Oral Health Promotion and Smoking Cessation Program Delivered via Tobacco Quitlines: The Oral Health 4 Life Trial

Abstract

Objectives. To assess the effects of a novel oral health promotion program (Oral Health 4 Life; OH4L) delivered through state-funded tobacco quitlines.

Methods. Using a semipragmatic design to balance experimental control and generalizability, we randomized US quitline callers (n = 718) to standard care or standard care plus OH4L. We followed participants for 6 months to assess effects on professional dental care and smoking abstinence. We collected data between 2015 and 2017.

Results. Participants were racially diverse (42% non-White) and socioeconomically disadvantaged. Most (71%) reported fair or poor oral health, and all were overdue for routine dental care. At 6 months, professional dental care and abstinence did not significantly differ between arms, but abstinence favored the experimental arm and was significantly higher among experimental participants at 2 months in a complete case sensitivity analysis.

Conclusions. OH4L was not effective for promoting dental care, but integrating oral health counseling with quitline counseling may offer some advantage for smoking cessation.

Public Health Implications. We offer a model for conducting semipragmatic trials and partnering with tobacco quitlines to evaluate population-level public health interventions.

Oral diseases such as cavities, gum disease, and oral cancer affect millions in the United States, resulting in unnecessary pain, potential disfigurement, emotional suffering, morbidity, and mortality.¹ The United States spends an average of $113 billion annually on dental care and another $6 billion on lost productivity because of missed work as a result of dental health issues.²

Smokers are at increased risk for oral disease as a result of their tobacco use, other lifestyle risk factors, and low rates of dental care utilization.^3–6 The lack of routine dental care is particularly problematic, because oral health issues then go undiagnosed and untreated. It also prevents smokers from being counseled about appropriate oral hygiene and quitting tobacco. In short, smokers are an important target group for oral health promotion efforts.

To reach smokers who do not routinely seek dental care, we proposed a novel strategy to embed an oral health promotion intervention in the standard tobacco cessation program offered through state-funded quitlines. The rationale for this was fivefold.

• First, smokers calling the quitlines have already made a commitment to behavior change. Therefore, they may be receptive to changing their oral health self-care or seeing a dentist.

• Second, state-funded tobacco quitlines are available nationwide and constitute a well-established, broadly accessible public health infrastructure. In 2016, more than 353 000 smokers received evidence-based treatment through publicly funded state quitlines,⁷ and treatment is provided by trained counselors who can also advise smokers about basic oral health care.

• Third, several common distraction tips for managing smoking cravings (e.g., brush or floss instead of smoking) or tips for motivating sustained tobacco abstinence (e.g., see a dentist to remove tobacco stains from teeth) simultaneously support the goals of quitting smoking and improving oral health care. Thus, addressing both behaviors concurrently could have synergistic effects.

• Fourth, partnering with state quitlines helps reach socioeconomically disadvantaged smokers. Many of these individuals are uninsured or have Medicaid coverage⁸ and, therefore, have limited access to dental care. In addition, many have limited education⁷ or are from racial or ethnic groups known to have higher rates of oral disease and lower oral cancer survival rates.⁹

• Fifth, many state quitline callers do not seek routine dental care,¹⁰ and both quitline callers and quitline providers support offering oral health promotion programs through the quitlines.^10–12

These factors demonstrate both a need and an opportunity to partner with state quitlines to improve oral health care among smokers; however, quitline providers have made it clear that interventions delivered through the quitlines must fit within the existing infrastructure and refer smokers to local low-cost dental providers rather than offering other types of intervention (e.g., treatment vouchers or transportation), which are not sustainable options for the quitlines.¹² Thus, we developed a theoretically grounded, comprehensive, multimodal oral health promotion program (Oral Health 4 Life; OH4L) designed to be integrated into a typical state quitline tobacco cessation program. We hypothesized that smokers receiving the OH4L program would be more likely to seek professional dental care and to quit smoking at long-term follow-up than would smokers who receive the standard quitline program. We report the main outcomes of this trial.

METHODS

We summarize key methodological details. Additional detail can be found in the published protocol.¹³ We enrolled participants between June 2015 and July 2016. We collected the data between June 2015 and March 2017.

