Inqr-43-01-02 54.65

Sara B. McMenamin Effectiveness of Different Jeffrey RideoutGifford Boyce-Smith Benefit Designs forTreating TobaccoDependence: Resultsfrom a Randomized Trial This research estimated the costs and effectiveness of three different benefit designs fortreating tobacco dependence: drugs only (nicotine replacement therapy patch, nasalspray, inhaler, and Zyban); drugs and counseling (drugs and proactive telephonecounseling); and drugs if counseling (drugs conditional on enrollment in counseling).
A sample of 393 adult smokers enrolled in a California preferred provider organizationwas randomly assigned to one of three study groups. After eight months, there were nosignificant increases in quit attempts or quit rates in the groups with covered drugs andcounseling compared to the group with drug coverage only. Therefore, costs rose with noincrease in quit rates when proactive telephone counseling was added to coverage ofpharmacotherapy, regardless of benefit design.
Tobacco use has been identified as the single, $8.6 billion in direct medical costs and $7.3 bil- largest preventable cause of morbidity and mor- lion in lost productivity from illness and prema- tality in the United States; it leads to an estimated ture death (Max et al. 2004).
440,000 deaths and more than five million years Recent clinical practice guidelines identify ef- of life lost each year (DHHS 2004). This places fective treatments for tobacco use and dependence a huge burden on society, with an estimated total and recommend that all patients who use tobacco cost of $158 billion per year (DHHS 2004). This be offered at least one of these treatments to help includes both the direct costs to the health care them stop smoking (Fiore et al. 1996, 2000). The system from the treatment of tobacco-related ill- 2000 U.S. Public Health Service (PHS) guideline ness (estimated at $82 billion per year), as well recommends two types of treatment for tobacco as the societal costs resulting from lost productiv- dependence: pharmacotherapy and counseling ity of workers due to premature death and disease services. The recommended first-line pharmaco- (estimated at $76 billion per year). In California, therapy treatments include bupropion SR (Zyban) there are approximately 43,000 tobacco-related and all Food and Drug Administration (FDA) deaths each year, representing 535,000 years of approved nicotine replacement therapy (NRT) life lost (Max et al. 2004). This translates into including gum, patch, nasal spray, inhaler, and Helen Ann Halpin, Ph.D., is a professor of health policy and director, and Sara B. McMenamin, Ph.D., is an assistantresearcher, both at the Center for Health and Public Policy Studies, School of Public Health, University of California, Ber-keley. Jeffrey Rideout, M.D., is vice president of Cisco Systems, San Jose´, and former medical director of Blue Shield ofCalifornia. Gifford Boyce-Smith, M.D., is director of quality at Blue Shield of California, San Francisco. This researchwas funded by a grant (no. 9RT-0096) from the Tobacco-Related Disease Research Program. Address correspondence toProf. Halpin at Center for Health and Public Policy Studies, University of California, Berkeley, 140 Warren Hall, #7360,Berkeley, CA 94720-7360. Email: [email protected] Inquiry 43: 54–65 (Spring 2006). Ó 2006 Excellus Health Plan, Inc.
Treating Tobacco Dependence lozenges. Effective counseling services identified ferent benefit designs (coverage for pharmaco- in the guideline include proactive telephone coun- therapy only, pharmacotherapy and counseling, seling, individual face-to-face counseling, and group counseling services. In addition, the PHS participation in counseling) with the same cost- guideline recommends that all health insurers sharing requirements; 2) to determine whether provide coverage for these effective treatments.
adding health insurance coverage for counseling Health insurance coverage of recommended to coverage for pharmacotherapy changes smok- tobacco dependence treatments has been demon- ing cessation behaviors and outcomes; 3) to deter- strated to increase quit attempts, use of treatments, mine whether the specific design of the counseling and quit rates (Curry, Grothaus, and McAfee and pharmacotherapy benefit (i.e., pharmacother- 1998; Schauffler et al. 2001a). However, the evi- apy either conditional on enrollment in counseling dence reviewed in the PHS guideline examines programs or not) affects smoking cessation behav- only the effectiveness of each drug or counseling iors among adult smokers; and 4) to estimate the service independently and provides no evidence relative costs of achieving the major quitting out- of the effect of combined pharmacotherapy and comes for each benefit design. The goal of the counseling or guidance on preferred benefit study was to learn how best to design a package of tobacco dependence treatments for a group of While there is widespread agreement in the insured adult smokers who are motivated to try field of tobacco control that health insurance should cover effective treatments for tobacco de-pendence, the optimal benefit design for covering these services remains uncertain. In 1999, a survey was conducted of all California health mainte-nance organizations (HMOs) to document the ben- This case study was an eight-month randomized efit designs for tobacco dependence treatments trial, conducted from May 1, 2001, through De- offered in their standard employer group health cember 31, 2001, that compared three different plan benefit packages (Schauffler et al. 2001b).
