Minutes (07-12-01)

1.	No formal group meeting occurred. Contributions to this report have been made by Cathy Backinger, Belinda Borrelli, Trish Jordan, and Geof Williams. We have voted to change our groups title to "Tobacco Dependence Outcome Group" from Smoking Outcome Group to be consistent with the PHS 2000 Guidelines. To keep in line with this, we also will adopt “tobacco user” rather than “smoker” as the way to refer to people who use tobacco regularly, because tobacco in all forms has important health consequences. (We don’t know what to call those exposed to second hand smoke or second hand tobacco juice.)
2.	In addition, the following changes in outcome measurement are offered for consideration based on the recommendations of the Society of Research on Nicotine and Tobacco (SRNT) subgroup headed by John Hughes at the University of Vermont. They have drafted an article discussing tobacco outcomes and making recommendations for their measurement. This paper has not been accepted for publication yet, so we cannot offer you the full text. You could e-mail John and ask for the text at this address john.Hughes@uvm.edu. We found it very helpful as it offered several insights to explain some of the difficulties we had establishing common outcomes across our groups. Our initial consensus as a group, led to making 6, and 12- or 18-month 7- and 30-day point prevalence (measured by the item: "Have smoked a cigarette, even a puff, in the past 7 days; if no in the past 30 days?" asked at 6 , 12 or 18 months), and the stage of change measures as our standard outcomes across sites. The SRNT paper advocates for continuous abstinence or prolonged abstinence be the primary measure and point prevalence be the secondary measure. Projects may not be able to add this measure at this time, but if you can, please consider doing so. Point prevalence is also an accepted measure — please do not remove it from your current outcomes, the issue is whether you can add prolonged abstinence. With this in mind, we have amended the "mid-sized" and "luxury measures" to include the prolonged abstinence measure. We have also added the tobacco use question from the mid-sized and luxury models to the economy model in order to detect use of all forms of tobacco in all measures. Therefore, if you used the economy model, consider adding the tobacco use question (# 8 in the amended economy model). The SRNT paper also makes the following important distinction between types of studies that tells us when to do the follow-up's (i.e., when to ask the follow-up question). So, we ask you first to identify which category of study you are and then to measure your outcomes at the recommended times. The paper notes three types of studies-only two of which are relevant to us. The timing of the follow-up measuring the prolonged abstinence, and point prevalence vary based on the type of trial you are conducting. The two types of trials are called "aid to cessation" trials, and "cessation induction" trials and they are described below. "Aid to cessation" trials typically test a treatment only among smokers who are currently willing to set a quit date (e.g., group behavioral therapy, or a medication). The experimenter sets a quit date for everyone, and therefore they know the intended quit date for everyone in the trial. Outcomes are then measured relative to that quit date (continuous abstinence), or that quit date plus a reasonable grace period-usually 2 weeks (prolonged abstinence). SRNT recommended outcomes for aid to cessation trials were: Primary outcomes: Prolonged abstinence after two week grace period was the primary outcome at 6, and 12, or 18 months. Prolonged abstinence is assessed with the following questions "Have you smoked even a puff since [date of the end of the grace period]? If yes, when did that first occur? Treatment failures are defined as those people who answer yes to the question above and then answer yes to this question "Have you smoked for 7 consecutive days or one day in 2 consecutive weeks? If yes, "When did that first occur?" Asking the questions in this manner allows you to identify the week of relapse so that it can be plotted using survival curve analysis. Survival curve analysis was recommended by the SRNT committee as an excellent method to report results. You can also easily calculate the duration of a quit attempt from this, which is correlated with future success. Secondary outcomes would be 7 and 30 day point prevalence asked at 6, and 12 or 18 months. Point prevalence is assessed with the following question "Have you smoked a cigarette, even a puff, in the past 7 days; if no in the past 30 days? asked at 6 , 12 or 18 months. The article also recommends repeated point prevalence which is