Paying Smokers to Quit: A Report from the New England Journal of Medicine

Should employers give their employees hard cash to stop using tobacco? The Disease Management Care Blog doesn’t know the answer to that either, but it’s a little closer to understanding the numbers thanks to this publication by Volpp and colleagues appearing in the latest issue of the New England Journal of Medicine. Of approximately 1900 smoker-employees of an unnamed company with multiple worksites, about 800 agreed to be randomly assigned to ‘community based tobacco cessation resources’ within 20 miles of each of the work site (442 persons) or being assigned to the community resource plus being paid $100 if the program was completed, $250 if there was tobacco cessation within 6 months of study enrollment and another $400 if still off tobacco 6 months later (436 persons). Participants were not required to actually use the community resource; it looks like they could quit on their own if they so chose. Cessation was confirmed with a highly accurate test for the presence of cotinine in the urine or saliva.

The bottom line quit rate at 9-12 months was 14.7% in the group that was incented vs. 5.0% in the group that had no incentive. There were only two urine or saliva samples that were positive for cotinine.

The DMCB counted 47 persons completed the community resource (at $100 each or $4700), 91 had cotinine confirmed abstinence within 6 months (at $250 each or $22,750) and 64 remained abstinent at about 12 months ($400 each or $25,600), yielding a total outlay of $53,050 or $828 per smoker who quit.

So if an employer has a thousand employees, and about 15% or 150 smoke (a not unusual prevalence rate) and approximately half agree to participate (75) and the quit rate is about 15%, that means that 11 persons would quit at a total cost to the company of about $9000. The authors quote a study that contrasts the $828 figure with a report from the Centers for Disease Control that states that active smokers experience $1,760 in lost productivity and $1,623 in excess medical expenditures per year.

Take away lessons from the Disease Management Care Blog:

Quit rates based in physician offices that are supplemented with the use of pharmacotherapy typically range from 11% to 30% at one year. The quit rate of about 15% in this program seems comparatively low. However, it has the advantage of offering more persons access to a tobacco program with a respectable outcome. As a result, while the relative number of quitters is lower, the absolute number is probably higher than what can be obtained in the traditional health care system.

While the number of persons trying to game the system seemed to be low (apparently saying they stopped and submitting a sample), that may have been helped by the sentinel effect of urine or saliva testing. This would seem to be an important part of any company-sponsored tobacco cessation program that involves cash incentives.

The cost of $9000 described in the scenario above is the direct cost to the company. Other costs, such as cotinine testing, physician services or drug use (for example, nicotine replacement therapy) represent other costs, especially in a self-insured company.

Last but not least, this appears to be another business opportunity for disease management organizations. This is not the first program to use incentives, but we now have a better idea on the role of cash in incenting persons to quit. DMOs can help recruit smokers, arrange for the community-based referrals, arrange the cotinine testing and adjudicate payment.

Chantix vs. Nicotine Patch for Tobacco Cessation & What About Disease Management?

The Disease Management Care Blog was struck by Wall Street Journal Health Blog’s (WSJHB) coverage of another varenicline (Chantix) peer review publication. Reported in Thorax, study participants were randomly assigned to Chantix or to a nicotine patch. While early abstinence rates favored Chantix, the one year quit rates (26% vs. 20%) failed to achieve statistical difference (p=.056). Unsurprisingly, it wasn’t Pfizer that alerted the WSJHB but GlaxoSmithKline. It makes the competing nicotine patch.

WSJHB writes the borderline p value suggests that Chantix was ‘a little bit’ more effective. As in, um, the results were a little bit statistically significant. Sorry guys, the interpretation is that the Chantix vs. patch rates did not achieve the conventionally accepted threshold that distinguishes random chance from a real effect.

This is bad news for Chantix’s manufacturer, Pfizer. Looks like the Mayo Clinic won’t need to change its on line information all that much. Many managed care organizations provide access to nicotine patches via vouchers or discount programs. Since the patch arguably works as well as Chantix, these patch promotion programs will continue and Chantix will continue to be subject to preauthorization that is often dependent on trying the patch 1st.

The DMCB took some additional time to review some other publications on Chantix here, here and here. What was striking about these studies was that Chantix’s success was always accompanied by multiple follow-up 10 minute tobacco cessation office visits. The DMCB interprets this as showing that Chantix’s quit rates are intertwined with a significant degree of ongoing counseling. In fact, we really don’t know how well Chantix works without counseling. What’s more, tobacco cessation guidelines echo the necessity of prescribing tobacco cessation medications in conjunction with close follow-up:

'There is a strong dose-response relationship between the intensity of tobacco dependence counseling and its effectiveness. Treatments involving person-to-person contact (ie, via individual, group, or proactive telephone counseling) are consistently effective, and their effectiveness increases with treatment intensity (eg, the number of minutes of contact). If the patient agrees to attempt cessation, the clinician should then assist in making a quit attempt and should arrange for follow-up contacts to prevent a relapse. The treatment of tobacco dependence, like the treatment of other chronic diseases, requires the use of multiple modalities. '

This is echoed in the Chantix web site and the package insert. Yet, despite the evidence, it’s unlikely that the market or insurance regulators will tolerate making access to Chantix contingent on the availability or use of counseling.

Other options include:

Paying the physicians: In contrast to other preventive care services, tobacco counseling is already covered by many insurers, including Medicare. The problem is that physicians are not taking advantage of it. Maybe it’s not enough money. Alternatives include linking the payment to presence of a Patient Centered Medicial home, pay for performance, a global fee for an episode of care or compensation referring Chantix users to another entity (hint, there's one in each state) that can provide the counseling.

Deploy disease management: Note that the counseling doesn’t necessarily have to be personally provided by a physician or in person. This is an opportunity for the Pharmacy Benefit Managers to provide the service or to partner with a disease management organization that can provide the counseling. Given the scalability and expertise of existing disease management programs, the DMCB suspects this is a more available and cost-effective option.