Abstract
Purpose. To evaluate the effectiveness of the Pharmacist-Managed Telephone Tobacco Cessation Clinic (PMTTCC) compared to the standard of care (SOC) at the Veterans Affairs San Diego Healthcare System. Method. A retrospective cohort study was performed investigating the proportion of veterans who quit smoking at 6 months while enrolled in the PMTTCC. Chart review was performed using the Veterans Affairs Computerized Patient Record System. The PMTTCC group included patients who had received medication and counseling from the tobacco cessation pharmacists. The cohort was compared to a matched SOC group who did not receive counseling, only tobacco cessation medication therapy through a primary care provider. The primary outcome for this study was patient-reported tobacco cessation at 6 months. Secondary outcomes were abstinence at 1 and 3 months. Results. A total of 1,006 patients were included in the analysis, 503 patients from the PMTTCC and 503 patients from SOC. The overall study population was 54 years old on average, 92.5% male, 70.0% Caucasian, 45.5% with history of psychiatric conditions, and had an average smoking history of 33-pack years. Patients in the PMTTCC group had statistically significant improvements in abstinence at 6 months versus the SOC group (81/503, 16.1% vs. 48/503, 9.5%; p < .0001). Quitters were older on average versus non-quitters (56.03 vs. 53.65 years; p = .01). Conclusion. Patients enrolled in the PMTTCC had improved tobacco abstinence rates at 6 months compared to SOC. Although the study was not designed to test for causality, the results lend support for using intensive tobacco cessation management in veteran population.
Introduction
Prior to the Surgeon General’s first report on the dangers of smoking released in January 1964, the per capita annual adult consumption of cigarettes in the United States had reached 4,345 cigarettes (U.S. Department of Health and Human Services, 2004). Despite improvements in reducing prevalence over the years, smoking remains the leading preventable cause of death in the United States (U.S. Department of Health and Human Services, 2004). Between 2000 and 2004, tobacco smoking was found to be the primary causal factor for at least 30% of all cancer deaths and has also been heavily associated with almost 80% of deaths from chronic obstructive pulmonary disease (Centers for Disease Control and Prevention [CDC], 2008). There is sufficient evidence linking tobacco use to a higher risk of cardiovascular diseases, damaging reproductive effects, peptic ulcer disease, dental diseases, and an overall diminished health status, which contribute to over 420,000 deaths per year (U.S. Department of Health and Human Services, 2004).
Although it is generally known that tobacco use is harmful, it has also been found that tobacco cessation at any stage is beneficial. The 2010 Surgeon General’s Report reviewed the benefits of tobacco cessation and reported as one of its major conclusions that there is no riskfree level of exposure to tobacco smoke (U.S. Department of Health and Human Services, 2010). Despite this, a significant percentage of the United States population continues to smoke. In 2008, approximately 20.6% of adults in the United States were considered current smokers, and this has remained consistent in recent years (CDC, 2009). One study performed by the CDC found that the prevalence of smoking among United States veterans between 2004 and 2007 was 27% compared to 21% among nonveteran adults (Brown, 2010). This study also found that the smoking prevalence among United States veterans with coronary heart disease was 43% compared to 31% in nonveterans with coronary heart disease, a statistically significant difference (Brown, 2010). In 2007, the veteran population in California had a smoking prevalence of 18% based on the enrollee survey reported by the Veterans Integrated Service Network in California and parts of Nevada, whereas the general population in California had a smoking prevalence rate 14.3% (California Department of Health Services, 2006). During the same year, the prevalence of tobacco use among veterans across all Veterans Integrated Service Networks ranged from 16.5% to 27%, significantly contributing to high morbidity and mortality rates in the veteran population (Kussman, 2008).
The veteran population generally has lower cessation rates than the general public, suggesting that veterans may be more difficult to treat (Roth, Andrus, & Westman, 2005). Roth et al. reported that veterans enrolled in a pharmacist-managed outpatient smoking cessation clinic (using a low-intensity counseling model) achieved a smoking cessation rate of less than 10%, whereas the general population achieved a cessation rate of 16% (Roth et al. 2005). In addition, An et al. (2006) reported that veterans at five distinct Veterans Affairs (VA) medical centers that received standard care achieved cessation rates of 4%, which further suggests that veterans may be more difficult to manage. There are potential confounding factors when evaluating tobacco cessation in the veteran population. One factor recently reviewed is a possible link to posttraumatic stress disorder. Patients with any diagnosed psychiatric disease states have lower tobacco cessation rates that those without (McFall et al., 2005). Data reported in 2007 by Gershon-Grand, Hwang, Han, George, and Brody estimate that between 45% and 60% of patients with posttraumatic stress disorder are also tobacco users. A study by Kirby et al. (2008) suggests that tobacco use may be a coping mechanism for some symptoms of posttraumatic stress disorder. A review published by Fu et al. (2007) acknowledged that conclusive data regarding appropriate treatment strategies for these patients are lacking and require further investigation.
