What's new in the 2008 US Guideline?

The key recommendations and a commentary

by Paul Aveyard and Jonathan Foulds

Paul Aveyard, NIHR Career Scientist, UK Centre for Tobacco Control Studies, Primary Care Clinical Sciences, University of Birmingham, Birmingham, B15 2TT, UK.
p.n.aveyard@bham.ac.uk

Jonathan Foulds, Professor, UMDNJ-School of Public Health Director, Tobacco Dependence Program, 317 George Street, Suite 210, New Brunswick, NJ 08901-2008, USA.
fouldsja@umdnj.edu

This Ask the Experts piece was written for treatobacco.net and peer reviewed before being published online in March 2009.

Suggested citation: Aveyard P, Foulds J. The 2008 US Clinical Practice Guideline: the key recommendations and a commentary. www.treatobacco.net 2009 Ask the Experts. Accessed 01 March 2009.

The US, English and many other guidelines can be downloaded from the www.treatobacco.net resource library, and the site provides many other useful links, including direct links to the Cochrane Reviews.

Summary

The purpose of this article is to summarise the US guideline, to comment on the process of its production and its conclusions and recommendations. In particular we examine some of the methodological processes involved in selecting literature and in commissioning new reviews of the literature. The US guideline quotes point prevalence abstinence rates in the intervention group whilst other guidelines quote different statistics as the metric of success. In this article we define these statistics and show their inter-relation, in particular drawing a distinction between rate ratios and odds ratios. We also discuss several key recommendations of the US guidelines, including the recommendation that more intensive behavioural support is more effective, which draws on indirect comparisons, and which we contrast with data from direct comparisons. We discuss the recommendations on brief advice, questioning whether recommending such extensive brief interventions and motivational enhancement are truly evidence-based, and the similarity of recent guidelines on smoking reduction. This is an interesting area where the evidence from trials is conclusive and yet guideline developers have declined to recommend these treatment programmes.

Introduction

An excellent method for creating clinical guidelines is well established [1]. An expert panel convenes and commissions reviews of the evidence including meta-analysis. Guidelines are produced based on consideration of the evidence, with expert opinion forming the background and filling in the bits where the evidence just does not quite reach. In many guidelines, the strength of evidence supporting recommendations is graded from A to C, where A represents evidence supported by a statistically significant effect in pooled randomized controlled trials and C represents expert consensus only. This method, or variations on it, is commonly used in developed countries, but perhaps it is not a method that can be so easily used in low income countries. It is expensive, costing hundreds of thousands of dollars to fund the time and expertise needed to review the literature to create evidence-based guidance, and to bring experts together to digest the evidence and formulate the guidance. Arguably too, where up-to-date reviews of key areas exist, it is more cost-effective to rely on these than commission reviews specifically for the guideline.

The US Clinical Practice Guideline [2] followed this exemplary process. The amount of literature reviewed and its breadth and depth distinguishes it from the previously published UK guidelines [3,4]. It also distinguishes it from the newly updated guidance for commissioners and providers of health services in England and Wales issued by National Institute for Health and Clinical Excellence (NICE) [5]. Even though NICE commissioned systematic reviews of key elements of the evidence, the scope of the work covered in these commissioned reviews was less than for the US Guideline. The NICE guidance instead drew on established evidence such as Cochrane reviews. The US approach took nothing as a given, relying on specifically commissioned external reviews.

In this commentary, we aim to give a summary of the guidance and comment on certain aspects of the methods of producing the guideline and some of the recommendations that follow. Inevitably, because we are familiar with the UK guidelines, this means that we contrast the US and UK guidelines in places, although the main aim is to describe and comment on the US guidance.

Box 1. The process of producing guidance in England

England produced national guidance on smoking cessation for the first time in 1998 and updated it in 2000. The process was as follows. Three experts commissioned by a national body used pre-existing reviews of the literature. In particular the group drew on reviews by the Cochrane Tobacco Addiction Review Group, but also on reviews developed for the American guideline by the then Agency for Health Care Policy and Research (AHCPR). The resulting guidance was peer reviewed by a panel of 18 international experts and revised. The guidelines were then reviewed by twenty professional agencies with an interest in tobacco dependence treatment, which also contributed to their professional adoption. Finally, the guidance was subject to peer review by the journal Thorax as part of the publication process. The 1998 and 2000 guidelines can be accessed free in the www.treatobacco.net resource library.