health insurance benefit designs for tobacco de- This study found that 31% of HMOs offered cov- pendence treatments. The three treatment groups erage for both pharmacotherapy and counseling, were designed based on three benefit designs 15% offered coverage for pharmacotherapy only, found in California HMOs (pharmacotherapy 15% offered coverage for counseling only, and only, pharmacotherapy and counseling, and phar- 38% offered coverage for pharmacotherapy con- macotherapy conditional on counseling) (Schauf- ditional on enrollment in a counseling program fler et al. 2001b). According to the findings of the (Schauffler et al. 2001b). This mix of benefit de- 2000 PHS clinical practice guideline, there is no signs offered in California HMOs raises questions statistically significant difference in the effective- regarding which design is optimal to most effec- ness of individual face-to-face, group, and pro- tively and efficiently aid smoking cessation in an active telephone counseling (Fiore et al. 2000).
enrolled adult population.
Proactive telephone counseling was chosen as Coverage for pharmacotherapy is more preva- the method of counseling for this study because lent than coverage for behavioral programs to treat it has been shown to have the highest participa- tobacco dependence in both the public and private tion rates compared to group counseling and sectors (McPhillips-Tangum et al. 2004; Halpin, face-to-face counseling, it is the most accessible McMenamin, and Keeler 2004). Therefore, two of the three counseling formats, and it is the for- of the major questions facing both public and pri- mat of counseling most covered by health plans vate policymakers are: 1) Is the effect of coverage in the United States (McPhillips-Tangum et al.
for pharmacotherapy treatments enhanced if 2004). Rates of participation in group counseling counseling services are also covered? 2) What is can be as low as 1% or less, while rates of partic- the effect of restricting coverage for pharmaco- ipation in telephone counseling can be as high as therapy to smokers who enroll in counseling? To 10% to 25% (Schauffler et al. 2001a). In addition, this end, this research has four main objectives: the physicians in the preferred provider organiza- 1) to document changes in quitting behaviors tion (PPO) in the study have not participated in among smokers in three treatment groups with dif- any organized training for smoking cessation, Inquiry/Volume 43, Spring 2006 and most physician offices do not have dedicated groups. The control group (drugs only, n ¼ personnel to perform such counseling.
126) received tobacco dependence treatment cov- While there is a considerable body of literature erage for pharmacotherapy only; the pharmaco- that demonstrates the efficacy of proactive tele- therapy included coverage for Zyban and NRT phone counseling compared to no interventions patch, inhaler, and nasal spray.1 The second on quit attempts and quit rates (Fiore et al.
group (drugs and counseling, n ¼ 140) received 2000; Stead, Lancaster, and Perera 2003), there tobacco dependence treatment coverage for the is a growing body of literature evaluating the aforementioned pharmacotherapy in addition to effectiveness of proactive telephone counseling coverage for proactive telephone counseling pro- as an adjunct to pharmacotherapy compared to vided through a nationally recognized program.
pharmacotherapy alone (Stead, Lancaster, and The third group (drugs if counseling, n ¼ 127) re- Perera 2003; Ockene et al. 1991; Lando et al.
ceived tobacco dependence treatment coverage 1997; Reid, Pipe and Dafoe 1999; Solomon et al.
for proactive telephone counseling and coverage 2000). Four trials have not found any additional for pharmacotherapy only if they enrolled in the effect of telephone counseling for those who telephone counseling program. The proportion use nicotine replacement therapy (Stead, Lancas- of subjects lost to follow-up at eight months ter, and Perera 2003), but none of these studies was 18%; the proportion completing the study specifically evaluated the effect of different ben- was 82%. However, all multivariate analyses in- efit designs on smoking cessation.
cluded 100% of the original sample in each study To recruit participants for this study, informa- group using an intent-to-treat model. All research tional postcards were mailed to all enrollees in was conducted with prior approval from, and in the individual and family plans of a large pre- accordance with the guidelines set forth by, the ferred provider organization operating in Califor- Committee for the Protection of Human Subjects nia (n ¼ 113,000 PPO enrollees). This postcard at the University of California, Berkeley.
invited smokers to participate in a research study Study participants were under no obligation on smoking cessation benefits, requiring them to use any of the tobacco dependence treatments to complete telephone interviews and providing covered under the study, and they were free to them with access to free or low-cost smoking ces- use these benefits just as they would any other sation methods (the pharmacotherapy under all covered services. Access to pharmacotherapy three benefit designs required a $15 co-pay, while benefits required a prescription. Enrollment in the counseling did not require cost sharing). At the proactive telephone counseling program re- the time of the study, this PPO offered cessation quired calling a toll-free number to register.