With these additional obstacles present in the veteran population, the Veterans Health Administration has taken a more proactive approach on tobacco cessation. In 2003, the VA Administration Directive 2003-042 stated that smoking cessation medications needed to be made available to all patients (Perlin, 2003). The Veterans Affairs San Diego Healthcare System (VASDHS) responded by creating a Pharmacist-Managed Telephone Tobacco Cessation Clinic (PMTTCC) to provide alternative access to tobacco cessation care and in adjunct to the currently available programs (e.g., standard of care [SOC]; Harris, Chen, Kim, & Nguyen, 2009).
To improve access to tobacco cessation care and with demonstrated efficacy in previous clinical studies, pharmacists at the VASDHS focused their efforts on developing a telephone-based clinic to reach a larger number of tobacco users (Harris et al., 2009). The tobacco cessation telephone clinic was part of a comprehensive provider program in which a collaborative scope of practice with a supervising physician granted clinical pharmacists prescribing privileges to enhance tobacco cessation treatment in the veteran population (Harris et al., 2009). The telephone-based clinic was originally designed to be a proactive model where initial calls consisted of intensive counseling for approximately 20 to 30 minutes with follow-up calls lasting for approximately 10 to 15 minutes. However, due to a lack of staffing and increased workload on clinical pharmacists, the clinic was modified into a reactive model where patients initiated the telephone calls. In this modified model, patients would call the tobacco cessation clinic to initiate treatment for the intensive counseling session and were given tobacco cessation medications after initial assessment by a clinical pharmacist. Patients were then instructed to make follow-up calls in order to get refills and counseling (Harris et al., 2009). To our knowledge, this was the first time a reactive model was initiated for tobacco cessation. We sought to evaluate the proportion of smoking veterans in the PMTTCC group who quit at 6 months compared to those in the SOC group.
Method
A retrospective cohort study was performed comparing tobacco cessation rates between patients who received the SOC treatment and patients enrolled in the PMTTCC at the VASDHS. Data was collected using the VA Computerized Patient Record System (CPRS). Both the SOC and PMTTCC data were collected from chart reviews from clinic notes which may include Tobacco Cessation Clinical Reminders, a CPRS tool to provide a brief intervention (Ask, Advise, Assess, Assist, Arrange).
For the SOC group, patients were included if they filled a prescription for any approved tobacco cessation medication between January 2003 and December 2005. These medications were on the VA national formulary (reflective of enrollment dates) and included nicotine replacement treatments (gum, lozenge, patch, nasal spray, oral inhaler), bupropion with a specific indication for tobacco cessation, or any combination therapy. Patients were included in the PMTTCC group if they had completed the enrollment process for medication and counseling from the tobacco cessation pharmacists between January 1, 2007, and June 30, 2007. Patients were excluded from either group if they were currently enrolled in another tobacco cessation program such as Mental Health Clinic–Tobacco Cessation Program or other Tobacco-Free classes offered through the VASDHS. Additionally, patients were excluded from the SOC group if they did not have any CPRS-documented follow-up with a primary care provider after filling their tobacco cessation prescription.
The primary outcome for this study was patient-reported tobacco cessation at 6 months. Additional measurements were taken at 1 and 3 months. Tobacco cessation was defined as any documentation by a health care professional within CPRS of the patient remaining abstinent from tobacco use over the specified time period. Those who were able to achieve tobacco cessation over this specified time period are referred to as “quitters.” Those who were unsuccessful in refraining from the use of tobacco-containing products during the specified time period were referred to as “nonquitters.”
