Abstract
Objectives. We sought to compare the effectiveness of a dental practitioner advice and brief counseling intervention to quit tobacco use versus usual care for patients in community health centers on tobacco cessation, reduction in tobacco use, number of quit attempts, and change in readiness to quit.
Methods. We randomized 14 federally funded community health center dental clinics that serve diverse racial/ethnic groups in 3 states (Mississippi, New York, and Oregon) to the intervention (brief advice and assistance, including nicotine replacement therapy) or usual care group.
Results. We enrolled 2549 smokers. Participants in the intervention group reported significantly higher abstinence rates at the 7.5-month follow-up, for both point prevalence (F1,12 = 6.84; P < .05) and prolonged abstinence (F1,12 = 14.62; P < .01) than did those in the usual care group.
Conclusions. The results of our study suggest the viability and effectiveness of tobacco cessation services delivered to low-income smokers via their dental health care practitioner in community health centers. Tobacco cessation services delivered in public dental clinics have the potential to improve the health and well-being of millions of Americans.
The prevalence of tobacco use is especially high in lower socioeconomic status (SES) populations in the United States.1 Intervening with these tobacco users can be difficult, in part because of their limited access to health care services. Community health centers were implemented as a means to address this concern; many are located in areas with health professional shortages. These health centers provide comprehensive primary care services (with 65% including dental services) to large numbers of low-income smokers and smokeless tobacco users.2 The typical patient visits the health center dental clinic multiple times per year, receiving treatment or prevention services, and it is routine for the patients to be on a periodic examination schedule.2 These dental visits provide an opportunity to offer tobacco cessation advice and counseling by the dental office team.
Organized dentistry has long recognized the connection between tobacco use and oral health and the appropriateness for oral health professionals to address tobacco use.3–6 The American Dental Association urges dentists to assist patients with tobacco cessation.3 The dental team routinely provides educational and preventive services to patients; this model can be extended to tobacco-related topics. With consideration to the oral health effects associated with chronic tobacco use, the dental visit provides a “teachable moment” during which the dental team can relate oral health and systemic problems to tobacco use and provide evidence-based brief interventions to patients who use tobacco.7–9
We recruited 14 federally funded public health dental clinics that serve diverse racial/ethnic groups in 3 states: 6 in Mississippi, and 4 each in New York and Oregon. Following recruitment, clinics were stratified by size and patient ethnicity (proportion of non-Hispanic Whites, Hispanics, and non-Hispanic African Americans), and randomized to 1 of the 2 study conditions.
Participants for this study were dental patients aged 18 years and older who were seen for a nonemergency visit to the clinic. All self-identified tobacco-using patients were asked by the front office staff to complete a brief survey regarding their tobacco use. Those patients who indicated current use of tobacco (in the past 7 days) were offered the opportunity to enroll in the study. Patients who completed the informed consent document and baseline survey were enrolled. Patients who declined to enroll were asked by the front office staff to indicate a reason by checking a box on the provided brief checklist. Enrollment of patients occurred over an approximately 6-month period within each practice, although 4 practices chose to enroll patients for longer than 6 months to achieve their recruitment goal.
We enrolled 2637 tobacco users at their dental health visit. All patients eligible to receive services at these health centers were at or below 200% of the federal poverty threshold as defined by the US Census Bureau (2006–2008). Participant characteristics are provided in Table 1. Ninety-seven percent of participants currently smoked. Patients had smoked for an average of 19.2 years (SD = 13.2) and smoked an average of 16.1 cigarettes per day (SD = 10.4). Seventy percent indicated that they were thinking about quitting in the next 30 days, with a mean score of 6.7 (SD = 2.7) on the readiness-to-quit scale10 (0 = no thought of quitting; 10 = ready to quit now). There were 62 (2.4%) tobacco chewers and 26 (1%) participants who both smoked and chewed tobacco.
TABLE 1 Participant Demographics From 14 Community Health Center Dental Clinics: Mississippi, New York, and Oregon, 2006–2008
| Demographics | Mean (SD) or % (No.) |
| Mean age, y | 40.5 (12.6) |
| % female | 57.2 (1508) |
| Race/ethnicity | |
| Non-Hispanic African American | 45.8 (1207) |
| Non-Hispanic White | 32.2 (849) |
| Hispanic | 15.8 (416) |
| American Indian | 2.8 (73) |
| Asian | 0.3 (7) |
| Hawaiian/Pacific Islander | 0.4 (10) |
| Other/more than 1 race/ethnicity | 1.9 (50) |
| Single marital status | 68.2 (1798) |
| High-school graduate or greater education level | 72.3 (1906) |
As depicted in Figure 1, 48 of the 2751 participants enrolled at baseline were deemed ineligible to participate. The 2 reasons for ineligibility were (1) being younger than the age of consent (< 18 years) and (2) not being a patient at the dental clinic. Sixty-six patients who enrolled in the study subsequently refused further participation, yielding a final sample of 2637 participants eligible for follow up.