The study was a collaboration between the Kaiser Permanente Washington Health Research Institute (KPWHRI); the University of California, Davis; and the Optum Center for Wellbeing Research. KPWHRI and the University of California, Davis designed all study materials. Optum collected and analyzed follow-up data. KPWHRI recruited participants, collected baseline data, and delivered the telephone-based counseling.

Study Design

The study was a 2-arm, randomized, semipragmatic trial. The design balanced tight experimental control (e.g., use of inclusion–exclusion criteria, structured follow-up data collection) with a pragmatic implementation plan in which we used established, usual care quitline systems (e.g., counselors and delivery system infrastructure). The combination of methodological rigor and pragmatic implementation provides greater confidence that study results will generalize to real-world conditions and increases the public health relevance of the findings.

We assessed quitline callers for eligibility (n = 8962) following registration with the Oregon, Nebraska, or Louisiana state quitlines if registration data indicated they smoked at least 5 cigarettes a day, were aged 18 years or older, were ready to quit smoking, could read and speak English, and were eligible for their state’s multicall quitline program. Eligibility criteria were had no dental appointment in the previous or upcoming 6 months, had at least some of their natural teeth, could receive text messages, and had Internet access. We excluded individuals if they were incarcerated, were receiving inpatient substance abuse treatment, had significant cognitive impairment or psychosis, planned to move in the next 6 months, or had a household member enrolled in the study. Additional detail on those excluded is provided in Figure 1. A total of 718 callers were eligible and enrolled in the study.

FIGURE 1—

Study CONSORT Diagram: Oral Health for Life (OH4L) Trial, United States, 2015–2017

^aExcludes people known to be ineligible per quitline intake data (n = 1072).

^bPrimary reasons were no Internet access (n = 2283), saw dentist in previous 6 months (n = 1051), lost all natural teeth (n = 693), previous diagnosis with psychosis (n = 373), no text-messaging capacity (n = 351), unwilling to discuss oral health (n = 303), dental appointment already scheduled (n = 233), and planning to move in next 6 months (n = 170).

^cIndividuals ineligible and mistakenly randomized. Immediately removed from sample, not offered treatment, and not followed for data collection. Reasons for ineligibility were not mutually exclusive and included enrolled in another study already (n = 1), lived with an enrolled participant (n = 1), did not own cell phone or could not receive text messages (n = 2), already quit smoking (n = 4), and not eligible per state’s change to eligibility criteria for multicall quitline program (n = 12).

^dLost to follow-up includes the following reasons: unable to be reached, refused, deceased, or too ill to participate at contact.

Eligible callers provided oral consent, completed a baseline interview, were randomized using an automated algorithm built into the quitline provider’s systems, and then initiated treatment. We stratified randomization by participants’ dental insurance (yes vs no or unsure) and quitline (Louisiana, Nebraska, or Oregon).

Intervention

We randomized participants to the standard quitline program plus generic health promotion text messages (control arm) or the standard care quitline program plus multimodal OH4L intervention (experimental arm). Each intervention is detailed in the published protocol.¹³ Briefly, the standard multicall quitline program consisted of either 4 or 5 planned calls (depending on the state), a mailed smoking cessation guide, and access to online cessation content. The program was grounded in social cognitive theory and cognitive behavioral therapy and focused on the standard counseling components demonstrated effective for tobacco cessation: problem solving, coping skills training, and support.¹⁴ We encouraged callers to set a target quit date during their first contact and then assisted them in quitting smoking and remaining abstinent during subsequent calls. As appropriate for each person, counselors used problem solving to help smokers better manage their mood and affect as it related to their ability to manage cravings and quit smoking. The standard program has been shown to be effective.^15–17

In addition to the standard behavioral intervention, each state offered an initial 2- or 4-week starter kit of nicotine replacement therapy. Medication access was determined by individual medical appropriateness and available state resources to provide the medication, regardless of study arm.

Participants in each arm also received 16 attention-matched text messages. Among controls, messages included general health behavior tips (e.g., recommendations for diet, physical activity) but not oral health messaging. In the experimental arm, tips focused on improving one’s oral health.

