These data imply that patients seen in public health dental clinics are at particular risk of developing tobacco-related health problems, have more oral health problems, and may especially benefit from exposure to a brief tobacco-use cessation program. The CDC estimated that in 1997, $17 billion (12 percent) of Medicaid expenses was spent on the treatment of tobacco-related disease.²

There is a clear need to provide tobacco-use cessation interventions to low-income people, who are more likely to use tobacco and less likely to have access to tobacco-use intervention services. Public health dental clinics provide a useful channel through which to reach this underserved population.

We based our study on a line of research conducted during the past 15 years in both health maintenance organization (HMO) and private dental practice settings.^3–6 This research showed that a brief intervention (including provision of advice to quit, written materials and a targeted self-help video to take home) delivered during the patient’s cleaning visit produced significant increases in quitting smokeless tobacco use in comparison with quit rates among patients who received usual care. The intervention was less effective for smokers than for users of smokeless tobacco.^3,5,6

The study presented here was conducted in two public health dental clinics in Oregon. The majority of patients served by public health clinics in Oregon are covered under the Oregon Health Plan (OHP), a state Medicaid program. According to a 1999 OHP tobacco-use survey, approximately 44 percent of enrolled patients were smokers, users of smokeless tobacco or both compared with 22 percent of adults in Oregon’s general population.

	METHOD

TOP ABSTRACT METHOD RESULTS DISCUSSION CONCLUSION AND CLINICAL... REFERENCES

 
Recruitment of public health clinics and dental practitioner training. We recruited two public health dental clinics to participate in the study. Both clinics are grantees in the U.S. Health Resources and Services Administration’s Community Health Center Program.⁷ One clinic was located in the inner city of Portland, Ore., and the other was located in Eugene, Ore., a smaller, less urban community. The Eugene clinic serves a more transient (that is, homeless and migrant worker) population.

Recruitment of practitioners and clinics presented challenges. In one clinic, there was no championship of or buy-in to the study (owing to leadership change within that clinic), and the clinic staff generally was not enthusiastic about participating in the project.

In the second clinic, although the practitioners and staff members were excited about participating, the administrators expressed concern about the cost of lost time owed to training and intervention activities. We found it necessary to provide a stipend in both clinics to allow the clinicians and staff members to attend the study training during work time.

Design. We employed a quasiexperimental time-series design in which all patients seen in the first clinic for three months before we trained the dental care providers in tobacco-use intervention received usual care and, thus, served as the control group. We classified as "intervention patients" those who were seen in this same clinic for three months subsequent to provider training and implementation of the protocol. In a second clinic, we trained all practitioners before they saw patients. Therefore, all patients enrolled in the second clinic were in the intervention group.

Intervention components. We based the intervention used in the study on the "five As" recommended in the U.S. Public Health Service’s Clinical Practice Guidelines (Fiore and colleagues⁸): Ask, Advise, Assess, Assist and Arrange (Box). We adapted the intervention for use in this study by breaking the Advise component into two parts:

– relate oral health findings to the patient;

– give direct advice to quit tobacco use.

In addition, we enhanced the "Assist" component to include assistance in setting a quit date, as well as provision of self-help materials and free nicotine patches or gum (for patients who set a quit date). We supplied all materials, including nicotine gum and patches, because tobacco-use cessation products were not covered by Medicaid in Oregon at the time we conducted the study. In addition, we trained all practitioners in brief motivational interviewing techniques for health care workers.⁹ Finally, for this study, we trained all practitioners and allied professionals (that is, dentists, dental hygienists, dental assistants and denturists) and encouraged them to provide the intervention to all patients seeking nonemergent care who used tobacco (that is, at all nonemergent-care dental visits, not just at dental hygiene visits). Delivery of the intervention was intended to be brief, ranging from three to five minutes per patient.

Practitioner training. Eight dentists, five dental hygienists, 10 dental assistants, two dental students, a denturist and five office staff members attended a three-hour workshop about the study intervention. The training included a description of each of the five As and didactic information on the oral health effects of smoking and smokeless tobacco use, the prescription and use of nicotine replacement therapy, and motivational interviewing techniques. In addition, we showed video vignettes modeling each intended behavior, and we conducted role-playing exercises to help the participants rehearse and reinforce the new behavioral skills.

Ninety-six percent of enrolled patients were current smokers. Seventy-one percent indicated that they were ready to quit smoking.