coverage for pharmacotherapy for enrollees in During the first call, the phone specialist used the group market, but did not provide coverage a formal protocol to assess the smoker's nicotine for any smoking cessation benefits to its individ- dependence, readiness to quit, motivation to use ual and family plan members.
behavioral techniques, and self-efficacy regard- Those who responded to the mailing (n ¼ 803) ing cessation.2 A quitting plan was developed were contacted by telephone to determine their and telephone follow-up calls were scheduled eligibility and, if eligible, to conduct the baseline for one year (during which the participant re- telephone survey. Those who were eligible and ceived four additional calls). If the participant completed the baseline interview (71%) were relapsed, s/he was sent a ‘‘recycle kit'' and the mailed a packet including a complete description specialist re-initiated the quitting process.
of the study, a self-help smoking cessation kit, in- Baseline and eight-month follow-up data on formation on any risks they might face as partic- smoking and quitting behaviors for all partici- ipants in the study, and two copies of a written pants were collected by telephone using a com- consent form (one for their records and one to re- puter-administered telephone interview (CATI) turn to the study). Figure 1 illustrates the disposi- system. Participants received $5 for each tele- tion of the sample from initial contact through phone interview that they completed. Reminder the eight-month follow-up.
postcards were sent at the midpoint of the study Study participants who completed the baseline to remind participants of their eligible benefits interview and returned written consent forms and to inform them that they had four more (69%) were randomized into three treatment months of eligibility. The follow-up was limited Treating Tobacco Dependence Figure 1. Flow diagram of tobacco dependence treatment randomization (*Recruitmentpackets mailed to all individual and family members in a large commercial PPO; packetsindicated that only smokers should respond to mailing. **Five participants were removed fromthe analysis because they obtained pharmacotherapy without enrolling in counseling) to an eight-month time frame, given the timing letter also explained the smoking cessation bene- of the start of the study and the practice of the fit options with special attention to the drug ben- health insurer to begin each enrollment year on efit. The letter required nothing specific of the January 1. Ideally, we would have liked to have physicians and was intended only to make them conducted the study for one full year and to aware of the study and the new benefits offered have collected outcome data one year post- under the study.
All physicians enrolled in the PPO network were sent a letter by the health plan informing Study Participants them about the study in general and the possibil- Eligibility for the study required being an adult ity that they may have patients participating. The 18 years of age or older, currently enrolled in Inquiry/Volume 43, Spring 2006 an individual or family (not group) plan in the utilization of the pharmacotherapy benefits and participating PPO, and a current smoker who participation in the proactive telephone counsel- had smoked at least one cigarette in the last seven ing program.
days. Respondents were not eligible to participate Data on quitting behaviors were collected by in the study if they had any of the following dis- self-report in the follow-up telephone survey.
qualifying health conditions: pregnancy, poor Data on filled prescriptions were provided by health, coronary artery disease, heart disease, ar- the pharmacy manager of the participating PPO rhythmia, heart attack or myocardial infarction, to confirm self-reported use of pharmacotherapy cardiovascular disease, angina pectoris, and con- benefits for each participant and for estimating gestive heart failure.
the cost of pharmacotherapy coverage for each Thus, the results from this case study are only group. Data confirming participant enrollment generalizable to smokers enrolled in individual in the covered proactive telephone counseling and family PPOs in the private health insurance program were provided by a contact at that pro- market; their applicability to smokers enrolled gram and used for estimating the cost of tele- in other types of health plans is not known. We phone counseling coverage for each group.
found several differences between our study par-ticipants and a population-based data set of Statistical Analysis smokers insured in HMOs and PPOs in Califor-nia.3 First, the smokers in our study were more For all demographic and control variables, pro- likely to be female and more likely to be white portions were estimated for the full sample and than the population of California smokers. And for each study group. The full sample included while the smokers in our study were more likely those respondents who were lost to follow-up be- to report wanting to stop smoking and contem- tween the baseline and follow-up surveys. Using plating quitting in the next month, they were ac- an intent-to-treat model, these respondents were tually less likely to have made a quit attempt in assumed not to have made a quit attempt, not to the last year and were no more likely to be plan- have quit during the study period, and not to have ning on quitting in the next month compared prevalent abstinence at eight months. Chi-square to all California smokers. We do not know the tests were performed to estimate whether there extent to which these differences are the result were any statistically significant differences in of differences in smokers enrolled in individual the characteristics of the three study groups at and family plan PPOs compared to other types baseline. Bivariate analysis using the chi-square of managed care plans, or the result of differences test also was conducted to analyze the associa- between those who volunteered to be part of a tions between treatment group and each outcome smoking cessation study and the general popula- tion of smokers.