Since the publication of the Thorax English guidelines the UK has adopted a new system for incorporating or rejecting treatments into its National Health Service (NHS) and producing national guidelines for all conditions, not just tobacco dependence. The organisation in England and Wales is called the National Institute for Health and Clinical Excellence (NICE). The NICE process was similar to the Thorax process in some ways, although the panel was much larger. It too drew on established reviews of work, such as the Cochrane reviews. However, it commissioned several reviews from systematic review teams in universities, which the panel considered. The documents can be found at: http://www.nice.org.uk/guidance/index.jsp?action=byTopic&o=7376&set=true

Some key features of the US Guideline

Box 2. Key US Guideline recommendations

Tobacco dependence is a chronic disease.
Healthcare providers need to identify and document tobacco use at every opportunity.
Clinicians should encourage all smokers willing to make a quit attempt to use behavioural support and medication.
Brief interventions are effective and clinicians should use these at every opportunity.
Individual, group, and telephone behavioural support are effective and effectiveness depends upon intensity.
Numerous effective medications exist and they should be used in every quit attempt except where contra-indications exist.
Behavioural support and medication are effective alone, but the combination is more effective than either alone.
Telephone quitlines are effective at reaching diverse populations and should be promoted.
If a person is unwilling to make a quit attempt, clinicians should use motivational strategies to encourage quitting.
Healthcare providers should ensure that effective and cost-effective treatments for tobacco dependence are provided routinely.

One feature to note is that the US Guideline describes itself as about the treatment of tobacco dependence. This is not simply to bring in smokeless tobacco, which is explicitly incorporated, but also serves to emphasise that this enterprise is about treating a disorder, tobacco dependence, and not so much about interfering in a person’s lifestyle. We think this is a helpful approach, as some clinicians around the world still see treating smokers as not their business [6].

The guideline starts with ten key recommendations. The first emphasises that tobacco dependence is a chronic disease and that it may require long term treatment or repeated attempts before tobacco abstinence is achieved. This is a critical point and implies that clinicians should treat tobacco dependence just as they would hypertension, asthma, diabetes schizophrenia or any other serious chronic illness [7]. The second emphasises that clinicians should take every opportunity to treat tobacco dependence. This recommendation is probably the same in smoking cessation guidance around the world and is certainly made in the UK guidance. However, in the UK, the system for rewarding general practitioners (GPs) or family physicians for quality care pays GPs only for treating smoking in the context of other disorders for which smoking is a risk factor, such as ischaemic heart disease. This illustrates that, in many contexts, smoking has yet to shift from being treated as a lifestyle risk factor to a disorder worthy of treatment, like hypertension. This guideline is unequivocal in asserting that these interventions are effective and, in view of the US health care system, that insurers and service providers should include tobacco dependence treatments as part of their standard portfolio of care.

A key recommendation of the Guideline is that behavioural support or counselling is effective. The guideline recommends individual, group or telephone support and unequivocally concludes that intensity relates to efficacy: more is better. Also, two particular forms of support, problem solving (review of the problems a person is having and finding solutions), and social support are claimed to have particular evidence of efficacy. The guideline panel recommended nicotine replacement therapy (NRT), bupropion, or varenicline as first line medication choices, and certain combinations of these medications: patch plus either gum or nicotine nasal spray, patch plus inhaler, patch plus bupropion. The US Guideline specifically recommends long term (>14 weeks) combined patch plus another medication, again consistent with its view of treating tobacco dependence as a chronic illness (p109 and p113). However, the guideline notes ‘Thus, neither high-dose nor long-term patch therapy appeared to produce benefit above and beyond that of nicotine patch therapy at the regular duration (6–14 weeks) and dose (14–25 mg)’ (p 113), which is somewhat at odds with this recommendation. The evidence from clinical trials is that longer term treatment [8] or relapse prevention interventions [9] have proved ineffective thus far and this does not support long-term treatment, with a few exceptions. Whether better selection of patients and more sophisticated methods of chronic treatment will prove effective, as many feel is likely, remains to be seen.

The US Guideline highlights some new areas where the evidence was strengthened and changes made in comparison to the old guidelines. These are that behavioural support is effective and adds significantly to medication and that quitline support is effective. The other key update is the strengthened evidence that the system of care makes a difference to the likelihood that someone with tobacco addiction will receive treatment that helps him or her overcome that addiction. Commissioners and managers of health care are urged to work with health professionals to ensure systems are in place to identify and treat tobacco dependence. This means in practice noting whether or not a health service user is a tobacco user and ensuring that the system prompts health care professionals to raise the topic of tobacco dependence and provide interventions. It also provides evidence that clinicians who have been trained to provide effective tobacco dependence treatment obtain better smoking cessation outcomes than those who have not (p130).