FIGURE 1 Participant flow and response rates from 14 community health center dental clinics: Mississippi, New York, and Oregon, 2006–2008.
The 14 community health clinic dental clinics were randomized to either the intervention condition (an intervention based on the 5 “A's” of the Clinical Practice Guideline for Treating Tobacco Use and Dependence)10 including nicotine replacement therapy (nicotine patches and lozenges) for patients setting a quit date or to the usual care group. The Clinical Practice Guideline advocates the use of a brief, office-based 5 “A's” model (Ask, Advise, Assess, Assist, and Arrange) with all tobacco-using patients.10 Clinics were stratified by size and race/ethnicity distribution prior to randomization. Using the RE-AIM (reach, effectiveness, adoption, implementation, and maintenance) framework,11 we compared the effects of dental practitioner advice and brief counseling to quit tobacco with usual care on patients' tobacco cessation, reduction in tobacco use, number of quit attempts, and change in readiness to quit.
Dentists, dental hygienists, and dental assistants in the intervention condition received a 3-hour in-service workshop on integrating and delivering a brief, tailored tobacco intervention to their patients, including proper use of nicotine replacement therapy.
The intervention consisted of (1) Ask—asking all patients about their tobacco use at every visit; (2) Advise—relating the oral effects of tobacco use to the patients' oral health status and advising patients to quit tobacco (a modified version of the Advise component); (3) Assess—assessing readiness to quit via brief verbal questions; (4) Assist—assisting those patients interested in quitting through a facilitated method, including setting a quit date, discussing pharmacotherapy, and providing free self-help materials and free nicotine replacement therapy to all patients setting a quit date; and (5) Arrange—arranging for follow-up by mail or phone for patients setting a quit date.
To facilitate implementation of the Assist component, each practice in the intervention condition was given a supply of nicotine patches and lozenges, as well as printed patient self-help materials. The printed materials were tailored for type of tobacco used (cigarettes or smokeless tobacco) and race/ethnicity (non-Hispanic Whites, non-Hispanic African Americans, or Hispanics). Also included was information on the local tobacco quit line, which providers were asked to give to all tobacco-using patients.
Practitioners in the usual care condition continued with their usual tobacco cessation practice with patients. After completion of the 7.5-month follow-up for all patients in their clinic, they participated in the in-service workshop and received all intervention materials.
The study was conducted from May 2005 to April 2009. Practitioners were enrolled between January 2006 and August 2007, and patients were enrolled from January 2006 through March 2008. Practitioner assessments were conducted between March 2006 and November 2008. Participant assessments began in March 2005, and 6-month follow-up assessments were completed in December 2008. Data analyses began in February 2009.
Participants completed the baseline survey in the clinic at enrollment, and were mailed follow-up surveys at 6-weeks and 7.5-months postenrollment (6-months postenrollment plus a 6-week grace period). If a follow-up survey was not returned within 2 weeks, a phone call was made in an attempt to complete the assessment.
This survey included questions assessing demographics as well as tobacco-use history. The tobacco use items included: (1) number of cigarettes smoked per day; (2) readiness to quit tobacco, assessed with the question “Are you seriously thinking about quitting smoking in the next 30 days?” and via a contemplation ladder10; (3) number of previous quit attempts; and (4) level of nicotine dependence as measured by 3 items13: “How many years have you smoked?”; “How soon after waking do you usually smoke your first cigarette?”; and “Do you experience strong cravings to smoke when you have to go more than 2 hours without a cigarette?”
At the 6-week and 7.5-month assessment, both point prevalence and prolonged abstinence were assessed. Point prevalence of tobacco use was measured by the questions, “Have you smoked, even a puff, in the last 7 days?” and “Have you used smokeless tobacco, even one dip, in the last 7 days?” Participants were considered abstainers if they reported not having used tobacco in the past 7 days. Prolonged abstinence was measured by the question “When did you last smoke a cigarette?” or “When did you last use chewing tobacco or snuff?” Response categories ranged from “less than 1 month ago” to “more than 6 months ago.” Self-reported abstinence was not biochemically verified because of the population-based nature of this trial,14 and because of reductions in the project budget at the time of initial funding.
