RESPONSE TO KOTZ ET AL. (2011): ESTIMATING THE RATE OF USAGE OF VARENICLINE AND OTHER MEDICATION FOR SMOKING CESSATION
Article first published online: 27 JUL 2011
Addiction © 2011 Society for the Study of Addiction
Volume 106, Issue 10, page 1869, October 2011
How to Cite
LANGLEY, T. E., HUANG, Y., MCNEILL, A., COLEMAN, T., SZATKOWSKI, L. and LEWIS, S. (2011), RESPONSE TO KOTZ ET AL. (2011): ESTIMATING THE RATE OF USAGE OF VARENICLINE AND OTHER MEDICATION FOR SMOKING CESSATION. Addiction, 106: 1869. doi: 10.1111/j.1360-0443.2011.03524.x
- Issue published online: 14 SEP 2011
- Article first published online: 27 JUL 2011
Our study  and the study by Kotz and colleagues [2,3] used different outcome measures over different time-periods, from data sources with different strengths and limitations; this is reflected in the differences in the results of the two studies. Kotz et al. used a series of monthly household surveys providing self-reported information on quit attempts in the past 3 months over a 3-year period. These data are very detailed, but cover a short time-period and may be prone to recall bias and significant sampling error due to small sample sizes. We used a validated measure of smoking cessation medication prescribing from primary care data in a monthly sample size of approximately 2.5 million over a much longer time-period, 9 years .
We acknowledge that not being able to look at prescribing in smokers is a limitation of our study; however, given the probably limited impact of changes in prevalence on the results of our study, our findings provide important information with regard to the public health impact of the introduction of varenicline. Rates of prescribing for smoking cessation medication in the overall population did not increase; therefore it is unlikely that, contrary to what may have been expected, the availability of a new medication increased quit attempts. By contrast, Kotz et al. looked at medication use in smokers who made a quit attempt and found a non-significant increase in medication use. We do not believe that not looking at the ‘population at risk’ of using smoking cessation medications is a limitation of our study; rather, it offers information as to the impact of varenicline on quitting in the general population, as opposed to information of its impact on quitters.
Kotz et al. explored the effect of varenicline on quit attempts by looking at changes in self-reported quit attempts during the study period. They found a decrease in quit attempts during the study period. Evidence suggests that English smoke-free legislation, which was introduced in July 2007, increased quitting behaviour before and around the time of the ban [5,6]. Their study included only half a year of data prior to legislation; therefore the decrease in the rate of quit attempts observed may reflect a return to previous rates. Our study used data covering the 7 years before the introduction of varenicline, enabling us to use powerful time–series methods which allowed us to take account of underlying trends in prescribing.
The apparent discrepancy between the results of our own study and that by Kotz and colleagues serves to highlight the strengths and weaknesses of different study designs, data sources and outcome measures that can be used to evaluate tobacco control initiatives, and the need for careful interpretation and triangulation of such studies to ensure appropriate conclusions.
Declarations of interest
Tim Coleman declares that, in the last 5 years, he has been paid for consultancy work by Johnson and Johnson and Pierre Fabre Laboratories (manufacturers of nicotine replacement therapy). However, this manuscript has not been discussed with any third parties. None of the other authors declare any potential financial conflicts.