We held two additional one-hour "minitraining" sessions at each clinic during the study period to reinforce the intervention protocol, answer questions, solve problems related to specific issues and train new staff members.

Participant enrollment and description. We enrolled 364 patients who used tobacco, all of whose income was at or below the federal poverty level at the time of their dental health visit. We encouraged front-office staff members and practitioners to distribute enrollment forms to all patients. However, the front-office staff members at the first clinic complied only poorly with this request. Therefore, we placed enrollment signs in the waiting rooms at the clinics encouraging patients who used tobacco to speak to the front-office staff members to enroll in the program. A range of 44 to 86 percent of eligible patients enrolled in the study across both sites.

The mean age of the participants was 38.6 years (standard deviation [SD] = 10.5), and 58 percent were men. Sixty-eight percent of the patients were white, 12.5 percent African-American, 5.2 percent Hispanic/Latino and 3.2 percent American Indian; 10.8 percent listed their race/ethnicity as "other." Most of the patients were single (65 percent) and had graduated from high school (81 percent).

Patients were similar across the two clinics in all variables except age (t[333] = 7.08; P < .001) and race/ethnicity (² [1, n = 364] = 10.08, P < .01). Participants in the Portland clinic were older and more likely to be black than participants in the Eugene clinic. Ninety-six percent of enrolled patients were current smokers. They had been smokers for an average of 21 years and had smoked an average of 15 cigarettes per day (SD = 7.7). Seventy-one percent indicated that they were ready to quit smoking.

Intervention group patients did not differ from control group patients in terms of sex, marital status, education, tobacco-related oral health effects, number of cigarettes smoked at baseline, readiness to quit smoking, or the proportion who smoked or used smokeless tobacco. However, patients in the usual care group reported significantly higher use of tobacco in the first 30 minutes after waking (72 percent) as compared with the intervention group patients (58 percent; ² [2, n = 360] – 9.23, P < .01). Patients who received usual care also were older (mean = 41.0 versus 36.2 years; t[341] = 4.33, P < .001) and less likely to be white.

We compared this sample of public health dental patients with private practice dental patients enrolled in a tobacco-use cessation study.^5,6 Patients in the two types of practices differed significantly on a number of demographic variables. Enrolled patients in the public health clinics were more likely to be male (58.0 versus 42.6 percent; ² [1, n = 4,495], P < .001), less likely to be white (68.5 versus 92.2 percent; ² [1, n = 4,434] = 214.68), more likely to be single (70.8 versus 35.6 percent; ² [1, n = 4,435] = 166.92), more likely to be younger (mean age = 38.6 versus 42.2 years; ² [1, n = 4,377]), and were less likely to have graduated from high school (78.4 versus 90.0 percent; ² [1, n = 4,468] = 46.58, P < .001). Although patients in the public health clinics were slightly more ready to quit smoking, we found no differences between patients in the two types of practices in terms of tobacco-related variables, including the average number of cigarettes smoked in a given day.

Measures. Baseline measures. Variables measured at enrollment included demographic variables, tobacco-related oral health effects, number of cigarettes smoked at baseline, readiness to quit smoking (as measured by Biener and Abrams’¹⁰ contemplation ladder), time to first cigarette after waking¹¹ and use of smokeless tobacco.

Outcome measures. We considered patients to have quit smoking if they reported that they "had not smoked at all, even a puff, in the last seven days." Outcomes included three- and six-month point prevalence cessation and repeated point prevalence cessation, at both three and six months.

We based the results of this study on self-reported, not biochemically validated, quitting. However, most research suggests that a person’s self-report is a valid measure of abstinence; furthermore, biochemical validation was impractical in this study owing to the high mobility of the sample.¹²

Provider protocol adherence. At the six-week assessment, patients reported whether the dental provider had, during their dental visit six weeks before, talked to them about their tobacco-related oral health problems, given them quitting tips, encouraged them to set a quit date, told them about nicotine replacement therapy or told them about the Oregon Tobacco Quit Line (a free telephone service available to all Oregon residents who want to stop using tobacco). Patients also reported whether they had received written materials, a video or both.

Women were significantly more likely to complete the six-month assessment, and those who had a high school education or higher were more likely to complete and return the assessment.

