The effectiveness of health care–based interventions for smoking cessation has been well-established, as reflected in the evidence-based Public Health Service clinical practice guideline for tobacco use. ⁷ The guideline is based on a five-step treatment model:

– ask patients about tobacco use status at every visit;

– advise all tobacco users to quit;

– assess a patient’s willingness to try to quit;

– assist the patient in quitting (provide cessation treatment or referral);

– arrange follow-up.⁷

The "5 A’s" model now is widely accepted as the standard for tobacco dependence treatment.

Clinicians in both medical and dental practices inconsistently deliver treatment and support to tobacco users, despite evidence of the effectiveness of these interventions, and there is great need to improve the implementation of the tobacco-use cessation guideline in practice.⁸ The starting point for the tobacco-use treatment model is adopting a system to identify all smokers within a clinical practice. Estimates of the rates at which dentists routinely ask patients about their tobacco use range from about 20 to 55 percent.^8–11 Implementation of the steps after identification also is not optimal. Although a representative national survey of oral health care providers in 1994 found that 66 percent advised most smokers who had been identified, less than one-third (29 percent) offered cessation assistance.⁹ The overall conclusion one can draw from the survey data of providers is that dentists and hygienists do not ask their patients about tobacco use routinely, and if they know a patient uses tobacco, they may advise the patient to quit but are unlikely to provide services to help him or her do so.

There has been little research concerning the implementation and effectiveness of clinical practice guidelines for smoking cessation intervention with adolescents. Research on adolescent smoking patterns suggests that identification of adolescent smokers might be particularly difficult, however. An examination of self-reported data in several intervention studies has indicated that adolescents’ reports of their own tobacco use are less likely to be accurate than are adults’ self-reports of tobacco use.¹² This is partially because of underreporting, perhaps motivated by social desirability or fear of legal or parental sanctions, but also because adolescents are likely to be sporadic smokers who do not label themselves as smokers.¹²

Research on adolescent smoking patterns suggests that identification of adolescent smokers might be particularly difficult.

We conducted a study to compare the identification of smoking status and provision of advice about smoking among adolescents during regular dental visits recorded on the health history forms in the patients’ charts to similar information obtained from the adolescents in telephone interviews. We hypothesized that

– adolescents would underreport smoking on the health history form compared with during the telephone interview;

– dentists would provide advice about smoking more often to adolescents identified as tobacco users on their health history forms;

– the rate of smoking cessation advice provided by dental staff members would be low.

In the study, we also examined whether patients’ characteristics and dental hygiene behaviors are related to the provision of cessation advice.

	METHODS AND SUBJECTS

TOP ABSTRACT METHODS AND SUBJECTS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Site. This study was conducted at HealthPartners, a not-for-profit managed care organization in Bloomington, Minn. HealthPartners owns and operates 16 staff-model dental clinics staffed by 60 dentists and 80 dental hygienists. The clinics offer a complete range of services to more than 100,000 enrollees in the Minneapolis/St. Paul metropolitan area. In 1996, the dental division of HealthPartners adopted a guideline to identify patients at risk of developing periodontal disease and to provide interventions to reduce their risk. The protocol calls for the measurement of tobacco use as a part of a health history update completed by patients.

Telephone interview. Subjects and procedures. We selected subjects who were dependents of HealthPartners members, who were between the ages of 14 and 17 years, and who were contacted by HealthPartners interviewers for screening and recruitment for a smoking prevention/cessation trial. They were eligible for the screening interview if they were due for a routine dental hygiene visit at a HealthPartners dental clinic and had scheduled an appointment for the visit, if they spoke English and if both they and their parents consented to their participation. Only one adolescent per household was eligible. A quarterly review of administrative dental records identified all adolescents in the specified age range who were due for dental checkups in the next six months. Approximately two months before the due date, we sent a letter to both the parents and the adolescents explaining the study and asking for consent to contact the adolescent by telephone for an interview. We asked parents to mail an enclosed postcard or call research staff if they did not want their children to participate.

On a weekly basis, we checked computerized records of scheduled dental appointments against the quarterly review to identify adolescents who had received the consent letter and had not actively refused. If we found a match, a trained telephone survey interviewer from the HealthPartners Survey Center contacted the adolescent, explained the study and proceeded with the telephone interview if the adolescent was willing. We made at least four attempts during a variety of periods to reach the adolescent for the telephone interview.

Demographic measures. Demographic characteristics measured at baseline included age, sex, ethnicity and the level of education attained by each parent or guardian. We collapsed the parental education measures into a single measure that indicated whether at least one parent had completed college.