Some methodological considerations

The guideline developers made some methodological decisions in selecting studies for meta-analysis with which not everyone will agree. Only trials were included. Studies needed to include six month follow up point prevalence abstinence data, but this was interpreted to mean follow up closest to six months provided that data were collected between five months and three years. However, this use of six month data as the primary source of evidence on long-term efficacy was checked by including longer term follow up data of about one year. This yielded similar results to those from the six month analysis. The panel included studies with self-report abstinence and biochemically confirmed abstinence, having previously checked that odds ratios using both types of data yielded similar results. However, for the outcomes in pregnancy, only biochemically confirmed abstinence was used because of the high rates of deception in pregnancy [10]. Intent to treat abstinence data were used in preference, although a few older studies that included only those followed up were also included. Most people would probably make similar decisions to these, although they may differ over detail. The Cochrane Tobacco Addiction Review Group uses similar approaches to analysing evidence.

One area of departure from the Cochrane methods and from the SRNT statement [11] on the optimum method for assessing the outcome of smoking cessation studies is the use of point prevalence abstinence as the outcome of preference. The reason for this is primarily expedience: more studies report this outcome than any measure of sustained abstinence. However, the guideline group advance arguments that this does not matter much. First, they suggest that point prevalence may actually be a better measure of sustained abstinence than is say prolonged or continuous abstinence. They argue that strict definitions of prolonged abstinence do not capture those who eventually establish a period of sustained abstinence, unlike point prevalence. The guideline developer may also have considered the fact that some treatments, for example varenicline [12] and NRT [13] make it less likely that a slip or series of slips turn into a relapse. Such an effect may be missed by some definitions of continuous abstinence. However, few people become continuously abstinent after early failure [14,15], so we are not convinced by this argument.

The second argument advanced for point prevalence over prolonged abstinence is that most people who relapse do so early and therefore that the rates do not differ much. In our view, the evidence suggests that about half of people who eventually relapse will have resumed smoking at six months (see relapse rates in these two reviews [16,17]), but we do not follow why this argues for point prevalence over prolonged abstinence as a preferred outcome measure. We would have used prolonged abstinence (based on the arguments above). However, as the primary output of the guideline reviews were odds ratios, the use of point prevalence figures may not matter. In the Cochrane review of NRT, which includes over 100 studies, the results of meta-analysis with point prevalence and prolonged abstinence were compared and found to yield similar results [8], and a similar result was found in another review [18]. The only issue would be the effect this might have on absolute effect size estimates.

Box 3. What is a quit rate?

Researchers in the field often use the term quit rate, but this can be unhelpful if it is not clearly defined in each case.

The different terms are most useful if we stick to a consensus definition, such as those suggested by the Society for Research on Nicotine and Tobacco (SRNT) [19].

Point prevalence abstinence means no tobacco use for a short period (typically 7 days) prior to a specific follow-up point.
Continuous abstinence means no tobacco use at all since the original quit date.
Prolonged abstinence means the same as continuous abstinence, except that a grace period of (typically) two weeks is allowed after quit date during which tobacco use does not invalidate subsequent abstinence.

Validated abstinence means that someone claiming abstinence has to provide some independent validation that they are not smoking or using tobacco. This typically requires blowing an exhaled carbon monoxide concentration below a cut-off (typically 8 or 10 parts per million) or having a salivary cotinine concentration below a cut-off (typically 15ng/ml). Even then, there is quite a lot of room for differences in interpretation as to who counts as having achieved prolonged abstinence and what happens to protocol violators, for example. An attempt to create consensus on this was published recently [20].

How do these definitions affect quit rates?

Imagine a trial of a new medication, in which 1000 participants are randomised to active or placebo arms, 500 in each arm.

At 6 months follow up, 175 people on active medication claim 7 day point prevalence quit rate of 35%, while 75 people on placebo claim a 7 day point prevalence quit rate of 15%.
Evidence shows that in such trials about 72% of people achieving 7 day point prevalence will have achieved prolonged or continuous abstinence [18]. Thus the self-reported continuous quit rates measured in this way may be 25% and 10%. However, some people who report abstinence (typically by telephone) will fail to attend to provide a validated sample and a few will be shown to be probably smoking. Imagine that about 30% of those claiming abstinence did not provide samples or failed validation. This would give validated prolonged abstinence rates or quit rates of 17% and 7%.

As well as expressing quit (validated prolonged abstinence) rates as 17% and 7% we could express the quit rate as the difference in quit rates or the absolute effect size i.e. 17%-7%=10%. Commonly this is turned into numbers needed to treat (NNT) by taking the reciprocal i.e. 1 / (0.17-0.07), which equals 10. This means that 10 people need to be offered treatment with the new medication to get one additional smoker to quit. This might be seen as expressing the true benefit of the medication, which feels a lot less pleasing than saying the quit rate is 35%.