Experimental participants also received scripted oral health counseling integrated into each quitline call; a mailed oral health promotion brochure and access to similar content online; and a toothbrush, floss, and sugar-free gum. Oral health content included recommendations on how to manage tobacco cravings by engaging in oral health behaviors (e.g., brushing, flossing, and chewing xylitol gum), referral information for local low-cost dental providers, and discussion of the benefits of proper oral health care. A team of behavioral scientists and oral health professionals developed the content, designing it for persons with low literacy. Access to the online OH4L content required a log-in ID and password. This allowed us to track use and prevented controls from accessing the site. To further prevent treatment contamination, different quitline counselors treated control and experimental participants.

Counselor Training and Fidelity Monitoring

Experimental arm counselors attended a 2-hour training during which they were taught to follow the detailed OH4L intervention protocol and practiced integrating the oral health counseling into standard cessation calls. To ensure fidelity to the scripting for each of the counseling sessions, fidelity coders reviewed audio recordings of the initial calls of each session (call sessions 1–5) for each experimental counselor plus a random sample of all calls. This resulted in 450 experimental calls (44%) being reviewed and coded for fidelity in the experimental arm. In the control arm, we reviewed and coded a random sample of all call recordings (n = 130; 13%). Coding in both arms followed a standard checklist to assess exposure to standard cessation advice, OH4L scripting, or other oral health advice.

Assessment

We monitored intervention contact with the quitline (number, duration, and content of intervention calls) and use of the OH4L Web site with automated data. We assessed self-reported survey items at baseline and 2 months and 6 months after enrollment. Survey assessors were blinded to treatment condition.

Baseline assessment included participant demographics, cigarettes per day, self-reported oral health and gum disease assessed using standardized items from the 2011 to 2012 National Health and Nutrition Examination Survey,¹⁸ and motivation and self-efficacy for quitting smoking and seeing a dentist in the next 6 months. We scored the last items on a 5-point Likert scale ranging from “not at all” to “very.” We assessed depression history using the 2 hallmark diagnostic criteria for clinical depression: ever having a period of 2 weeks or more during which (1) one felt down, depressed, or hopeless, or (2) one had little interest or pleasure in doing things.¹⁹

The primary outcome measures were point prevalent smoking abstinence (self-report of no smoking, even a puff, in the last 7 days) at 6 months and receipt of professional dental care in the last 6 months at follow-up. We assessed receipt of dental care with a self-report item from the 2012 National Health Interview Survey, which assessed time since last seeing a dentist (including orthodontists, oral surgeons, and dental hygienists).²⁰ We defined utilization as a report of seeing a dental provider in the last 6 months at either the 2- or 6-month assessment. Because previous use in the last 6 months at baseline was an exclusion criterion, we considered any dental visits reported during study follow-up postenrollment services. To discourage misreporting, we told participants at the beginning of each survey that we may contact their dental provider to verify their receipt of services if they reported seeing a dentist. We collected contact information but did not contact providers.

Analyses

Analyses conformed to the detailed plan outlined in McClure et al.¹³ We used descriptive statistics to characterize the sample, treatment utilization, and treatment fidelity. We used logistic regression to quantify group differences for the 2 main outcomes and to estimate odds ratios (ORs). For the smoking cessation outcome, the model included data from both follow-up time points to improve power; we addressed correlation within individuals using generalized estimating equations. Models adjusted for prespecified baseline covariates and stratification variables. We have presented covariates with the outcome results, which we chose on the basis of those known or expected to affect outcomes. In all analyses, we used an intent-to-treat approach with participants classified on the basis of their assigned arm, regardless of treatment exposure. In the primary analysis, we imputed the outcome of nonrespondents as smokers and as not having had a dental visit, and we included all randomized participants in the analysis. With a sample of n = 718, the study had 80% power to detect a difference of 5.6% in dental care utilization and 7.6% in smoking abstinence.

We also conducted 2 sensitivity analyses using different approaches to handle missing data: (1) included only participants with observed outcome data (complete case analysis), and (2) used multiple imputation methods to impute missing outcomes. We carried out multiple imputation of the primary outcomes among those with missing outcome values on the basis of Rubin’s method.²¹ We then estimated treatment effects by averaging across 5 multiple imputation data sets and accounting for certainty introduced by the imputed outcomes. The number of missing cases at each follow-up is presented in the CONSORT diagram (Figure 1).

The original study sample included an additional 19 individuals who we randomized in error and immediately removed from the study sample without them receiving treatment or providing follow-up data. We also analyzed main outcomes including these individuals; however, the results were unchanged so they are not presented.