Logistic regression models were estimated for each of the major outcomes of interest to assessthe independent association of the two treatment Main Outcome Measures groups covering counseling with each outcome The primary outcomes of interest were: making using the ‘‘drugs only'' group as the referent a quit attempt (stopped smoking for one or more group. The models were run controlling for: days during the study because they were trying to 1) smoking characteristics at baseline (made quit and not for some other reason), quitting dur- a quit attempt in lifetime, number of cigarettes ing the study (stopped smoking for seven or more smoked per day, age started smoking regularly, days in a row during the study because they were stage of readiness to quit, used drugs in a prior trying to quit and not for some other reason), and quit attempt, prior use of Wellbutrin for non- prevalent abstinence (had not smoked a cigarette smoking related diagnosis), 2) demographic char- for seven or more days in a row at the eight- acteristics (age, gender, income, race), and 3) month follow-up interview). A second outcome doctor visit during the study period. These varia- was the total cost of coverage of tobacco depen- bles were selected based on previous research on dence treatments under the three different benefit the determinants of quitting smoking (Kabat and designs and the standardized cost per quitting out- Wynder 1987; DHHS 1990). The doctor visit come. Additional outcomes of interest included variable measured any doctor visit. While indi- Treating Tobacco Dependence vidual doctors were not informed regarding abstinence rates at eight months averaged 16% which of their patients were participating in the across all groups, ranging from 13% to 19%.
study, prior research finds that more than half In addition, utilization of the pharmacotherapy of all doctors in California advise their patients benefit did not vary across treatment groups. On to quit smoking during an office visit. Thus it average, 20% of subjects filled a prescription for was important to control for physician contact one of the covered medications, with essentially over the study period.
no variation observed across the three study Adjusted odds ratios and 95% confidence in- groups. This pattern held true for Zyban (11%), tervals were estimated from the coefficients in the NRT patch (8%), and NRT nasal spray or the logistic models. Costs of treatment for each inhaler (4%). However, for the two treatment group were estimated based on utilization of the groups for which proactive telephone counseling treatments and the costs of each covered drug was covered, statistically significant differences (for a 12-week course of treatment) to the PPO, were observed in enrollment in the proactive tele- the cost of enrollment in the proactive telephone phone counseling program. While 8% of the sub- counseling program, and the cost of the self-help jects in the drugs and counseling group enrolled in kit sent to all study participants.
telephone counseling, approximately three timesas many subjects in the drugs if counseling group (24%) enrolled in the telephone counseling.
When we examined multiple treatment use Randomization and Demographics among participants in the two study groups For all demographic, smoking, and control vari- who had coverage for counseling in addition to ables, there were no statistically significant differ- drugs, we found that few chose only counseling.
ences across the three groups with the exception In the drugs and counseling group, of the 34 sub- of income level (Table 1). The drugs and counsel- jects who used any treatment, 30 chose to use ing group reported lower incomes compared to drugs; four chose only counseling. This pattern the other two study groups, although the partici- is similar in the drugs if counseling group; of pants as a whole had relatively high incomes, re- the 29 subjects who signed up for counseling, flecting their ability to purchase health insurance just seven used only counseling. In effect, fewer in the private individual market. Upon further than 3% of the sample across all three study examination, it appears that this observed dif- groups used counseling only.
ference is a purely random result. The study Differences in the characteristics of smokers participants were predominantly white (90%), who did and did not use the covered treatments female (66%), age 40 or older (67%), and over the course of the trial suggest that smokers smoked less than one pack of cigarettes per day who used covered treatments were more likely: (84%). They were also a highly motivated group to have been at a higher stage of readiness to quit with good access to the health care system, with (planning to quit), to have visited a physician in 94% reporting that they would like to stop smok- the past year, and to have reported using drugs ing, 86% having made at least one quit attempt in in a previous quit attempt (Table 3). In contrast, their lifetime, and 19% reporting using medica- those who did not use any covered treatments tion in their most recent quit attempt.
were more likely to report that they did not wantto quit and that they were in the pre-contemplationstage of readiness. No differences were observed in the groups who did and did not use covered A simple bivariate analysis of quitting outcomes treatments as measured by age, gender, income by treatment group found no statistically signifi- and race, or by number of cigarettes smoked cant differences across the groups in quit at- per day, age started smoking regularly, if they tempts, quit rates during the study, or prevalent had ever made a quit attempt in their lifetime, abstinence rates at eight months (Table 2). The and their use of Wellbutrin prior to the study average rate of making a quit attempt across all groups was 48%, ranging from 43% to 55%. Quit We also were interested in knowing whether rates during the study averaged 31% across all certain smokers preferred one type of treatment groups, ranging from 26% to 37%. Prevalent over the other and what characteristics were Inquiry/Volume 43, Spring 2006 Baseline characteristics by treatment group Number % Number % Number % Number % Chi-square Smoking characteristics Number of cigarettes smoked per day Age started smoking regularlya Made quit attempt in lifetime Tried to quit last year Number quit attempts in past year (of those who tried to quit)a Used medication in most recent quit attempt Use of Wellbutrin prior to study period Visit to doctor within last yeara Stage of readiness Don't want to quit Note: The stages of readiness are defined as follows: Planning (planning on quitting in the next 30 days); contemplation(contemplating quitting in the next six months, but not planning on quitting in the next 30 days); pre-contemplation (notcontemplating quitting smoking in the next six months, but would like to quit smoking); don't want to quit (does not want to quitsmoking). P values are in parentheses.