point prevalence measured at several times (e.g., did the participant reported that they had not smoked a cigarette, even a puff for 7 days; or if no for 30 days at 6, and 12 months). "Cessation induction" trials test a treatment among all smokers (not just those willing to set a quit date). In these trials, the participants determine if and when they set a quit date. Therefore, follow-up is tied to the onset of the initiation of treatment. SRNT recommended outcomes for cessation induction trials are: Primary outcomes: Prolonged abstinence of 6 months or greater (for us 12, or 18 months) from the start of the intervention was the primary outcome. Prolonged abstinence is assessed with the following questions "Have you smoked even a puff since [date of the end of the grace period]? If yes, when did that first occur? Treatment failures are defined as those people who answer yes to the question above and then answer yes to this question "Have you smoked for 7 consecutive days or one day in 2 consecutive weeks? If yes, "When did that first occur?" Secondary outcomes: Point prevalence of 6 months, and 12 or 18 months from the start of the intervention. . Point prevalence is assessed with the following question "Have you smoked a cigarette, even a puff, in the past 7 days; if no in the past 30 days? asked at 6 , 12 or 18 months. SRNT also recommended anyone using tobacco of any type (cigars, chew, pipe) be considered a smoker, even if they report successfully quitting cigarettes. Thus, if someone has quit smoking cigarettes, but now smokes cigars or a pipe this would be considered a treatment failure. One member of our group (Belinda Borelli), indicated that follow-ups in cessation induction trials are tricky (we all agreed). In order to try to make cessation induction trial outcomes more comparable to public health trials, their site decided to tie follow-ups to the end of the treatment period. In this trial a maximum time of 8 weeks is allowed for the intervention-so all outcomes are relative to that time. Another site (Geof Williams) has a 6 month intervention. We tie our outcomes to the initiation of the intervention (date of randomization is the first day of treatment). In addition, all people who have reported not smoking at 6 months from randomization, are asked their stop date, and are called back for a non-biochemically validated point prevalence, and prolonged abstinence outcome 6 months from their stop date. Neither system is perfect, and each study site will need to decide how to proceed. We did all agree to recommend prolonged abstinence and point prevalence as the preferred measures for all BCC sites with smoking interventions.
3.	Finally, Geof Williams mentioned at the BCC meeting that we have just had an article accepted at Health Psychology in which we assessed three indicators of smoking abstinence and combined them in a latent variable. The model fit nicely, and as a clinician he likes this measure because two of the indicators reflect days not smoked during the follow-up time. Our three indicators for smoking change are: 7 day point prevalence (cotinine validated): "Have you smoked a cigarette, even a puff, in the past 7 days” yes or no What was the date of the last cigarette you smoked (you could add “even a puff” to have this match the prolonged abstinence measure)? __ Month __Day __Year. The indicator is calculated by the number of days difference between the date of completing the questionnaire and the date of not smoking. In the past six months (or since the beginning of the study), what was the single longest time that you have not smoked a cigarette, even a puff? __Years _months __weeks and/or __days? Alternate suggestions for indicators 2 and 3 are: Since our last contact on ____, have you smoked any cigarettes at all even a puff?” Since our last contact on ___, what was the single longest time that you have not smoked a cigarette, even a puff? You will need to know if the time period they specify is continuous with non-smoking days before the “last contact” or not if you are to calculate longest number of days not smoking.
4.	The following three references should be added to our BCC reference list: Hughes, J. R., Keely, J. P., Niaura, R. S., Ossip-Klein, D. J., Richmond, R. L., & Swan, G. E. Measures of abstinence in clinical trials: Issues and recommendations. SRNT Subcommittee on Abstinence Measures.(Unpublished manuscript-contact Dr. Hughes for permission to cite at john.Hughes@uvm.edu) Velicer, W. F., & Prochaska, J. O. (in press). A comparison of four smoking cessation outcome measures. Nicotine and Tobacco Research. Williams, G.C., Gagne, M., Ryan, R. M., & Deci, E. L. (in press). Facilitating autonomous motivation for smoking cessation. Health Psychology.