Participants who reported quitting were asked about their use of quitting strategies, resources, materials, and medications. At the 6-week assessment, all participants were asked about their receipt of intervention components, including discussion of pharmacotherapy and the tobacco quit line. For participants who continued to use tobacco, we assessed secondary outcomes, including number of cigarettes smoked per day, number of quit attempts, level of readiness to quit, and level of tobacco dependence.
We collected 6-week outcome data from 85% (n = 2240) of eligible participants. At 7.5 months, we collected data from 71% (n = 1875) of participants. Differences were found in return rates between the 2 conditions, with usual care participants returning the survey at a slightly higher rate than did those in the intervention condition (73.9% vs 69.3%; χ21, 2549 = 6.67; P < .01). Women were more likely than were men to return surveys (75.4% vs 66.1%; χ21, 2549 = 26.65; P < .001). Compared with nonrespondents, respondents were older (41.67 years vs 37.64 years; t2540 = −7.64; P < .001), had smoked longer (19.98 years vs 16.52 years; t2507 = −5.94; P < .001), and had a higher level of education (t2486 = −2.89; P < .01). In addition, those patients who were not Hispanic (73.6% vs 59.5%; χ21, 2549 = 33.36; P < .001) were more likely to complete a 7.5-month assessment than those who were Hispanic.
Because of the extremely low number of tobacco chewers and number of those who both smoked and chewed tobacco, we elected to perform analyses solely on those who only smoked tobacco. Therefore, the final sample size on which analyses were conducted was 2549.
Analyses were conducted for both patient and practitioner to assess differences by condition at baseline. Patients in the intervention condition were more likely to be single (χ21, 2549 = 6.56; P < .01), older (t1, 2549 = 3.43; P < .001), to have smoked for more years (t1, 2549 = 2.44; P < .05), to have higher levels of craving (t1, 2549 = 2.58; P < .01), to be thinking about quitting (χ21, 2549 = 1.03; P < .001), and to be more ready to quit (t1, 2549 = 8.23; P < .001). There were no differences at baseline for practitioners by condition.
Our primary outcome was prolonged abstinence at the 7.5-month assessment (6 months postenrollment plus a 6-week grace period). We also report on point-prevalence cessation, as defined by no tobacco use in the past 7 days, at the 7.5 month assessment. Because only a very small proportion of participants chewed tobacco (2.4%) or smoked and chewed tobacco products (1%), we report the self-reported quitting rates for the entire sample and for each of the 3 racial/ethnic groups within the sample only for those participants who smoked tobacco (Table 2). We used multiple imputation procedures to impute missing data.15,16 For all smokers, the intraclass correlations (ICCs) for both 7.5-month point prevalence and prolonged abstinence were 0.009. An analysis within racial/ethnic group showed an ICC of 0.015 for point prevalence and 0.008 for prolonged abstinence for Hispanics, and an ICC of 0.014 for both outcomes for non-Hispanic African Americans. For non-Hispanic Whites, the ICC was 0.014 for point prevalence and 0.003 for prolonged abstinence.
TABLE 2 Self-Reported Quit Rates at 7.5-Months Follow-Up Among Participants From 14 Community Health Center Dental Clinics (n = 2549): Mississippi, New York, and Oregon, 2006–2008
| Entire Sample | Non-Hispanic African Americans | Non-Hispanic Whites | Hispanics | |||||
| Outcome | Usual Care (n = 1155), % | Intervention (n = 1394), % | Usual Care (n = 511), % | Intervention (n = 656), % | Usual Care (n = 390), % | Intervention (n = 431), % | Usual Care (n = 185), % | Intervention (n = 217),a % |
| Point prevalenceb | 6.8 | 11.3* | 7.7 | 12.2* | 4.9 | 10.7** | 7.6 | 7.4 |
| Prolonged abstinencec | 1.9 | 5.3** | 2.0 | 6.5*** | 2.3 | 4.6* | 1.1 | 3.2 |
a Not significant for point prevalence and prolonged abstintence.
b No tobacco use in past 7 days.
c No tobacco use 6-months postenrollment plus a 6-week grace period.
*P < .05; **P < .01; ***P < .001.
Using a generalized linear model (taking into account a dichotomized outcome variable), we conducted analysis of variance with clinics as a random, nested factor within condition and patients nested within clinic for both outcomes, for all participants, and within each racial/ethnic group. After we controlled for differences at baseline, participants in the intervention condition reported significantly higher abstinence rates at the 7.5-month follow-up, for both point prevalence (F1, 12 = 6.84; P < .05) and prolonged abstinence (F1, 12 = 14.62; P < .01).