a A few variables have some missing responses and therefore have n's that vary from those indicated at the top of the table: Income:total n ¼ 378, drugs n ¼ 123, drugs and counsel n ¼ 137, drugs if counsel n ¼ 118. Age started smoking: total n ¼ 384, drugs n ¼ 125,drugs and counsel n ¼ 137, drugs if counsel n ¼ 122. Number quit attempts: total n ¼ 144, drugs n ¼ 47, drugs and counsel n ¼ 52,drugs if counsel n ¼ 45. Doctor visit: total n ¼ 385, drugs n ¼ 123, drugs and counsel n ¼ 140, drugs if counsel n ¼ 122.
associated with different treatment choices. We had used drugs in a previous quit attempt were found that smokers who were taking Wellbutrin less likely to use bupropion (Zyban) as a covered prior to the start of the study were more likely to use NRT as a covered benefit, and those who The results of the logistic regressions using Treating Tobacco Dependence Quitting behaviors by treatment group (n ¼ 388) Intermediate outcomes Did at least 1 covered treatment (pharmacotherapy or counseling) Filled Rx for pharmacotherapy Filled Rx for bupropion (Zyban) Filled Rx for NRT patch Filled Rx for NRT spray/inhaler Enrolled telephone counselinga Quit attempt during study Quit during study Prevalent abstinence Note: It was possible for study participants to fill a prescription for more than one type of pharmacotherapy. P values are inparentheses. NA ¼ not applicable.
a n ¼ 262 (excludes drugs only group).
an intent-to-treat model, which included all 393 assess the effects of different benefit designs that study subjects originally enrolled in the trial, add health insurance coverage of proactive tele- confirm the findings observed in the bivariate phone counseling services to pharmacotherapy analysis: that neither of the treatment groups with coverage for treating tobacco dependence on coverage for proactive telephone counseling re- the use of covered services, quitting outcomes, ported higher quit-attempt rates, higher quit rates and benefit costs. The findings presented here during the study or higher prevalent abstinence in the context of a health insurance benefit design rates at eight months compared to the drugs only are consistent with the previously published stud- group, regardless of benefit design (Table 4).
ies, which find that the addition of proactive tele- Using the drugs only group as the referent group, phone counseling to pharmacotherapy does not only one of the adjusted odds ratios for the three positively impact smoking cessation practices quitting outcomes was statistically significant for and quit rates compared to pharmacotherapy one of the treatment groups with covered coun- alone (Stead, Lancaster, and Perera 2003; Ock- seling. This exception was the drugs and counsel- ene et al. 1991; Lando et al. 1997; Reid, Pipe, ing group, which had lower odds of making a quit and Dafoe 1999; Solomon et al. 2000). In ad- attempt compared to the drugs only group.
dition, this is the first study to examine two The costs of treating tobacco dependence different benefit designs for covering both phar- under each of the benefit designs varied approx- macotherapy and counseling to treat tobacco de- imately twofold due to the added costs associated pendence. The results from this study show that with proactive telephone counseling (Table 5).
although there was an increase in use of proactive The group with drug coverage only consistently telephone counseling among those in the drugs if had the lowest standardized costs per study par- counseling group, this group did not have any ticipant ($85) and the lowest costs for achieving higher quit rates compared to the group with un- each of the major study outcomes.
linked counseling and pharmacotherapy benefits.