RESULTS

Participant flow is noted in Figure 1. Baseline participant characteristics are presented in Table 1. The sample (n = 718) was racially diverse (41.6% non-White or mixed race), 86.5% had an annual household income less than $40 000, 15.1% had a college degree, 46.8% were employed, and 32.7% had dental insurance. Participants smoked an average of 19 cigarettes per day. On average, they had moderate to high levels of motivation and confidence both to quit smoking and to see a dentist in the next 6 months. Most had not seen a dentist (80.2%) or had not had their teeth cleaned (86.8%) in more than a year. Seventy-one percent rated their oral health as “fair” or “poor.”

TABLE 1—

Baseline Demographics of Oral Health 4 Life (OH4L) Study Participants: United States, 2015–2016

Demographic	Control (n = 360), Mean ±SD or No. (%)	Experimental (n = 358), Mean ±SD or No. (%)	All (n = 718), Mean ±SD or No. (%)
Age, y	44.5 ±12.2	44.1 ±12.2	44.3 ±12.2
Cigarettes per day	18.9 ±9.5	19.3 ±9.7	19.1 ±9.7
Motivation to see dentist in next 6 moa	4.4 ±1.1	4.4 ±1.1	4.4 ±1.1
Motivation to stop smokinga	4.9 ±0.5	4.9 ±0.4	4.9 ±0.4
Self-efficacy for seeing dentist in next 6 mo	4.0 ±1.3	3.9 ±1.4	4.0 ±1.4
Self-efficacy for quitting smokinga	4.5 ±0.9	4.5 ±0.9	4.5 ±0.9
Female	223 (62.1)	220 (61.5)	443 (61.8)
Hispanic or Latino	5 (1.4)	16 (4.5)	21 (3.0)
Race
White	208 (58.1)	209 (58.7)	417 (58.4)
Black	113 (31.6)	95 (26.7)	208 (29.1)
Other or multiracial	37 (10.3)	52 (14.6)	89 (12.5)
Dental insurance	118 (32.8)	117 (32.7)	235 (32.7)
Married or living as married	144 (40.2)	154 (43.5)	298 (41.9)
Currently employed	168 (47.3)	165 (46.2)	333 (46.8)
Disabled (employed and unemployed)	99 (27.6)	101 (28.6)	200 (28.1)
History of depression	215 (59.7)	211 (58.9)	426 (59.3)
Income, $
< 20 000	218 (62.6)	212 (62.0)	430 (62.3)
20 000 to < 40 000	85 (24.4)	82 (24.0)	167 (24.2)
40 000 to < 60 000	30 (8.6)	31 (9.1)	61 (8.8)
≥ 60 000	15 (4.3)	17 (5.0)	32 (4.6)
Education
< high school degree	59 (16.7)	64 (18.1)	123 (17.4)
GED or high school degree	135 (38.1)	131 (37.1)	266 (37.6)
Some college, technical, or trade school	115 (32.0)	96 (26.8)	201 (29.4)
≥ college degree	45 (12.7)	62 (17.6)	107 (15.1)
Last dental cleaning > 1 y ago or never	312 (86.9)	307 (86.7)	619 (86.8)
Last dental visit > 1 y ago or never	294 (81.7)	282 (78.8)	576 (80.2)
Oral health self-rating
Excellent	4 (1.1)	1 (0.3)	5 (0.7)
Very good	24 (6.8)	28 (7.9)	52 (7.3)
Good	75 (21.1)	75 (21.1)	150 (21.1)
Fair	131 (36.9)	137 (38.6)	268 (37.8)
Poor	121 (34.1)	114 (32.1)	235 (33.1)

Note. GED = general equivalency diploma.

^aMeasured on 5-point Likert scale from “not at all” (0) to “very” (5).

Dental Care Utilization and Smoking Abstinence

Use of professional dental services did not differ between groups in the primary analysis (18.6% of controls vs 18.4% of experimental participants; adjusted OR [AOR] = 1.03; confidence interval [CI] = 0.69, 1.53; Table 2). Outcomes were unchanged when we excluded nonrespondents (AOR = 1.09; CI = 0.71, 1.66) or imputed missing values using multiple imputation methods (AOR = 1.05; CI = 0.69, 1.59).