As shown in Table 2, the intervention was effective for non-Hispanic African Americans and non-Hispanic Whites, but not for Hispanics. Because there were fewer Hispanic participants, we evaluated the effect sizes across racial/ethnic groups to determine whether a decreased sample size could explain the lack of significance of the intervention for Hispanics. However, this analysis replicated our findings, showing a difference in effect sizes as a function of race/ethnicity and that Hispanics were less likely to quit smoking as a result of the intervention.
Analyses were conducted to determine the effect of the intervention by geographic location. Because only 2 clinics within each condition were located in Oregon and New York, we ignored the design effect, and analyzed outcomes by using a χ2 analysis. The 7.5-month point prevalence and prolonged abstinence rates by state are displayed in Table 3. For 7.5-month point prevalence, participants in Oregon and Mississippi in the intervention condition quit at significantly higher rates than did those in usual care. Across all 3 locations, prolonged abstinence rates were higher for intervention condition participants than for usual care participants.
TABLE 3 Self-Reported Quit Rates by State at 7.5-Months Follow-Up Among Participants From 14 Community Health Center Dental Clinics: Mississippi, New York, and Oregon, 2006–2008
| Mississippi | New York | Oregon | ||||
| Outcome | Usual Care (n = 534), % | Intervention (n = 381), % | Usual Care (n = 340), % | Intervention (n = 770), % | Usual Care (n = 281), % | Intervention (n = 243), % |
| Point prevalencea | 6.2 | 11.8** | 10.0 | 10.5 | 3.9 | 13.2** |
| Prolonged abstinenceb | 2.1 | 6.3** | 2.1 | 4.9* | 1.4 | 4.9* |
a No tobacco use in past 7 days.
b No tobacco use 6-months postenrollment plus a 6-week grace period.
*P < .05; **P < .01.
For those patients who did not quit using tobacco products, we examined several secondary outcomes. Controlling for baseline differences between conditions, we used a generalized linear mixed model to examine the effect of the intervention on whether these patients made a quit attempt, a change in amount smoked, and a change in readiness to quit. We found no significant differences on any of these secondary outcomes as a function of condition.
We also assessed the effect of the intervention on these secondary variables by racial/ethnic group. Among those non-Hispanic African Americans who did not quit, those in the intervention condition were more likely to make a quit attempt and to increase their readiness to quit than were participants in usual care. There were no significant differences for the other racial/ethnic groups.
We used logistic regression to examine the effect of baseline measures of tobacco use history on prolonged abstinence. Condition was included in the model as a covariate. The predictors included age, gender, marital status, education, cigarettes smoked per day, time to first cigarette in the morning, thinking about quitting, level of readiness to quit, cravings, and number of serious quit attempts during the past year. We included the interaction of condition with each of the predictors in the model, and we used backward elimination to remove nonsignificant interactions. No interactions were significant. After removal of all interactions and nonsignificant main effects, only thinking about quitting significantly predicted prolonged abstinence (odds ratio [OR] = 2.74; 95% confidence interval [CI] = 1.44, 5.19; P < .01).
Among patients within the intervention condition, 66.5% reported receiving the reading materials, and the majority of these patients reported reading them (96.7%); 16.9% reported using nicotine replacement therapy, and 10.9% reported receiving quit line counseling. On the basis of the criteria outlined by Baron and Kenny,17 we evaluated whether these 3 components of the intervention condition (using nicotine replacement therapy, receiving reading materials, and calling the quit line) mediated the effect of the intervention condition on our primary outcome, prolonged abstinence, measured at 7.5 months. As noted earlier, the effect of condition on prolonged abstinence was significant (OR = 3.19; 95% CI = 1.93, 5.27). In addition, each of the 3 components varied significantly as a function of the intervention condition (using nicotine replacement therapy: OR = 10.99; 95% CI = 6.73, 17.94; receiving reading materials: OR = 2.12; 95% CI = 1.78, 2.51; calling the quit line: OR = 3.24; 95% CI = 2.22, 4.74). Thus, 2 of the 3 criteria for mediation were met.
Although in univariate analyses each of these 3 variables predicted prolonged abstinence (using nicotine replacement therapy: OR = 3.80; 95% CI = 2.36, 6.13; receiving reading materials: OR = 1.67; 95% CI = 1.08, 2.59; calling the quit line: OR = 2.69; 95% CI = 1.53, 4.72), in multivariate analyses, with all 3 predictors in the model, only using nicotine replacement therapy (OR = 3.26; 95% CI = 1.99, 5.32) and calling the quit line (OR = 2.04; 95% CI = 1.14, 3.66) predicted prolonged abstinence. Thus, these 2 variables met Baron and Kenny's criteria for mediation. With these 2 variables in the model, the effect of the intervention condition was reduced slightly but remained a significant predictor of prolonged abstinence (OR = 2.39; 95% CI = 1.41, 4.05), suggesting that using nicotine replacement therapy and calling the quit line partially mediated the effect of the intervention on prolonged abstinence. The Sobel test confirmed an indirect effect of the intervention through using nicotine replacement therapy (4.23; P < .001) and calling the quit line (2.28; P < .05). The indirect effect of receiving reading materials was not significant.