There are three major findings from this case study that are particularly noteworthy. The first is that adding coverage for proactive telephone Previous studies have examined the effects of counseling to coverage for pharmacotherapy for telephone counseling as an adjunct to pharmaco- treating tobacco dependence did not increase quit therapy on cessation rates. This case study, how- attempts or quit rates among adult smokers. The ever, represents the first randomized trial to second is that linking access to drugs to counsel- Inquiry/Volume 43, Spring 2006 ing by restricting drug coverage to those enrolled Comparison of smoking and in proactive telephone counseling did not act as demographic characteristics of study a barrier to use of these medications or as a com- participants who used treatments and plement that enhanced the effect of the drugs.
those who did not use treatments Regardless of benefit design, the rates of use ofZyban, the nicotine patch, and nicotine nasal spray did not vary across the three treatment treatments treatments groups. The third finding is that the cost of adding coverage for telephone counseling to Smoking characteristics at baseline a pharmacotherapy benefit was substantial, in- Number of cigarettes creasing the costs approximately twofold to smoked per day (mean) Age started smoking regularly achieve each major quitting outcome. Coverage Made a quit attempt for pharmacotherapy only was clearly the most efficient benefit design for treating tobacco Stage of readiness to quit* dependence, achieving similar outcomes at a Don't want to quit (%) Pre-contemplation (%) The first major finding raises serious questions Contemplation (%) about the policies adopted by employers or health plans to add coverage for proactive tele- Used drugs to quit in previous quit (self-report)* (%) phone counseling services to treat tobacco depen- Use of Wellbutrin prior dence if pharmacotherapy (Zyban and NRT) is covered. In fact, our trial found that quit attempts (pharm. records) (%) were lower in the group that received unlinked Visit to doctor during drug and telephone counseling benefits and no study period* (%) different in the group that received linked tele- phone counseling benefits compared to those who received coverage for drugs only. In addi- tion, even though the smokers in the treatment group with linked drug and counseling benefits $50,000–$75,000 (%) enrolled in the proactive telephone counseling at nearly three times the rate compared to the treatment group with unlinked drug and coun- White, non-Hispanic (%) seling benefits, there were still no differences ob- served for these two groups in quit attempts or quit rates compared to the drugs only group after * There is a difference at the p < .05 level.
eight months.
The second finding is that, at least for proactive telephone counseling, requiring enrollment in be higher if smokers receive both counseling and counseling in order to obtain drug coverage does pharmacotherapy rather than just drugs alone.
not deter smokers from getting prescriptions for The second is to establish a barrier to control covered tobacco dependence medications. The costs and utilization by limiting access to drugs rates at which smokers filled prescriptions for to only those who demonstrate their willingness Zyban and the nicotine patch or nasal spray were and interest in quitting by committing to partici- no different for those smokers who had to first pate in a proactive telephone counseling pro- sign up for telephone counseling compared to gram. Our results suggest that there is little those who did not. Many employers and health reason to believe that linked policies will accom- plans currently structure their smoking cessation plish either of these objectives.
benefits such that access to covered pharmaco- The third finding that adding coverage for tele- therapy is linked to enrollment in counseling phone counseling to a pharmacotherapy benefit (Schauffler, Mordavsky, and McMenamin 2001b).
nearly doubles the cost with no added value in There are at least two possible objectives for such terms of outcomes suggests that the most efficient policies. The first is the belief that quit rates will benefit design is to limit coverage to pharmaco- Treating Tobacco Dependence Adjusted odds ratios (ORs) of quitting behaviors by treatment group Quit during study Prevalent abstinence Group: Drugs only (referent) Group: Drugs and counseling Group: Drugs if counseling Note: Analysis controls for smoking characteristics at baseline (number of cigarettes smoked per day, age started smoking regularly,made a quit attempt in lifetime, stage of readiness, used drugs in prior quit attempt, used Wellbutrin prior to study period),demographic characteristics (age, gender, income, race), and doctor visit during the study period. In parentheses are 95% confidenceintervals.
therapy, including Zyban and nicotine replacement therapy is covered. In addition, employers and therapy, except for those smokers for whom phar- health plans may want to cover proactive tele- macotherapy is not indicated or desired. The cost phone counseling, but only for those smokers of enrollment in the proactive telephone counsel- for whom pharmacotherapy is not medically in- ing program was $185 per smoker, which added dicated or desired; research studies and meta- more than $2,000 to the total costs of coverage analyses have found that proactive telephone for the drugs and counseling group, and more counseling interventions alone statistically sig- than $5,000 to the total costs of coverage for nificantly increase the odds of quitting smoking the drugs if counseling group; there were no over less intensive treatments or no treatment added benefits observed for either group in terms (Fiore et al. 2000; Stead, Lancaster, and Perera of increased quit attempts or quit rates compared 2003; Borland et al. 2001). However, the litera- to the drugs only group.
ture to date and the results of this study do not in- These findings indicate that employers and form the question of whether adding coverage for health plans may achieve a significant impact face-to-face counseling by a health care provider on quitting behaviors and smoking rates at a rela- to a pharmacotherapy benefit increases quit and tively low cost by covering pharmacotherapy abstinence rates. Additional research is needed only. Our results suggest that rather than adding to address this question.
coverage for proactive telephone counseling to An alternative to coverage for employers and a pharmacotherapy benefit, health care dollars health plans that want to provide access to coun- may be more efficiently used if only pharmaco- seling services, particularly for those smokers Costs of tobacco dependence treatment coverage by group (n ¼ 388) Coverage cost ($) Drugs and counseling Drugs if counseling Cost per covered treatmenta Self-help kit ($27) NRT nasal spray/inhaler ($427) Proactive telephone counseling ($185) Total cost of treatments Cost/study participant Cost/quit attempt during study Cost/quit during study a Pharmacy costs are based on a 12-week course of treatment. All costs are based on actual utilization of treatments by studyparticipants in each group over the eight months of the study.
