Since nicotine replacement therapy (NRT) was introduced 30 years ago its full potential has remained underdeveloped and under-researched. A new drug, Varenicline, acting on the same nicotinic receptors, appears to give better efficacy [1]. For those wishing to stop smoking, the NRT marketing licence still indicates a restricted regimen that is the same for those self-treating by buying over the counter and those in the care of specialist clinicians: (i) start using NRT on the day that smoking stops; (ii) use a limited dosage while not smoking; (iii) stop using NRT if smoking resumes; and (iv) use only for 10–16 weeks, regardless of progress. There is no evidence that this protocol is optimally effective at a population level, and it is unlikely to be flexible enough to meet the differing needs of individual smokers. The only options involve a choice of route by which nicotine is absorbed (skin, nose, throat or mouth), and whether nicotine is taken passively/chronically (skin patch) or actively/acutely (gum, lozenge/sublingual tablet, inhalator, nasal spray). Regardless of product or route, nicotine is absorbed far more slowly than when inhaled via cigarettes and in practice nicotine levels from this NRT protocol are, at most, about half those from smoking [2].

Paradoxically, for those who do not wish to stop smoking in the near future, the NRT licence allows long-term use while still smoking as a means of potentially reducing cigarette consumption [3]. The indications for those trying to stop and those not trying to stop, although apparently very different, are both highly conservative in some respects. Neither allows much freedom with dosage nor long-term use after smoking has ceased. Ironically, because there can be no effective control over the dose and duration of NRT use when it is purchased over the counter, smokers have real freedom with medicinal nicotine only when they are not in the care of clinicians. The reason manufacturers and clinicians have tolerated such restrictions, and not explored the potential for improved efficacy through more liberal regimens, probably lies with the influence of some regulators and tobacco control advocates. These, in the past, have viewed smoking as a habit maintained by the force of the tobacco industry, rather than as an addiction to inhaled nicotine, that for many requires clinical intervention on a par with that for other addictive substances. The manufacturers themselves can also be criticized for their timidity in not pressing the regulators to broaden the clinical indications of existing products, and for not developing improved products. In this issue, Shiffman & Ferguson (p. 557) conduct a meta-analysis of four studies where smokers were treated with the nicotine patch prior to their quit attempt, and provide good evidence to support this more flexible use of NRT [4].

The level of ‘dependence’ on smoking is measured most commonly by scores on inventories such as the Fagerström Test for Nicotine Dependence and the Heaviness of Smoking Index [5]. Such measures are weighted towards daily cigarette consumption and early morning smoking, both underpinned by the degree of reinforcement from inhaled nicotine. In numerous cessation studies higher scores on these measures have been associated with failure to quit. Indeed, ‘dependence’ appears to be the most consistent of a relatively small set of antecedent characteristics that predict whether a smoker is likely to succeed or fail in a quit attempt. Others include gender, age, social grade, motivation to stop, partner smoking and self-efficacy. As motivation is usually high prior to a quit attempt, ‘dependence’ may be the easiest predictor to modify ahead of a quit attempt, in the hope of improving the chance of subsequent success. Essentially, this means reducing smoking behaviour and consumption.

The practice of encouraging a reduction in consumption before a quit attempt appears to have been discounted prior to the introduction of NRT. This may have been because when smokers try to cut down they tend to smoke the remaining cigarettes more intensively to maintain adequate nicotine intake, and so each cigarette may take on a proportionately higher value and be more reinforcing. Hence, when smokers have to make an unaided effort to cut down there is no effective reduction in ‘dependence’. The alternative is to provide a means by which little or no effort is required to reduce both the habitual behaviour and the reinforcing value of inhaled nicotine without experiencing aversive withdrawal symptoms. As early as 1942 Johnston observed that, by taking nicotine through means other than smoking, both the desire to smoke and tobacco consumption are reduced [6]. Notably, in the three studies under review where cigarette consumption was recorded there was a significant reduction during the period of patch pretreatment. Unfortunately, the potentially more revealing change in nicotine intake from smoking during the pretreatment phase was not reviewed. Notwithstanding the absence of evidence for the mechanism, the effect due to nicotine pretreatment found across the four studies was a twofold improvement in cessation—an effect similar to that for 12 weeks of patch treatment starting on quit day [7].

However, unlike most treatment reviews that demonstrate efficacy, this one does not provide sufficiently homogeneous data to formulate a simple new treatment protocol for clinicians and smokers. The four studies differed with respect to dosage, length of pretreatment period and post-quit treatment, follow-up duration, use of placebo and, most importantly, the inclusion of concomitant treatments (mecamylamine and low-yield cigarettes), both in the pretreatment phase and beyond. Only Schuurmans et al.'s study provides evidence for the efficacy of pretreatment patch use free of the presence of other concomitant interventions [8]. Therefore, we have some way to go before guidance can be given to clinicians as to the optimal NRT pretreatment protocol. Another trial, similar in design to that of Schuurmans et al., would help in this respect. In the meantime, there is no reason why smokers should not be prescribed NRT before they make a quit attempt, just as they can be when trying to cut down.