(2003; 211 pages)
Smoking cessation is generally defined as complete abstinence from the use of smoked tobacco. The duration of the studies varied from 12 weeks to 24 weeks, and used patient self-report questionnaires or interview data to assess quitting smoking. Almost all studies confirmed the self-reported data using one or more of the following biological measurements: expired carbon monoxide £ 10 ppm from the quitting day until the end of treatment and follow-up (1,3,10 - 12,15,17,21,25,26,28 - 32), salivary cotinine levels ≤ 20 ng/ml (3,11,13,16,24,28,31,33,34) and urinary cotinine levels of 317 ng/ml or less (21).
Adherence to smoking cessation therapy. The most widely used definition of adherence to treatment was "using the nicotine replacement therapy continuously at the recommended dose in the instructed manner for the entire 16-h (17) (or 24-h) time period" (1,10,12,13,17,20,29,30,32,35,36).
Some studies assessed adherence by comparing the number of used and unused systems returned each week with the number of days that had elapsed between visits (18,21,29). Others counted the total number of days on which patients did not use the systems during the treatment period, more than 5 days missed, or not wearing patches at night, were considered nonadherence (7).
Others defined adherence as "perfect compliance with treatment protocol and/or not missing any scheduled follow-up visits" (1,8,18). Bushnel et al. defined adherence as attending ≥ 75% of smoking cessation classes (26).
Few reports provided detailed data on adherence such as number of prescribed doses taken during a monitored period, monitored days during which the correct number of doses were taken or whether or not the prescribed intervals between doses taken were respected.
Drop-out. Patients may drop out from treatment for several reasons. These include patient-related factors, physician decision and adverse effects of the drug. Regardless of the reason for dropping out, patients who do so are usually found to be smoking at follow-up (25).
The way in which dropouts are handled can make it difficult to compare studies in this area. It is important to consider the reasons for dropping out to achieve accurate estimates of adherence. Those who drop out for reasons related to the treatment need to be distinguished from those who dropped out for reasons related to the study itself. Some patients drop out because they experience adverse events or withdrawal symptoms. As with studies in other therapeutic areas, these patients should be classified as nonadherent. Another important reason for dropout is the failure to stop or reduce smoking despite following the treatment. Many relapsed smokers stop using the prescribed NRT (37) when they fail to quit smoking despite having been adherent to NRT (21,36).We consider that these patients should be counted as treatment failures for the purpose of calculating smoking cessation rates, but not for adherence rates. Side-effects were the main reason given for dropout in the studies reviewed (1,9,11 - 13,16 - 18, 22,35 - 40). Other patient-related reasons for stopping therapy were failure to recall the receipt of a prescription (20), unwillingness to continue in the study (1,9,10,13 - 17) lack of a self-perceived need for treatment and lack of a perceived effect of treatment (1,9,13,16,36 - 40). Physicians reported discontinuation of therapy due to lack of efficacy or complete failure to stop or to reduce smoking after therapy had been started (1,3,8 - 10,17,18,21,22,24,29,36 - 38,41 - 43) and elevated carbon monoxide (17).