BACKGROUND/RATIONALE:
Numerous randomized controlled trials (RCTs) both inside and outside of the Veterans Health Administration (VHA) have demonstrated the efficacy of smoking cessation pharmacotherapies. Hence, both inside and outside the VHA, there has been wide-scale distribution of smoking cessation pharmacotherapies including the implementation of VHA directive 2003-042, National Smoking and Tobacco Use Cessation Program. This directive indicates that "cessation medications need to be made available to all smokers interested in quitting, regardless of whether or not the patient is willing to attend a cessation program." Despite the broad distribution of these smoking cessation pharmacotherapies, declines in adult smoking prevalence have stalled in the past 5 years. One in five adults continues to smoke and this is true in the VHA as well.

Recently outside the VHA, the generalizability of RCT results to the community setting has been challenged, as cross-sectional population studies have shown lower risk of relapse among persons not using cessation pharmacotherapies. The quit rates of the wide-scale distribution of cessation pharmacotherapies in the VHA are unknown. Given the evidence for lack of effectiveness of wide-scale distribution outside of the VHA, there is an urgent need to evaluate the value of the broad distribution of smoking cessation pharmacotherapies in the VHA, as funding ineffective services may be resulting in the loss of scarce public health resources. Hence this study evaluated VHA national data on the outcomes of VHA directive 2003-042, which mandated the broad distribution of smoking cessation pharmacotherapies in the VHA.

OBJECTIVE(S):
Using the Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM) framework as a guide, the objectives of this study were to: 1) Determine the reach of smoking cessation pharmacotherapies (percentage of smokers that receive smoking cessation pharmacotherapies over a ten-year period, trends over time, and patient-level factors associated with receipt of these smoking cessation pharmacotherapies); 2) Using health factors data, determine the effectiveness of smoking cessation pharmacotherapies by comparing quit rates among those who receive and do not receive specific smoking cessation pharmacotherapies; 3) Determine adoption by identifying the characteristics of providers, clinics, and facilities prescribing smoking cessation pharmacotherapies; 4) Determine the cost per quit of implementing smoking cessation pharmacotherapies; and 5) Based on the findings, make recommendations on the maintenance of the program.

METHODS:
This retrospective, cohort study used electronic medical record (EMR) data to evaluate the distribution of tobacco cessation pharmacotherapy in the VHA. Inverse probability of treatment weights were used to balance baseline differences in demographic characteristics and comorbidities among those who did and did not receive tobacco cessation pharmacology. Smoking quit rates were identified from longitudinal evaluation of the health factors (clinical reminder) data. Costs were estimated based on the direct cost of the medication (VA acquisition cost), indirect costs for labor of pharmacy departments, and prescribing health care encounter cost. Provider, clinic, and facility characteristics associated with the index tobacco use assessment were summarized from EMR.

Cox regression models were used to estimate the adjusted rate ratios of time to pharmacotherapy initiation (Reach). Multivariable logistic regression was used to estimate quit status (Effectiveness). Logistic regression models that included all provider characteristics as well as patient demographic and clinical characteristics and facility as fixed effects were used to determine the independent associations between a given provider characteristic and pharmacotherapy initiation (i.e., odds of initiation in those with the characteristic relative to those without the characteristic) (Adoption). The incremental cost-effectiveness ratio (ICER) was estimated as the difference in mean cost per additional quit over the follow up period (7 to 18 months) after the assessment date (Implementation). Based on these analyses, recommendations for continuing the distribution of tobacco cessation pharmacotherapies in the VHA were made (Maintenance).

FINDINGS/RESULTS:
Aim 1 (Reach) Findings. Smoking cessation pharmacotherapy in the VHA increased from 13.8% in 2004 to 25.6% in 2013. In 2011, current tobacco users (N=838,309) who were more likely to initiate pharmacotherapy (treated) included those with psychiatric disorders (depression, bipolar disorder, non-alcohol substance use disorder, other anxiety, and post-traumatic stress disorder), chronic pulmonary disease, peripheral vascular disorders, and younger Veterans (adjusted rate ratios (ARR) ranged from 1.03 to 1.92, all p<0.001). Veterans less likely to receive pharmacotherapy (untreated) were those with schizophrenia, other psychosis, several other medical conditions, and those who were male and Hispanic (all p<.001).