Inquiry/Volume 43, Spring 2006 who cannot or do not want to use pharmacothera- Finally, these findings suggest that national py, is to refer them to a state telephone quit line guidelines for tobacco dependence treatments to receive counseling services. In 2003, 34 states need to be updated and revised to address the offered counseling services through a toll-free tele- effectiveness of combined therapies. The 2000 phone quit line to assist smokers in their cessation PHS clinical practice guideline was silent on attempts (Center for Tobacco Cessation 2004).
In light of the heterogeneity of tobacco depen- pharmacotherapy and counseling. At present, dence treatment benefit designs that are offered in more than half of the Medicaid programs and practice today, it is important that employers and health maintenance organizations in the United health plans take into account the relative impacts States cover treatments for tobacco dependence of different benefit designs for smoking cessation to most efficiently and effectively increase quit McMenamin, and Keeler 2004). The continued attempts and quit rates, and ultimately improve development of both public and private health the overall health status of their populations.
insurance benefits for smoking cessation would Our findings add to the growing body of literature be greatly aided by a comprehensive review of on tobacco dependence treatment indicating that the evidence of the effectiveness of combined it is a drug benefit which produces the observed pharmacotherapy and counseling services. This increases in quit attempts and quit rates at the is particularly important given pressures to lowest cost to employers and health plans, with control rising health care costs. It is in the inter- no value added by covering proactive telephone ests of employers, health plans, and state and counseling (Stead, Lancaster, and Perera 2003), federal governments to design cost-effective regardless of benefit design, when Zyban and benefits that achieve desired outcomes at the nicotine replacement therapy are covered.
lowest cost.
The funder of this research, the California Tobacco- A. Bupropion SR—Contraindications: history of Related Disease Research Program (TRDRP), was seizure or history of eating disorder; Side effects: in- established after the passage of California's Proposi- somnia, dry mouth; Dosage: 150 mg every morning tion 99 in November 1988. Proposition 99 instituted for three days, then 150 mg twice daily (begin treat- a $.25 per pack cigarette tax, of which 5% was ear- ment one to two weeks pre-quit); Duration: seven to marked for research on tobacco-related disease. This 12 weeks, maintenance up to six months; Available led the California State Legislature to authorize the creation of TRDRP, which is administered by the B. Nicotine patch—Side effects: local skin reac- University of California.
tion, insomnia; Dosage and duration: 21 mg/24hours (two weeks), 14 mg/24 hours (two weeks), 1 The PHS 2000 Clinical Practice Guideline (Fiore 7 mg/24 hours (four weeks), or 15 mg/16 hours et al. 2000) defines Bupropion SR and NRT as fol- (eight weeks); Available as: Nicoderm CQ (over lows: Bupropion SR (bupropion sustained-release) the counter [OTC] only), generic patches (prescrip- is a non-nicotine aid to smoking cessation originally tion and OTC), Nicotrol (OTC only).
developed and marketed as an antidepressant. Its C. Nicotine inhaler—Side effects: local irritation mechanism of action is presumed to be mediated of mouth and throat; Dosage: six to 16 cartridges/ through its capacity to block the re-uptake of dopa- day; Duration: up to six months; Available as: Nic- mine and norepinephrine centrally. It is available as otrol Inhaler (prescription only).
Zyban (used for smoking cessation) or Wellbutrin D. Nicotine nasal spray—Side effects: nasal (used for depression). Nicotine replacement therapy irritation; Dosage: eight to 40 doses/day; Duration: refers to a medication containing nicotine that is three to six months; Available as: Nicotrol NS intended to promote smoking cessation. There are four nicotine replacement therapy delivery systemscurrently approved for use in the United States.
Taken from the Suggestions for the Clinical Use of These include nicotine chewing gum, nicotine in- Pharmacotherapies for Smoking Cessation. U.S.
haler, nicotine patch, and nicotine nasal spray.
Public Health Service. http://www.surgeongeneral.
