BACKGROUND/RATIONALE:
Smoking is the leading preventable cause of death in the US, and health care systems are able to reach a large number of smokers and effectively engage them into evidence-based tobacco treatment. Effective cessation approaches include medications and behavioral treatment (including face-to-face counseling, telephone counseling and text messaging). While delivery of smoking cessation medications has increased, it has proven more difficult to increase the delivery of behavioral treatments. Within primary care, health care providers are effective at counseling smokers to quit, yet few providers actually deliver longitudinal counseling due to time pressures and competing priorities [1]. The two main evidence-based approaches available in the U.S. for behavioral tobacco dependence treatment include Quitlines (to deliver telephone counseling) and text messaging programs. Tobacco Quitlines are effective and our prior studies have explored different approaches to increasing use of Quitlines, yet many questions remain about how best to do this within health care system. Similarly, text messaging is effective at helping smokers to quit, but is rarely used in health care. While 11% of smokers reported having used a quit smoking program that involves text messaging no studies to date have examined health care-based approaches to increase engagement in text messaging programs, such as VA's SmokefreeVET. Our prior research shows that nearly all smokers are capable of receiving text messages and most feel that it would help them to quit [2]. National mandates call for health care systems to refer all smokers to treatment, yet there is little guidance on how to achieve this objective. Studies from behavioral economics and other fields suggest that how the referral is framed is extremely important [3]. Systems typically use an opt-in approach where the default is "no treatment", unlike conditions such as diabetes or hypertension where we treat people unless they opt-out of treatment. An opt-out approach has proven much more effective in other settings, such as organ donation, screening for sexually transmitted diseases and recruitment for research studies. Aside from an uncontrolled UK study among pregnant women, no studies have tested an opt-out approach for smokers in ambulatory care [4].

OBJECTIVE(S):
Our specific aims are: (1) To compare the effectiveness of an opt-out approach to an opt-in approach at helping patients quit smoking in primary care; (2) To compare the costs and benefits of an opt-out approach to an opt-in approach to smoking cessation in primary care.; and (3) To evaluate patient and staff factors affecting implementation of the opt-out and opt-in interventions in primary care.

METHODS:
We propose a Type I hybrid effectiveness/implementation study to evaluate two population-based approaches for increasing use of Quitlines and text messaging at two VA sites. Specifically, we will test the default bias, examining whether an opt-out approach to referral is more effective than an opt-in approach. We will randomly assign teams to either an opt-out or opt-in approach to referring smokers to treatment. In the opt-out approach, the default is that everyone is referred to treatment unless they actively choose not to be. In the opt-in approach, people are offered treatment but must actively choose to enroll in it.

FINDINGS/RESULTS:
We have no findings or results to report at this time, as we have not yet implemented the intervention.

IMPACT:
If effective, this approach would provide a simple, easily scalable method for increasing engagement in behavioral treatment and for boosting long-term abstinence rates.

---

External Links for this Project

NIH Reporter

Grant Number: I01HX002455-01A1
Link: https://reporter.nih.gov/project-details/9504820

---

PUBLICATIONS:

None at this time.

DRA: Cancer, Cardiovascular Disease, Lung Disorders
DRE: Prevention, Treatment - Implementation, TRL - Applied/Translational
Keywords: Comparative Effectiveness, Implementation, Models of Care, Substance Use and Abuse
MeSH Terms: none