Drug therapy in both groups was restricted to a formulary tier system based on the VA National Formulary and local tobacco cessation guidelines. First-line agents (those without prescribing restrictions) included nicotine patches, nicotine gum, and bupropion sustained-release. Other tobacco cessation medications included nicotine nasal spray, nicotine oral inhaler, nicotine lozenges, varenicline, and any combination (nicotine patch with a short-acting nicotine replacement therapy; nicotine replacement therapy with bupropion sustained-release) were also available condition to previous failure with first-line agents (nicotine patch, nicotine gum or bupropion).
Demographic information collected included age, race, diagnosis of psychiatric illness, years of tobacco use, quantity of tobacco used, amount of previous quit attempts, and medications used in previous quit attempts. Specific to the PMTTCC group, the number of follow-up phone calls was documented.
Statistical Analysis
Means and standard deviations were presented for continuous data; frequency and percentages were presented for discrete data. Test for normality was performed using the Shapiro–Wilk test. Continuous data were analyzed using Student’s t test or Mann–Whitney U when appropriate. Categorical data were analyzed using Pearson’s chi-square or Fisher’s exact test when appropriate. Univariate analyses included a comparison of quitters versus nonquitters within the PMTTCC group at 1, 3, and 6 months. Cochran–Mantel–Haenszel test was performed to determine if the proportions of quitters at 6 months were different between the PMTTCC and SOC groups controlling for gender. Homogeneity of odds ratio was evaluated using the Breslow–Day test. Sensitivity analysis was performed using logistic regression to confirm the primary results of abstinence at 6 months while controlling for gender, age, smoking history, and psychiatric history. Statistical significance was set a priori at p < .05 (two-tailed). Data analysis was performed using SPSS statistical software (SPSS Inc., Chicago, IL, USA) Version 18.0.
Results
Study Population
A total of 1,006 patients were included based on inclusion and exclusion criteria. Thirteen patients were excluded from the PMTTCC group due to enrollment in another tobacco cessation group. The PMTTCC and SOC groups had 503 patients each (baseline characteristics listed in Table 1). The majority of patients were male in both groups (92.5% overall). The mean age was approximately 54 years in both groups without significant differences between groups (p = .956). The majority of patients were Caucasian in both the PMTTCC (n = 337, 71%) and SOC (n = 367, 73%) groups, followed by African Americans (11% and 19%, respectively; p < .001). Both groups had a significant number of patients with a prior psychiatric history, but the PMTTCC group had a significantly higher percentage than the SOC group (52% vs. 39%; p < .001).
Baseline Characteristics
Smoking history was not significantly different between the PMTTCC (33.37 years) and SOC (32.36 years) groups (p = .206). However, the SOC group had significantly fewer prior quit attempts versus the PMTTCC group (p < .001). In evaluating the medications used for the quit attempt, the groups did show statistically significant differences in treatment strategies (p < .005). The SOC group was more significantly likely to prescribe nicotine only therapy versus the PMTTCC group (461/503, 91.7% vs. 361/503, 71.8%; p < .001). In the PMTTCC group, 32.2% (162/503) of the patients used bupropion only or in combination with a nicotine product compared to only 8.3% (42/503) in the SOC group. See Table 1 for complete details.
Primary and Secondary Outcomes
PMTTCC had higher abstinence rates at all three times points. At 1 month, the PMTTCC group had a higher rate of patient-reported tobacco cessation, with 187 patients (37%) remaining abstinent compared to only 116 patients (23%) in the SOC group (p < .001). At 3 months, the PMTTCC group continued to have higher abstinence, with 119 patients (24%) remaining tobacco-free whereas the SOC group had 64 patients (13%) report abstinence (p < .001). At 6 months, the primary outcome measure, abstinence, remained higher in the PMTTCC group (N = 81, 16%) compared to the SOC group (N = 48, 10%; p < .001). Both groups saw noticeable decreases across the three time points in abstinence. Full results are presented graphically in Figure 1.