The side effects, dosages, duration, and availability of the pharmaceuticals used in this study are as 2 The proactive telephone counseling was conducted by a nationally recognized firm using standard pro- Treating Tobacco Dependence tocols accepted in the field. Nicotine dependence regarding cessation were assessed using a 1–10 was assessed using a modified Fagerstrom Toler- ance Scale. Readiness to quit was measured 3 Data from the 2000 California Behavioral Risk using Prochasksa's (Prochaska, DiClemente, and Factor Survey were used as a comparison for our Norcross 1992) five stages of change. Motivation sample. This data is available at http://www.
to use behavioral techniques and self–efficacy Borland, R., C.J. Segan, P.M. Livingston, and N.
Kabat, G.C., and E.L. Wynder. 1987. Determinants of Owen. 2001. The Effectiveness of Callback Quitting Smoking. American Journal of Public Counseling for Smoking Cessation: A Random- ized Trial. Addication 96(6):881–889.
Lando, H.A., S. Rolnick, D. Klevan, et al. 1997.
Center for Tobacco Cessation. 2004. Quitting Telephone Support as an Adjunct to Transdermal Isn't Easy, Getting Help Is. Washington, Nicotine in Smoking Cessation. American Jour- D.C.: Center for Tobacco Cessation. Available nal of Public Health 87(10):1670–1674.
Max, W., D.P. Rice, H-Y Sung, et al. 2004. The Economic Burden of Smoking in California.
Curry, S.J., L.C. Grothaus, and T. McAfee. 1998. Use Tobacco Control 13(3):264–267.
and Cost Effectiveness of Smoking-Cessation McPhillips-Tangum, C., C. Bocchino, R. Carreon, Services Under Four Insurance Plans in a Health et al. 2004. Addressing Tobacco in Managed Maintenance Organization. New England Journal Care: Results of the 2002 Survey. Public Health of Medicine 339(10):673–679.
Research Practice and Policy 1:1–11.
Department of Health and Human Services (DHHS).
Ockene, J.K., J. Kristeller, R. Goldberg, et al. 1991.
1990. The Health Benefits of Smoking Cessation: Increasing the Efficacy of Physician-Delivered A Report of the Surgeon General. Atlanta, Ga.: Smoking Interventions. Journal of General In- Centers for Disease Control and Prevention, ternal Medicine 6(January/February):1–8.
National Center for Chronic Disease Prevention Prochaska, J.O., C.C. DiClemente, and J.C. Norcross.
and Health Promotion, Office of Smoking 1992. In Search of How People Change. Appli- cations to Addictive Behaviors. American Psy- ———. 2004. The Health Consequences of Smoking: A Report of the Surgeon General. Atlanta, Ga.: Reid, R.D., A. Pipe, and W.A. Dafoe. 1999. Is Centers for Disease Control and Prevention, Telephone Counseling a Useful Addition to National Center for Chronic Disease Prevention Physician Advice and Nicotine Replacement and Health Promotion, Office on Smoking and Therapy in Helping Patients to Stop Smoking? A Randomized Controlled Trial. Canadian Med- Fagerstrom, K.O., and N.G. Schneider. 1989. Mea- ical Association Journal 160(11):1577–1581.
suring Nicotine Dependence: A Review of the Schauffler, H.H., S.B. McMenamin, K. Olson, et al.
Fagerstrom Tolerance Questionnaire. Journal of 2001a. Variations in Treatment Benefits Influence Behavioral Medicine 12(2):159–182.
Smoking Cessation: Results of a Randomised Fiore, M.C., W.C. Bailey, S.J. Cohen, et al. 1996.
Controlled Trial. Tobacco Control 10:175–180.
Smoking Cessation. Clinical Practice Guideline Schauffler, H.H., J.K. Mordavsky, and S.B. McMe- No. 18. AHCPR Publication No. 96-0692. Rock- namin. 2001b. Adoption of the AHCPR Clinical ville, Md.: U.S. Department of Health and Human Practice Guideline for Smoking Cessation: A Services, Public Health Service, Agency for Survey of California's HMOs. American Journal Health Care Policy and Research.
of Preventive Medicine 21(3):153–161.
———. 2000. Treating Tobacco Use and Depen- Solomon, L.J., G.M. Scharoun, B.S. Flynn, et al.
dence. Clinical Practice Guideline. Rockville, 2000. Free Nicotine Patches Plus Proactive Md.: U.S. Department of Health and Human Telephone Peer Support to Help Low-Income Services, Public Health Service.
Women Stop Smoking. Preventive Medicine Halpin, H.A., S.B. McMenamin, and C.L. Keeler.
2004. State Medicaid Coverage for Tobacco Stead, L.F., T. Lancaster, and R. Perera. 2003.
Dependence Treatments—United States, 1994– Telephone Counseling for Smoking Cessation.
2002. Morbidity and Mortality Weekly Report Cochrane Database Systematic Review (1):

Source: http://inq.sagepub.com/content/43/1/54.full.pdf