Cigarette use. We asked subjects if they had ever smoked a cigarette in their lives. If they had, they were asked the number of times they had smoked (one time, two to five times, more than five times) and whether they had smoked in the previous 30 days. If they had smoked in the previous 30 days, we asked if they smoked on a daily basis. We used this series of questions to form a single measure of tobacco use with four levels that indicated increasing level of use: never smoked or smoked only once; smoked more than once, but not in the previous 30 days; smoked in the previous 30 days, but not daily; and smoked daily. In the cases of subjects who reported that they had never smoked, we asked three questions about the likelihood that they would start smoking in the future.¹³ If the answers to these questions indicated any inclination to try smoking, we considered him or her a susceptible nonsmoker.

In the cases of subjects who reported that they had never smoked, we asked three questions about the likelihood that they would start smoking in the future.

Dental care. We asked the subjects four questions about dental care. Two questions about personal dental practices asked how often they usually brushed their teeth and flossed. Response options for the brushing question were "never," "less than once per day," "one time per day," "two times per day" and "three or more times per day." Response options for flossing were "never," "one time per week," "two times per week," "three times per week" and "four or more times per week." A third question asked whether it was important to keep the teeth and gums healthy; response options were "very," "somewhat" and "not important." Finally, we asked the subjects about having received smoking advice: "Has a dentist or dental hygienist ever talked with you about smoking?" (response options: "yes" or "no").

Chart audit. Subject selection. We selected for the chart audit a stratified random sample of 300 subjects who had completed the telephone interview. We based strata on the adolescents’ reports of smoking status during the telephone interview; 100 had smoked in the preceding 30 days, 100 were susceptible nonsmokers, and 100 were ex-smokers (had smoked previously, but not in the previous 30 days) or were never-smokers who scored as not being susceptible to starting smoking.

Procedures and measures. Information concerning a patient’s health history is gathered routinely at each visit at HealthPartners dental clinics on a self-completed form given to the patient before he or she sees the dentist or the dental hygienist. For adolescents, either the patient or the patient’s parent might complete the form. The smoking question on the health history form lists tobacco as one of several risk factors for dental or oral disease and asks the patient to check "Yes" or "No" to indicate whether he or she uses it. If "Yes" is chosen, the form prompts the patient to provide information about the type and amount of tobacco used.

We reviewed the dental charts for the subjects selected for the audit for smoking status, as reported in the patient- or parent-completed health history form in the dental chart. The chart auditor simply noted the answer to the smoking question. The reference date for the chart audit was the date on which the screening telephone survey was completed. We selected for review the health history form completed closest in time to the reference date (either before or after this date). We excluded from the chart audit any subjects who did not have a completed health history form dated within six months of the reference data.

Analyses. Agreement between the smoking status as indicated on the health history form and by self-report during the telephone interview consisted of calculating statistics to measure overall agreement and the sensitivity and specificity of information on the health history form when compared with the self-report in the telephone interview. Since both indicators of smoking are self-report measures, we elected to use the telephone interview as the gold standard for the sensitivity-specificity analysis. We chose the interview because we considered underreporting more likely than overreporting in both contexts and because reported smoking rates were higher on the telephone interview. We conducted multivariate analyses using logistic regression techniques. Owing to low numbers in some of the response categories of predictor variables (for example, few subjects reported that they never brushed their teeth), we collapsed some categories before analysis.

	RESULTS

TOP ABSTRACT METHODS AND SUBJECTS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Sample. Telephone interview. We identified 6,031 adolescents as being due for a dental hygiene visit, based on a six-month checkup schedule. A total of 3,845 adolescents of these actually scheduled a dental hygiene visit. However, 18 were duplicates, 52 had siblings already enrolled in the study, and 173 had an appointment date outside the period we had set for the telephone interviews. We completed telephone interviews with 32 percent of the remaining teens (1,162 of 3,602). The majority of those not interviewed declined participation by postcard (n = 985) or telephone (n = 312), or were not reached or available after four attempts (n = 699). Other teens were not eligible because of age or medical condition (n = 34) or because their consent letters were returned as undeliverable (n = 80). In addition, we did not interview some teens (n = 330) because their appointment time duplicated a time slot filled by another teen who had completed an interview, making them ineligible for the smoking cessation trial.

Chart audit. Of the 300 subjects we selected for chart audit, 20 either did not have a health history form or did not have a form completed within six months of the date of the telephone interview. The final chart audit sample consisted of 280 subjects.

Sample characteristics. Table 1 shows the characteristics of the 1,162 subjects who completed the telephone interview and the 280 we selected for chart audit. There were fewer subjects in the older age groups in the interview sample. About 36 percent of the interview sample had tried smoking, and 13.4 percent had smoked in the previous 30 days. The higher smoking rates in the chart audit sample reflect the fact that the sample was selected from the interview sample to comprise one-third of those who had indicated that they had smoked cigarettes in the previous 30 days. Other differences between the two samples also probably are driven by this selection criterion.