Our results replicate those of our previous research18–20 and our pilot work in public dental clinics.21 The intervention appears to have been effective at facilitating cessation among low-income dental patients who are interested in quitting their use of tobacco products. These patients are at increased risk for a variety of health problems.22 Improving access to an effective intervention for tobacco use would have a broad impact on public health.
The intervention appears to be particularly effective for non-Hispanic African American smokers. We believe this was attributable, in part, to the fact that the recruitment flyers and self-help materials were culturally tailored for African Americans and shown to be effective in previous studies. Historically, this population has been less likely to benefit from smoking cessation interventions delivered in health care settings,23 and African Americans are at increased risk for developing cancer, heart disease, and stroke compared with non-Hispanic Whites.1 An intervention such as this, delivered routinely in public dental clinics, could yield a substantial reduction in smoking-related diseases in this underserved population.
Unfortunately, the intervention did not appear to be efficacious for Hispanics. One possible explanation for this is that, although the materials were translated into Spanish, they were not specifically culturally tailored, which may have limited their effectiveness. Second, the vast majority of our Hispanic participants were recruited from clinics located in New York City, where few participants from other ethnic groups were recruited, thus confounding Hispanic ethnicity with location. During this study, New York instituted a series of tax increases on cigarettes, such that in 2008, the tax on a pack of cigarettes was $5.00, the highest in the United States. The prevalence of smoking in New York City dropped from 18.4% in 2006 to 15.8% in 2008. In addition to the usual decrease in smoking observed in conjunction with price increases,24,25 the revenue garnered from this tax was routed into cessation resources to smokers, including a large media campaign promoting the tobacco quit line and nicotine patch “giveaways.” This likely contributed to the unusually high rate of quitting among usual care patients observed in our study, thereby affecting the relative impact of the intervention. Our intervention provided the same tobacco cessation resources (e.g., referral to the quit line and free nicotine replacement therapy) as did New York State.
This study had 2 major limitations. First, although we were interested in examining differences as a function of race/ethnicity, there were a limited number of Hispanics in the sample. Second, race/ethnicity was confounded with location. As noted previously, in New York, the state that included the highest proportion of Hispanics, the public health department initiated a large quit-smoking campaign. Because race/ethnicity varied by clinic and geographical location, the findings regarding the effects of the intervention by race/ethnicity must be interpreted cautiously.
Our results support the viability and effectiveness of tobacco cessation services delivered to low-income smokers via their dental health care practitioner in a public health setting. The results of this study confirm that public health dental practitioners are committed to providing broad services to their patients, who may suffer from more severe oral and general health problems, partially because of tobacco use. Tobacco cessation services delivered in public dental clinics have the potential to improve the health and well-being of millions of Americans.
Acknowledgments
This research was funded by a grant from the National Institutes of Health, National Cancer Institute (grant R01-CA107442).
The authors gratefully acknowledge the following community dental clinics for graciously participating in this research project: Copiah Comprehensive Health Center, Hazelhurst, MS; G. A. Carmichael Clinic, Canton, MS; G. A. Carmichael Clinic, Yazoo City, MS; Greater Meridian Clinic, Meridian, MS; Hattiesburg Community Clinic, Hattiesburg, MS; Jackson Hinds Comprehensive Health Center, Jackson, MS; Jacobi Medical Center, Bronx, NY; Montefiore Medical Center, Bronx, NY; North General Hospital, New York, NY; Klamath Open Door Clinic, Klamath Falls, OR; Siskiyou Community Health Center, Grants Pass, OR; Sydenham/Grant Neighborhood Family Care Center, New York, NY; Umpqua Community Health Center, Roseburg, OR; and White Bird Clinic, Eugene, OR.
Our special appreciation goes to our research staff—Maureen Barckley, Josephina Batista, Berry Broadbent, Katie Clawson, Barbara Eisenhardt, Debbra Hunter, Julia Martin, Katie Newman, Erin Peterson, and Christopher Widdop—for their outstanding contributions to this research study and article.
Human Participant Protection
This study was approved by the institutional review boards of the Oregon Research Institute, Columbia University Medical Center, University of Mississippi Medical Center, Albert Einstein College of Medicine of Yeshiva University, Montefiore Medical Center, North General Hospital, and Harlem Hospital Center.
