline.org/resource/resmgr/research/ researchguideapril2013.pdf. Parks, M.J., Slater, J.S., Rothman, A.J., & Nelson, C.L. (2015). Interpersonal communication and smoking cessation in the context of an incentive-based program: Survey evidence from a telehealth intervention in a low-income population. Journal of Health Communication, Advance Online Publication. http://dx.doi.org/10.1080/10810730. 2015.1039677. Patten, C.A., Smith, C.M., Brockman, T.A., Decker, P.A., Hughes, C.A., ... Zhu, S. -H. (2011). Support-person promotion of a smoking quitline: A randomized control trial. American Journal of Preventive Medicine, 41(1), 17–23. Prochaska, J.O., Redding, C.A., & Evers, K.E. (2008). The transtheoretical model and stages of change. In K. Glanz, B.K. Rimer, & K. Viswanath (Eds.), Health behavior and health education: Theory, research, and practice (pp. 97–121) (4th ed.). San Francisco, CA: Jossey-Bass. Rollnick, S., Miller, W.R., & Butler, C.C. (2008). Motivational interviewing in health care: Helping patients change behavior. New York, NY: The Guilford Press. Rothman, A.J., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin, 121(1), 3–19. Rothschild, M.L., & Gaidis, W.C. (1981). Behavioral learning theory: Its relevance to marketing and promotion. Journal of Marketing, 13, 263–269. Sanson-Fisher, R.W., Bonevski, B., Green, L.W., & D'Este, C. (2007). Limitations of the randomized controlled trial in evaluating population-based health interventions. American Journal of Preventive Medicine, 33(2), 155–161. Schuck, K., Bricker, J.B., Otten, R., Kleinjan, M., Brandon, T.H., & Engels, R.C.M.E. (2014). Effectiveness of proactive quitline counselling for smoking parents recruited through primary schools: Results of a randomized controlled trial. Addiction, 109, 830–841. Sigmon, S.C., & Patrick, M.E. (2012). The use of financial incentives in promoting smoking cessation. Preventive Medicine, 55, S24–S32. Slater, J.S., Henly, G.A., Ha, C.N., Malone, M.E., Nyman, J.A., ... McGovern, P.G. (2005). Effect of direct mail as a population-based strategy to increase mammography use among low-income under-insured women ages 40 to 64 years. Cancer Epidemiology, Biomarkers & Prevention, 14(10), 2346–2352. Soet, J.E., & Basch, C.E. (1997). The telephone as a communication medium for health education. Health Education & Behavior, 24, 759–772. Solomon, L.J., Marcy, T.W., Howe, K.D., Skelly, J.M., Reinier, K., & Flynn, B.S. (2005). Does extended proactive telephone support increase smoking cessation among lowincome women using nicotine patches? Preventive Medicine, 40, 306–313.
114
J.S. Slater et al. / Addictive Behaviors 52 (2016) 108–114
Spoth, R., Rohrbach, L.A., Greenberg, M., Leaf, P., Brown, H., ... Society for Prevention Research Type 2 Translational Task Force Members and Contributing Authors (2013). Addressing core challenges for the next generation of type 2 translational research and systems: The Translation Science to Population Impact (TSci Impact) framework. Prevention Science, 14, 319–351. Stead, L.F., Hartmann-Boyce, J., Perera, R., & Lancaster, T. (2013). Telephone counselling for smoking cessation. Cochrane Database of Systematic Reviews, 8, CD002850. Stead, L.F., Perera, R., & Lancaster, T. (2007). A systematic review of interventions for smokers who contact quitlines. Tobacco Control, 16, i3–i8. Stewart, M.J., Kushner, K.E., Greaves, L., Letourneau, N., Spitzer, D., & Boscoe, M. (2010). Impacts of a support intervention for low-income women who smoke. Social Science & Medicine, 71, 1901–1909. Thomas, S., Fayter, D., Misso, K., Ogilvie, D., Petticrew, M., ... Worthy, G. (2008). Population tobacco control interventions and their effects on social inequalities in smoking: A systematic review. Tobacco Control, 17, 230–237. U.S. Census Bureau (2012). State and county quickfacts: Minnesota. http://quickfacts. census.gov/qfd/states/27000.html Vidrine, J.I., Shete, S., Cao, Y., Greisinger, A., Harmonson, P., ... Wetter, D.W. (2013). Ask-advise-connect: A new approach to smoking treatment delivery in health care settings. Journal of the American Medical Association Internal Medicine, 173, 458–464. Volpp, K.G., Troxel, A.B., Pauly, M.V., Glick, H.A., Puig, A., ... Audrain-McGovern, J. (2009). Randomized, controlled trial of financial incentives for smoking cessation. New England Journal of Medicine, 360, 699–709.
Weinstein, N.D., Sandman, P.M., & Blalock, S.J. (2008). The precaution adoption process model. In K. Glanz, B.K. Rimer, & K. Viswanath (Eds.), Health behavior and health education: Theory, research, and practice (pp. 123–147) (4th ed.). San Francisco, CA: Jossey-Bass. Willett, J.G., Hood, N.E., Burns, E.K., Swetlick, J.L., Wilson, S.M., ... Levinson, A.H. (2009). Clinical faxed referrals to a tobacco quitline: Reach, enrollment, and participant characteristics. American Journal of Preventive Medicine, 36, 337–340. Witte, K., & Allen, M. (2000). A meta-analysis of fear appeals: Implications for effective public health campaigns. Health Education & Behavior, 27, 591–615. Witte, K., Meyer, G., & Martell, D. (2001). Effective health risk messages: A step-by-step guide. Thousand Oaks, CA: Sage Publications, Inc. Wootton, R. (2012). Twenty years of telemedicine in chronic disease management—An evidence synthesis. Journal of Telemedicine and Telecare, 18, 211–220. Wootton, R., Jebamani, L.S., & Dow, S.A. (2005). E-health and the universitas 21 organization: 2. Telemedicine and underserved populations. Journal of Telemedicine and Telecare, 11, 221–224. Zhu, S. -H., Gardiner, P., Cummins, S., Anderson, C., Wong, S., ... Gamst, A. (2011). Quitline utilization rates of African-American and white smokers: The California experience. American Journal of Health Promotion, 25, S51–S58. Zhu, S. -H., Lee, M., Zhuang, Y. -L., Gamst, A., & Wolfson, T. (2012). Interventions to increase smoking cessation at the population level: How much progress has been made in the last two decades? Tobacco Control, 21, 110–118.