Follow-up assessments of patients. We obtained follow-up data at six weeks and three and six months postenrollment. The protocol for follow-up assessments was this:

– We mailed questionnaires to all participants with an enclosed $2 bill as an incentive for returning the forms. We also entered respondents who returned completed questionnaires into a drawing for $100.

– We mailed a second questionnaire to the participants who had not returned the first questionnaire two weeks after the first mailing.

– We telephoned those who did not respond to the second mailing, and we administered the questionnaire orally.

Two hundred four (56 percent) patients completed the six-week assessment. Of these, 126 (62 percent) responded to the first mailing; 42 (21 percent) responded to the second mailing and 36 (18 percent) completed the assessment by phone. One hundred eighty-six (51 percent) patients completed the three-month assessment; 96 (52 percent) responded to the first mailing and 52 (28 percent) to the second mailing, and 38 (20 percent) completed the assessment by phone. One hundred seventy-one (47 percent) completed the six-month assessment; 83 (49 percent) responded to the first mailing and 45 (26 percent) to the second, and 43 (25 percent) completed the assessment, by phone. Two hundred thirty-four (61 percent) completed either the three- or the six-month assessment, and 134 (36 percent) completed both. As compared with patients at the Eugene clinic, which served a more transient population, a significantly higher proportion of patients in the Portland clinic completed follow-up assessments. Nine patients from both clinics dropped out of the study after enrollment and asked for no further contact (2 percent). One hundred sixteen patients relocated during the six-month follow-up period (32 percent). We could not locate 68 of these patients (or 18 percent of the total sample).

Those who completed the six-month assessment and those who did not complete it were similar in race/ethnicity, marital status, intervention group, age, use of smokeless tobacco, use of cigarettes, number of cigarettes smoked per day at baseline, years smoked, time to first cigarette after waking or readiness to quit smoking. However, women were significantly more likely to complete the six-month assessment (54 versus 40 percent; ² [1, n = 367] = 6.12, P < .05), and those who had a high school education or higher were more likely to complete and return the assessment (49 versus 47 percent; ² [1, n = 366] = 3.84, P < .05).

Statistical analyses. We used logistic regression, controlling for baseline differences between patients who received usual care and patients who received the intervention, to examine the effect of the intervention on patients’ reports of quitting tobacco use. This statistical technique yields an odds ratio, which can be interpreted as the odds of quitting tobacco use if the patient is in the intervention group, as compared with participating in the usual care group. Using ² analyses, we examined the differences in patients’ reports of receiving components of the intervention protocol between patients in usual care and patients in the intervention group and between the two clinics. We used logistic regression to identify which intervention components were related to patient tobacco-use cessation, as well as the odds of quitting tobacco use if the patient received that intervention component. We further examined whether the receipt of these intervention components explained the effect of the intervention on quitting tobacco through conducting another logistic regression and examining if the previously significant effect of the intervention on outcomes became nonsignificant when we controlled for the receipt of the intervention components.

As usually is found in effectiveness trials in health settings in which the intervention is delivered by provider staff, we found adherence to the protocol to be inconsistent.

	RESULTS

TOP ABSTRACT METHOD RESULTS DISCUSSION CONCLUSION AND CLINICAL... REFERENCES

 
Outcomes. Self-reported quitting. The differences in self-reported quitting by condition were significant across all endpoints (Table 1).

Although the probability that those in the intervention group would quit was three to four times higher than that those in the usual care group would quit, owing to the small sample size, the intervention effect was not significant on repeated point prevalence (report of nonuse at both three and six months). There were no differences in quit rates as a function of the intervention site. Those in the Portland clinic were similar to those in the Eugene clinic at all endpoints in terms of quitting.

Patient report of provider adherence to the protocol. To assess provider adherence to the intervention protocol, we used the six-week patient reports regarding whether the dental provider had talked to them about a variety of topics. As usually is found in effectiveness trials in health settings in which the intervention is delivered by provider staff, we found adherence to the protocol to be inconsistent. However, as shown in Table 2, a significantly higher proportion of patients in the intervention group reported that providers delivered the intervention components compared with patients in the control group. Also, the proportion of intervention patients (14 percent) reporting that they had received telephone counseling if they had set a quit date was higher than that of the control patients (3 percent; ² [1, n = 206] = 9.13, P < .01).