TABLE 2—

Dental Care Utilization and Smoking Abstinence Among Oral Health for Life (OH4L) Participants: United States, 2015–2017

	Total, No. (%)	Control, No. (%)	Experimental, No. (%)	AOR (95% CI)a
Dental care in last 6 mo (yes)b	132 (18.4)	67 (18.6)	65 (18.2)	1.03 (0.69, 1.53)
Smoking abstinencec
7-d PPA at 2 mo (yes)	243 (33.8)	116 (32.2)	127 (35.5)	1.22 (0.89, 1.69)
7-d PPA at 6 mo (yes)	230 (32.0)	109 (30.3)	121 (33.8)	1.29 (0.93, 1.78)

Note. AOR = adjusted odds ratio; CI = confidence interval; PPA = point prevalent abstinence.

^aDental care OR adjusted for baseline sex, age, depression history, motivation, self-efficacy, state quitline, and dental insurance. Smoking abstinence OR adjusted for baseline sex, age, cigarettes per day, depression history, motivation, self-efficacy, current use of a stop-smoking medication, state quitline, and dental insurance.

^bMissing values imputed as not having seen dental provider (n = 718).

^cMissing values imputed as smoking (n = 718).

Among participants who saw a dental provider at 2 months, most (90.0% of controls [n = 27] and 81.3% in the experimental arm [n = 26]) offered contact information for their provider. A similar proportion provided contact information at 6-month follow-up (89.7% [n = 52] and 92.6% [n = 50]) in the control and experimental arms, respectively.

Smoking abstinence rates were not significantly different at 6 months in the primary analyses (30.3% vs 33.8%; AOR = 1.29; 95% CI = 0.93, 1.78; Table 2) or when missing values were accounted for using multiple imputation (AOR = 1.26; 95% CI = 0.91, 1.74). Abstinence rates were significantly higher in the experimental arm at 2 months in the complete case analysis (39.1% vs 46.4%; AOR = 1.42; 95% CI = 1.01, 2.00; P = .04) and favored the experimental arm at 6 months (40.8% vs 43.8%; AOR = 1.37; 95% CI = 0.95, 1.96; P = .09).

Intervention Utilization and Fidelity Monitoring

Participants in each arm completed a similar number of quitline counseling calls (control: mean = 2.8; SD = 1.5; experimental: mean = 2.9; SD = 1.5). Most experimental participants viewed the mailed oral health booklet (n = 246; 68.7%), but few accessed the online content (n = 35; 9.8%). Only a few people opted out of receiving study text messages (9 control; 7 experimental).

Fidelity to the quitline counseling protocol was very good in the calls sampled for review. Counselors routinely discussed strategies for quitting smoking (94.4% experimental calls; 96.9% control calls), setting a target quit date (94.2% experimental calls; 99.2% control calls), and use of pharmacotherapy (91.3% of experimental calls; 97.7% of control calls). Among experimental participants, the OH4L protocol was read as scripted in 76.9% of cases, mostly followed (1 or more scripted talking points missed) in 15.6% of cases, and not delivered in 7.6% of cases. Most experimental participants were provided referral information for local, low-cost dental clinics (84.0%), advised to have their teeth cleaned (71.1%), and advised of the benefits of professional dental care (92.4%). Only 1 control (0.8%) received any oral health advice, and none received the scripted OH4L counseling.

DISCUSSION

The OH4L intervention did not significantly improve dental care or smoking cessation at 6-month rates relative to standard quitline care among socioeconomically disadvantaged callers to the state tobacco quitlines. This is important because tobacco quitline stakeholders (service providers and purchasers) previously expressed interest in an intervention like OH4L, provided it could effectively increase dental care without negatively affecting cessation rates.¹² As it turned out, integrating the oral health counseling into the standard tobacco quitline program did not have a negative impact on cessation and may have had a positive effect on people’s ability to quit smoking at 2 months on the basis of our sensitivity analyses, but there was no differential effect on dental care utilization.

We conducted the study with methodological rigor that prevented treatment contamination across arms, ensured the OH4L intervention was delivered with high fidelity to the protocol, and encouraged accurate self-report of professional dental care. Moreover, actual quitline counselors delivered the intervention using the standard care infrastructure, and we recruited all participants from regular quitline callers. Together these design choices enhance our confidence that the results reflect real-world outcomes. As a result, we have confidence that the comprehensive OH4L program is not an effective means of increasing professional dental care among state quitline callers and does not warrant future dissemination for this purpose.