Proportion of Patients Who Quit Smoking at 1 Month, 3 Months, and 6 Months
Quitters Versus Nonquitters
Univariate analysis between all quitters and nonquitters at 6 months for the PMTTCC and SOC was performed to identify relationships between variables and abstinence (Table 2). Within the PMTTCC group, patient’s age at baseline was significantly different between quitters and nonquitters. The mean age among the 81 patients who quit was 56.44 (SD = 11.31) years versus 53.32 (SD = 10.85) years among the 422 patients who were unable to quit (p = .03). In addition, mean number of follow-up calls made during quit attempt were statistically different between quitters and nonquitters for the PMTTCC group. Patients in the PMTTCC group who were able to quit at 6 months had a mean of 1.27 (SD = 1.70) follow-up phone calls whereas patients who continued smoking had a mean of 0.43 (SD = 0.80) follow-up phone calls (p < .001). There were no significant differences in packs per day (PPD), number of previous quit attempts, ethnicity, gender, history of psychiatric illness, or medications used between quitters and nonquitters within the PMTTCC group. In the SOC group, patients who continued to smoke had a 0.98 PPD (SD = 0.57) mean whereas patients who were able to quit had a mean of 0.78 PPD (SD = 0.63) at baseline (p = .026). In addition, 25% of the SOC quitters had a psychiatric illness whereas approximately 41% of the SOC smokers who continued to smoke had a psychiatric illness (p = .043).
Comparison of Quitters and Nonquitters at 6 Months (PMTTCC Group Only)
Data for 80 patients who quit smoking and 419 in patients who continued smoking were available.
Data for 420 patients who continued smoking were available.
Data for 454 patients who continued smoking were available.
Logistic Regression
Sensitivity analysis using logistic regression confirmed our primary aim that patients in the PMTTCC group had a higher likelihood of quitting at 6 months compared to patients in the SOC group when controlling for gender, age, smoking history, and psychiatric history (odds ratio = 2.72; 95% confidence interval: 1,74, 4.23). Breslow–Day test suggested no difference in likelihood of quitting at 6 months between males and females (p = .366) ensuring homogeneity of odds ratio between genders. Figure 2 illustrates the proportion of patients quitting at 6 months stratified by gender.
Abstinence Rates at 6 Months Between Intervention Groups Stratified by Gendera
Discussion
The results from this retrospective analysis among a veteran population provide insight into variables that can positively influence an individual’s tobacco cessation attempt. At 6 months, the PMTTCC group had a higher proportion of patient-reported tobacco cessation than the SOC group. One likely reason for this is that the intensity of contact and counseling received by patients in the PMTTCC group was different compared to the SOC group. All patients in the PMTTCC group received an initial phone call of 20 to 30 minutes with a pharmacist evaluating the patients’ readiness to begin the quit attempt and which medications would aid them best.
In addition to this initial call, patients received follow-up calls of 10 to 15 minutes. A small proportion of patients benefited from a high number of follow up calls. However, it is uncertain what kind of effect this may have among patients not enrolled in the PMTTCC. Patients in the SOC group may have had follow-up calls from their providers, but this was not measured as this information was not readily retrievable. Consequently, we were unable to control for this in the regression model. According to Graham et al. (2011), regardless of the wealth of information presently available through the Internet, patients who received additional telephone counseling had higher tobacco cessation rates. These data were also further supported by the most recent 2008 Clinical Practice Guidelines, which recommend using a combination of medication and counseling which has been proven to be more effective than either intervention alone (U.S. Department of Health and Human Services, 2008). Future studies will need to document the number and type of follow-up calls made to patients attempting to quit smoking and the relationship to tobacco cessation.
Although data have long supported an increase in tobacco cessation rates as a result of primary care providers inquiring about smoking cessation, providing additional counseling and pharmacotherapy can potentially increase the proportion of patients quitting (Sherman, Yano, Lanto, Simon, & Rubenstein, 2005). An et al. (2006) showed that when comparing 6-month abstinence rates between telephone clinics and SOC, the proportion of patients quitting ranged from 4.1% to 13.0%. In the Treatment of Tobacco Use and Dependence: 2008 Update, a meta-analysis of 58 studies demonstrated that proactive telephone counseling produced a cessation proportion of 13.1% compared to 10.8% with other programs at 6 months (U.S. Department of Health and Human Services, 2008). Moreover, Hollis et al. (2007) reported cessation proportion of 24.3% with intensive telephone counseling and nicotine replacement therapy, further supporting the effectiveness of telephone counseling.