Aim 2 (Effectiveness) Findings. Adjusting for inverse probability of treatment weights (IPTWs) and covariates, the odds of quitting (N=589,862) were 24% higher among those treated versus untreated (OR=1.24, 95% CI = 1.23, 1.25, p<0.001). The marginal probabilities of quitting were 16.7% for the untreated versus 19.8% for the treated based on the weighted model.

Aim 3 (Adoption) Findings. Patients who were treated by providers that were younger (p<0.001) and female (p<0.001) were more likely to initiate tobacco cessation pharmacology. Compared to physicians (referent group), adjusted odds ratios (AORs) showed that patients who were treated by pharmacists were more likely to initiate pharmacotherapy (AOR=2.62 (95% CI=2.33, 2.95, p<0.001), while patients who were treated by physician assistants (AOR=.88), registered nurse (AOR=.95), nurse practitioner (AOR=.94), and licensed practical/vocational nurse (AOR=.90) were slightly less likely to initiate pharmacotherapy (p<.001 for all groups). Compared to outpatient primary care clinics, patients treated in outpatient substance abuse treatment clinics were more likely to initiate pharmacotherapy (AOR=1.22, 95% CI=1.14, 1.31, p<0.001) than those in outpatient psychiatry (AOR=.91), outpatient surgery (AOR=.61), and outpatient other (AOR=.85) (all p<.001). Compared to those seen in primary care CBOCs, other classes of VHA facilities, for example, health care center (AOR=.89), multi-specialty community-based outpatient clinics (CBOC) (AOR=>85), and VHA medical centers (AOR=ranging from .69 to .81, low complexity to high complexity), were more likely to initiate pharmacotherapy (p<.001 for all). Patients treated in VHA facilities located in urban areas were more likely to initiate pharmacotherapy than those in rural/highly rural/insular areas (AOR=.95, CI=.93, .97, p<0.001).

Aim 4 (Cost Implementation) Findings. After one year, recently screened tobacco users (N=589,862) who initiated tobacco cessation pharmacotherapy incurred $143.79 in additional treatment costs and had a 3.1% absolute increase in tobacco quit rates, an incremental cost-effectiveness ratio of $4,705. This ratio is significantly cost-effective compared to a willingness to pay of at least $5,600/quit (p<0.001).

Aim 5 (Maintenance Recommendations).
-Although rates of tobacco cessation pharmacotherapy in the VHA have increased substantially and are higher than some other published studies (e.g., 18.7%), much higher rates may be achievable, as 33.7% was achieved in patients of health maintenance organizations participating in the clinical trials network of the National Cancer Institute. Particular subgroups of Veterans (e.g., those with schizophrenia, other psychosis, most medical conditions, males, and Hispanics,) could be targeted to improve the reach of tobacco cessation pharmacotherapy.
-Continuation of the policy to distribute tobacco cessation pharmacy is warranted, as quit rates were higher among those receiving it. Quit rates among those who did not receive pharmacotherapy were higher than shown in other studies, suggesting that behavioral counseling may also have an impact or some are simply opting to quit on their own.
-Targeting training to low prescribing providers (e.g., non-pharmacists), clinics (e.g., outpatient psychiatry), facilities (VHA medical centers), and those in rural areas may increase the adoption of provider prescribing of tobacco cessation pharmacotherapies.
-Since tobacco cessation pharmacotherapy is one of the most cost-effective medical interventions, efforts should be made to maintain and enhance the program even further.