Predictors of provider advice. Only 8.9 percent of all subjects interviewed reported that a dentist or a dental hygienist had ever talked with them about smoking. In the chart audit sample (n = 280), however, 25 percent of adolescents who identified themselves as smokers on the health history form reported having received advice about smoking, compared with 9.5 percent of non-smokers (P = .006). These results also were reflected in bivariate and multivariate analyses of the entire sample (Table 3, page 1118), especially if the adolescent reported smoking on a daily basis (odds ratio = 2.62, 95 percent confidence interval 1.35 to 5.10). Other characteristics found to be associated with provider advice included older age and lower levels of parental education.

	DISCUSSION

TOP ABSTRACT METHODS AND SUBJECTS RESULTS DISCUSSION CONCLUSION REFERENCES

Our study examined the validity of one method of screening for adolescent tobacco use and provides evidence about the link between the identification of tobacco use and advice given to adolescents by health care professionals about this risk factor. Our results suggest that when a patient indicates smoking activity on the health history form, dentists are significantly more likely to provide him or her with advice about cessation. The positive association between report of smoking on the health history form and provision of cessation advice by the health care professional suggests that the report of smoking on the health history form serves as a prompt to dental staff members to talk to the patient about smoking. This finding is consistent with well-established findings among adults that including smoking status as a vital sign in medical settings leads to a greater likelihood of providers’ offering cessation advice.⁷

The results also indicate, however, that tobacco use is underreported on the health history form. Comparison of smoking as reported on the health history form and smoking as reported during a telephone interview conducted within six months of the dental visit found that only 38 percent of adolescents who reported having smoked in the previous 30 days and 58 percent of those who reported smoking daily were identified as tobacco users on the health history form. There are several possible reasons for this. It is probable that the wording of the tobacco-use item on the health history form results in lower levels of disclosure. Recent studies indicate that many adolescents who smoke do not consider themselves to be smokers. This is particularly true when they are social or experimental smokers.¹² Adolescent smokers who smoke infrequently and do not label themselves as smokers easily could decide that a question about tobacco use, such as that found on the health history form, does not apply to them. It also is possible that the form was completed by a parent who was unaware that the adolescent smoked, or that the adolescent was unwilling to disclose smoking activity in the dental setting because a parent was present or because he or she was concerned that the parent would find out about the smoking from the dentist. Another possibility is that the adolescent was unwilling to disclose his or her smoking to the dental staff because he or she believed that practitioners would not approve of this behavior. Finally, the result might have been due in part to a methodological limitation of the study—the fact that the dental visit and the phone interview could have occurred up to six months apart—and therefore the difference in self-report on the health history form and in the phone interview could have reflected a real change in behavior. In other words, an adolescent might have started using tobacco in the time between completing the health history form and taking part in the phone interview.

A further striking finding of this study is that, although identification on the health history form as a smoker is related to significantly greater recall of advice from the dentist about smoking cessation among adolescents, the rate of recall of advice was low. Only 25 percent of adolescents identified on the health history form as smokers recalled ever having received advice from dental staff members to quit smoking. This finding suggests that dental team members often fail to attend to smoking status as indicated on the health history form or fail to provide advice to smokers or, if they do provide advice, fail to provide advice that is memorable or effective.

	CONCLUSION

TOP ABSTRACT METHODS AND SUBJECTS RESULTS DISCUSSION CONCLUSION REFERENCES

 
With a caveat regarding the methodological problem of the time interval between the two self-reports of tobacco use, we make the following recommendations on the basis of our study findings:

– Screening questions for identifying tobacco users included in instruments like the health history form should leave little room for interpretation. A standard question used to ascertain smoking activity among adolescents and young adults is whether any tobacco use has occurred in the previous 30 days.¹⁵ This question about actual behavior would be more likely to identify low-rate users than would a general question about tobacco use that leaves it to the adolescent to decide the criteria for tobacco use. Sargent and DiFranza¹⁵ suggested a set of 10 questions for identifying adolescents who are at risk of starting to use tobacco.

– Efforts could be made to ensure that the adolescent, rather than the parent, completes the behavioral section of a health history form and that this be done in privacy. It might be possible, for instance, for the adolescent to complete the form in the dental examination room rather than in the reception room.

– Dental staff should have a clear policy concerning whether the smoking status of adolescent patients is considered confidential information that should not be disclosed to parents. Conveying this policy to adolescent patients might improve accuracy.

– Advice about smoking cessation certainly should be provided to adolescents who are identified as smokers. It also is recommended that preventive advice be provided to adolescents who indicate that they have not smoked in the previous 30 days.

Further research in this area is needed. Important questions involve the best wording and placement of questions used to identify adolescent smokers; increasing the salience of this issue for dental providers; and, most importantly, the effectiveness of efforts of dentists and dental staff members to prevent initiation of tobacco use and encourage its cessation among adolescents.