Prediction of tobacco-use cessation from receipt of intervention components. We sought to identify specific intervention components that were related to patient outcomes. A series of univariate logistic regressions showed that receipt of information about nicotine replacement therapy (OR = 4.13; 95 percent CI = 1.03, 16.54, P < .05), the receipt of a self-help video (OR = 8.75; 95 percent CI = 2.10, 36.53), the receipt of tobacco-use cessation materials (OR = 4.40; 95 percent CI = 1.04, 18.61, P < .05), encouragement to set a quit date (OR = 4.13; 95 percent CI = 1.03, 16.54, P < .05) and the receipt of information about the Oregon Tobacco Quit Line (OR = 5.17; 95 percent CI = 1.25, 21.46, P < .05) predicted repeated point prevalence of cessation at both three and six months. In another logistic regression, we examined the effect of the intervention on repeated point prevalence of cessation, controlling for patient report of receipt of these intervention components. The results of this analysis showed that the effect of the intervention was due to patients’ receipt of all five of these intervention components.

Conducting tobacco-use cessation interventions in public dental clinics presented unique challenges.

	DISCUSSION

TOP ABSTRACT METHOD RESULTS DISCUSSION CONCLUSION AND CLINICAL... REFERENCES

 
Conducting tobacco-use cessation interventions in public dental clinics presented unique challenges. For example, the patient population was difficult to track for follow-up assessments, the clinic administrators were concerned about the cost of intervention activities and the effect of those activities on the limited time available for service delivery, and it was necessary to offer a stipend to the clinics to guarantee full attendance at all of the training sessions.

Despite these challenges, our results suggest the viability and effectiveness of tobacco-use cessation services delivered to smokers with low income via their dental health care practitioners in a public health setting. This intervention was much more effective than similar ones conducted with smokers in private practice dental settings. Several factors could have influenced our success in helping public health patients quit smoking. The patients seen in this public health study may not have had previous access to cessation resources and assistance. Although they may have been ready to quit for some time, they may have lacked the resources and support to do so. Dental patients with low SES smoke at disproportionately high rates, yet they have the fewest opportunities to avail themselves of treatment. It also is possible that these patients had not been educated previously about the oral health effects of tobacco use by their dentist or dental hygienist.

Furthermore, the public health setting may have been more conducive for delivering the intervention. Perhaps the intervention provided a structure that the public dental health practitioner was more able to incorporate in routine practice. In our previous randomized trials, the dental hygienist was the primary tobacco-use cessation interventionist. In this public health clinic pilot study, dentists, as well as allied professionals, delivered the intervention. Public health dental clinic patients are more likely to receive treatment from a dentist for acute care, rather than visiting the hygienist for routine, preventive care. This gives dentists a unique opportunity to intervene with patients who use tobacco. In addition, public health dentists are committed to providing broad services to their patients who may have more severe oral and general health problems related to their tobacco use.

Limitations. Although the design of this study was an efficient test of the intervention model, it did not control for two threats to internal validity (the extent to which outcome can be confidently attributed to the intervention).¹³ Although not plausible, historical factors other than dental provider implementation of the intervention protocol could have been related to historical changes in smoking behavior among intervention patients. In addition, patients’ responses to the surveys could have varied across time.

Furthermore, the generalizability of our findings from this pilot study is limited because only two clinics in Oregon participated. These two clinics may not be representative of all public dental clinics.

Finally, the high attrition of subjects at follow-up assessments further limited the generalizability of our findings. However, the attrition analyses suggest that those who did not respond did not differ from those who did regarding the variables known to be associated with tobacco-use cessation.

	CONCLUSION AND CLINICAL IMPLICATIONS

TOP ABSTRACT METHOD RESULTS DISCUSSION CONCLUSION AND CLINICAL... REFERENCES

 
The results of our quasiexperimental study suggest the viability and effectiveness of tobacco-use cessation services delivered to low-income users of tobacco via their dental health care practitioners. Plus, the intervention may have been even more effective if all of the practitioners adhered more closely to the intervention protocol. The potential reach and effect of a public dental clinic–based tobacco-use cessation program are great. There are approximately 475 federally funded community or neighborhood health centers with dental clinics and another 200 to 250 Indian Health Service and other community clinics throughout the United States.^14,15 The percentage of patients who use tobacco at these clinics is much higher than the national average. Even with a modest cessation rate such as the one reported herein, the public health impact of a brief tobacco-use cessation intervention would be significant.