IMPACT:
Impact of Aim 1 (Reach). The VHA has substantially increased the reach of tobacco cessation pharmacotherapy. However, initiation of tobacco cessation pharmacotherapy among underserved subgroups (e.g., those with most medical comorbidities, some psychiatric comorbidities, and demographic subgroups) could be improved.

Impact of Aim 2 (Effectiveness). The effectiveness of cessation pharmacotherapy supports continuation of the VHA policy directing that tobacco cessation pharmacology treatment be made available for VHA patients interested in quitting.

Impact of Aim 3 (Adoption). Particular groups of providers, clinics, and facilities may need to be targeted and trained to increase provider adoption of tobacco cessation pharmacotherapy.

Impact of Aim 4 (Implementation/Cost). Tobacco cessation pharmacotherapy distributed in the VHA delivers sufficient value to justify its cost.

Impact of Aim 5 (Maintenance). The VHA has made substantial impact in reaching Veterans with tobacco cessation pharmacotherapies, improving quit rates, improving provider prescribing, and providing cost- effective tobacco cessation treatment. These interventions, as well as targeting Veterans less likely to be reached, and providers, clinics and facilities less likely to adopt tobacco cessation prescribing practices, have the potential to improve quit rates even further, decreasing tobacco-related morbidity and mortality among Veterans who use tobacco that are treated by the VHA.

---

External Links for this Project

NIH Reporter

Grant Number: I01HX001764-01
Link: https://reporter.nih.gov/project-details/8864008

---

PUBLICATIONS:

Journal Articles

1. Barnett PG, Chow A, Flores NE, Sherman SE, Duffy SA. Changes in Veteran Tobacco Use Identified in Electronic Medical Records. American journal of preventive medicine. 2017 Jul 1; 53(1):e9-e18. [view]
2. Barnett PG, Ignacio RV, Kim HM, Geraci MC, Essenmacher CA, Hall SV, Sherman SE, Duffy SA. Cost-effectiveness of real-world administration of tobacco pharmacotherapy in the United States Veterans Health Administration. Addiction (Abingdon, England). 2019 Aug 1; 114(8):1436-1445. [view]
3. Duffy SA, Ignacio RV, Kim HM, Geraci MC, Essenmacher CA, Hall SV, Chow A, Pfeiffer PN, Sherman SE, Bohnert KM, Zivin K, Barnett PG. Effectiveness of tobacco cessation pharmacotherapy in the Veterans Health Administration. Tobacco Control. 2019 Sep 1; 28(5):540-547. [view]
4. Rozek LS, Virani S, Bellile EL, Taylor JMG, Sartor MA, Zarins KR, Virani A, Cote C, Worden FP, Mark MEP, McLean SA, Duffy SA, Yoo GH, Saba NF, Shin DM, Kucuk O, Wolf GT. Soy Isoflavone Supplementation Increases Long Interspersed Nucleotide Element-1 (LINE-1) Methylation in Head and Neck Squamous Cell Carcinoma. Nutrition and Cancer. 2019 Mar 12; 71(5):772-780. [view]
5. Robinson CH, Yankey NR, Couig MP, Duffy SA, Sales AE. The Veterans Health Administration Registered Nurse Transition-to-Practice Program: A Qualitative Evaluation of Factors Affecting Implementation. Journal for nurses in professional development. 2018 Nov 1; 34(6):E8-E22. [view]
6. Ignacio RV, Barnett PG, Kim HM, Geraci MC, Essenmacher CA, Hall SV, Chow A, Pfeiffer PN, Sherman SE, Bohnert KM, Zivin K, Duffy SA. Trends and Patient Characteristics Associated with Tobacco Pharmacotherapy Dispensed in the Veterans Health Administration. Nicotine & tobacco research : official journal of the Society for Research on Nicotine and Tobacco. 2018 Sep 4; 20(10):1173-1181. [view]

DRA: Cancer, Cardiovascular Disease, Lung Disorders
DRE: Treatment - Comparative Effectiveness
Keywords: none
MeSH Terms: none