Patients who were older by 1 year had higher odds of quitting at 6 months compared to patients who were younger regardless of intervention or gender. There are several reasons why this trend might hold true. Older patients tend to have more medical comorbidities such as hypertension, chronic obstructive pulmonary disease, or a previous cardiovascular event that may provide patients extra motivation to remain tobacco-free. As supported by the 2008 guidelines, patients with such medical comorbidities would benefit greatly from tobacco cessation (U.S. Department of Health and Human Services, 2008). Duffy, Biotti, Karvonen-Gutierrez, and Essenmacher (2011) evaluated the effect of medical comorbidities on the motivation level in tobacco cessation attempts in veterans and reported history of arthritis, diabetes, lung disease, or stroke was associated with a patient’s motivation to quit tobacco use. According to Duffy et al., although some variables (e.g., ethnicity, psychiatric history, number of previous quit attempts, and medications used) may have been different among the two groups at baseline, they had no statistically significant effect when evaluating quitters and nonquitters. This was also observed in our logistic regression model when gender, age, smoking history, and psychiatric history were controlled. In the SOC group, the quitters smoked less, suggesting that SOC may be more appropriate for less nicotine dependent veterans. In addition, there was statistical difference between quitters and nonquitters for psychiatric illness in the SOC group, yet this was not observed in the PMTTCC group. PMTTCC may be a better option for veterans with psychiatric illnesses. Future analysis will be needed to establish PMTTCC for psychiatric patients.
There are several limitations to this study. First, this study was based on an unvalidated practice paradigm—the reactive model. The reactive model requires patients who are motivated to enroll themselves into the clinic, creating a layer of complexity that is difficult to measure. Patients also initiated their own telephone encounters in this reactive model. As a result, patients in the PMTTCC group may be more motivated to be tobacco-free at 6 months. Second, in accordance with our study design, if no documentation was recorded in CPRS, that patient was assumed to have been unsuccessful in the quit attempt. It is possible that some patients were in fact successful and the health care provider did not document this in the electronic medical record. Since SOC providers (e.g., primary care, etc.) were nonspecific to tobacco cessation, this could have lead to underreporting of quitters in the SOC group if effectiveness of tobacco cessation therapy was not documented in their follow-up medical notes after patient initiation. There have been no established reports that demonstrate SOC would be more prone to underreporting relative to the PMTTCC group. Future studies will need to clarify this potential bias.
Third, all of the data on abstinence failures or successes in both the PMTTCC and SOC groups are based on patient self-report. There were no laboratory data (e.g., CO2 or nicotine levels collected via blood, urine, or saliva) to validate whether the patients actually had quit smoking. Fourth, the results of this study may limit generalizability to a nonveteran population. Nicotine patch was the most frequent medication prescribed due to VA formulary preference during the period of the study. The PMTTCC group had more prescribing for bupropion, most likely due to their familiarity with the medication. During 2003-2007, the VA national formulary placed combination therapy as a second- or third-line option, available to patients only if they had already failed nicotine replacement therapy or bupropion therapy alone. Once the 2008 Clinical Practice Guidelines were released suggesting that combination therapy was most effective, the VA national formulary was updated to reflect these data. Outside practice in a nonveteran population may have already implemented the guidelines ahead of its release and may show potentially different outcomes in the proportion of abstinence at 6 months. Fifth, the PMTTCC group had significantly higher prior quit attempts. Zhu, Lee, Zhuang, and Garnst (2012) found a positive link between higher number of quit attempts and better cessation outcomes, suggesting that the PMTTCC group who had higher number of quit attempts may be more likely to be more motivated to quit than the SOC group.
Last, this study was not designed to test for causality. A prospective study that randomly assigns patients to the PMTTCC and SOC within the VA would validate the results of our study. Differences in motivation and validating self-reports would eliminate confounding introduced in a retrospective analysis. However, our findings offer encouragement to continue providing veterans’ access to the PMTTCC for their tobacco cessation needs. Regardless of limitations, this retrospective cohort study supports the current research and clinical practice guidelines, giving further insight to the challenges faced by patients undergoing tobacco cessation attempts.
Conclusion
In conclusion, this retrospective cohort study evaluated data from over 1,000 veteran patients who were either enrolled in the reactive PMTTCC program or received tobacco cessation therapy as SOC. Patients in the PMTTCC group had higher abstinence rates than the SOC group at 1, 3, and 6 months. A higher number of patient-initiated follow-up phone calls were observed in patients who were quitters at 6 months in the PMTTCC group. Based on this, the PMTTCC reactive clinic was revised starting February 2010 by making the clinic more proactive with scheduled follow-up calls for each patient. By scheduling phone calls to patients, the PMTTCC clinic hopes to increase the amount of counseling the patients receive during their quit attempt, which will ideally further increase tobacco cessation rates. Future prospective studies would be beneficial in confirming the results of our institution.