Anecdotal evidence from reviewing the counseling call audio recordings suggests that many people were interested in receiving professional dental care but had concerns about cost and access to treatment. We sought to address these issues by informing callers about low-cost dental providers in their local area and providing contact information for these clinics; however, providing this information was not enough to overcome these barriers, even with motivational support and encouragement. It is beyond the scope of this study, but in future analyses we will explore the extent to which the intervention influenced secondary oral health behaviors (brushing, flossing) or potential mediators of oral health behavior change (motivation, confidence, knowledge). These analyses will indicate whether the intervention shows promise as a means for promoting better oral health self-care as opposed to professional dental care.

It is notable that smokers receiving the OH4L intervention had a 29% to 37% increased odds of quitting smoking at 6 months compared with people offered the standard quitline program and an absolute difference in quit rates of 3.5% to 6.1% (depending on the approach used to account for missing data). Although not statistically significant, these differences are a potentially promising signal that pro–oral health counseling and tips (e.g., brushing one’s teeth or chewing xylitol gum to help manage cravings) may benefit smokers who are trying to quit. Future research should explore how counselors can more effectively offer oral health promotion tips to support cessation efforts. If similar effects could be achieved with a more minimal quitline intervention, it could be a cost-effective strategy for boosting smoking abstinence rates on a population level.

Finally, the study is noteworthy because it demonstrated the operational feasibility of working with state-funded quitlines to promote healthy behaviors concurrent with smoking cessation. Only 3 other studies to our knowledge have attempted this, 2 focused on reducing weight gain when quitting smoking^22,23 and 1 on alcohol.²⁴ State tobacco quitlines provide services to hundreds of thousands of smokers annually, including a high proportion of low-income smokers. Considering this wide reach, tobacco quitlines offer untapped potential to deliver important public health interventions to these high-risk, high-need smokers. Others interested in nationwide public health interventions should consider partnering with tobacco quitlines.

Strengths and Limitations

In addition to the strengths we have noted, our study had several other notable strengths and limitations. For example, the study relied on self-reported smoking and dental care. Self-report is recommended best practice when assessing smoking status in large, geographically dispersed, randomized trials that involve no face-to-face contact and when missing values are imputed as smokers.²⁵ Requiring biochemical confirmation of smoking status is not recommended because it can lead to missing data unrelated to smoking and bias study results. Thus, our use of self-reported abstinence is a strength, even though self-report is often a study limitation. We applied the same logic to our dental care outcome but attempted to minimize misreporting by asking respondents for their provider’s contact information and permission to verify self-reported treatment use. Because we did not actually contact providers, we cannot confirm the veracity of participants’ self-report, but most respondents who reported seeing a dental provider did provide contact information. Thus, we believe our use of this modified bogus pipeline strategy^13,26 helped mitigate the inherent limitations of using self-reported dental care as an outcome.

Requiring participants to log in to view online content was also a study strength and limitation. This feature allowed us to monitor treatment exposure and prevent contamination in the control arm but may have discouraged some participants from viewing the Web site.

Next, our focus on socioeconomically disadvantaged smokers is a strength because of the significant oral health needs of this population, but the fact that so many participants were low income and had no dental insurance likely limited their ability to obtain dental care. As a result of our focus on this target demographic group, we cannot draw conclusions about how effective OH4L might be if offered through commercially funded quitlines to smokers with greater economic resources or higher rates of dental insurance. In our previous research we found that insured, commercial quitline callers also have unmet oral health needs,^11,27 so future research should explore whether OH4L or similar interventions would be more effective if targeted to smokers with greater access to dental care.

Conclusions

The OH4L intervention showed promising effects on cessation but was not an effective strategy for increasing the utilization of professional dental care among state quitline callers. This conclusion is supported by the methodological rigor of this semipragmatic, randomized trial. The findings point to several issues that warrant further research, including the use of pro–oral health tips to encourage cessation and the effectiveness of oral health promotion programs targeting less economically disadvantaged smokers. Finally, this trial can serve as a model to others interested in partnering with state tobacco quitlines to deliver public health interventions to socioeconomically disadvantaged smokers.