The Thorax English guidelines focus more on the incremental extra quitters gained by providing the active treatment as compared with a control treatment. Thus it would quote the benefit of treatment as 10%, while the US guideline quotes the odds ratio and the percentage achieving point prevalence abstinence in the intervention group, which in this case might be 35%. Thus the same data can be presented differently and requires understanding to interpret carefully. The term quit rate could probably apply to any of these rather different percentages and so it should be avoided in scientific writing except where previously defined. (The US guideline would in fact have quoted validated point prevalence abstinence rates in this particular trial).

A further complication is that researchers differ in how they quote the ratio of quit rates. The most straightforward way is to use rate ratios or relative risks. If a person is twice as likely to quit smoking with a medication, this is another way of saying that the rate ratio or RR is 2.0. However, for various reasons related to statistical properties and the output of statistical programmes, researchers commonly report odds ratios (OR) and this was the output of the US guidelines. It is common for people to interpret odds ratios as rate ratios, meaning that if an odds ratio is 2.0 they say ‘people were twice as likely to quit smoking’. However, interpreting odds ratios as rate ratios can be misleading. The higher the abstinence rates, the more misleading the OR is as an estimate of the rate ration (Figure 1). When the quit rate in the control group is high (say more than 10%), the OR has no intuitive meaning. It is perfectly valid as a statistic, but lacks intuitive meaning.

Figure 1 The relation between odds ratios and rate ratios depends upon the quit rate in the control group

A key area for all smoking cessation guidance is brief interventions by physicians and other health care professionals to encourage quitting and support it. It is key because there is a large gap between what health professionals could achieve and what is being achieved in most health care systems. It is worth noting here the distinction between brief advice (to stop smoking) and brief interventions, which generally include brief advice, but often other elements too. It is noteworthy how guideline developers take evidence and create the guidance. For example, almost all the studies in the US Guideline meta-analysis contrast brief advice to quit with no advice. However, the Guideline suggests that doctors should implement the famous 5As:

ask (about smoking status)
advise (to stop)
assess (willingness to quit)
assist (prescribe, refer for support)
arrange (further support or to motivate reluctant quitters)

These are arguably useful, but they are not supported by the evidence reviewed, which largely concerns only the one A advice. Furthermore the Guideline recommends motivational enhancement, including specifically Motivational Interviewing (MI), as a specific technique for smokers who are not keen to quit. It notes that MI requires skill that needs to be learnt and cites a series of references to support its use. However, the evidence in this area was not subject to formal inclusion, appraisal, and meta-analysis as is the case in other areas, such as for medication. The 5Rs, which describe the content of motivational interviews, are supported by reference to two studies, which again have not gone through formal appraisal and meta-analysis. This is an important area, in which the US Guideline is asserting that ‘more is more’ ie. that more intensive interventions are more effective and that it is desirable to encourage health care systems to implement these.

The 'more is more' approach contrasts with the English Thorax guidelines, which simply recommend brief advice to stop and the offer of therapy [3,21]. However the updated UK NICE guidance on brief intervention suggests that a brief intervention on smoking cessation should be 5-10 minutes, thus also moving to the more is more approach [22]. A concern with this is that the biggest effect is probably between no intervention and some form of brief advice and support. We would argue that raising the bar higher could deter any kind of intervention and any recommendation to suggest more than the briefest intervention should be based on good evidence of incremental effectiveness. Other countries, such as New Zealand, have gone for very simple brief intervention recommendations, the ABC system [23]. It is an empirical question as to whether suggesting more extensive or less extensive brief interventions is the right strategy for guidelines to recommend. This is probably a key area to examine in future revisions.

Another area of difference between the Cochrane approach and the US Guideline is the use of indirect comparisons. A direct comparison occurs when a trial randomises participants to either intensive or brief behavioural support, for example. The relative effect of intensive versus brief support from that trial is securely estimated and unconfounded by differences in participant characteristics because of the randomisation. Indirect comparisons make use of a common reference standard. Thus a fair number of trials have randomised smokers to brief support or no support, while other trials randomised to intensive support or no support. The common feature is the reference group with no support. This common group can be used to estimate the effects of intensive support versus brief support by combining estimates from both groups of trials statistically. An important concern in indirect comparison is that this comparison is not randomised because estimates from different trials are compared. These trials may have different inclusion criteria and had participants with different likelihood of achieving abstinence. It is, however, possible to adjust these indirect comparisons for some of these differences and the use of random effects regression is recommended to reflect the greater uncertainty inherent in indirect comparisons [24]. The US Guideline used indirect comparisons in an informal way, without adjustment or without formally combining estimates. In the particular case of behavioural support, odds ratios for brief support and more extensive support were estimated from studies and tabulated. The odds ratios for more intensive support versus no support were greater than those for briefer support versus no support and this was used to conclude that more intensive support was more effective.

The Cochrane approach does use indirect comparisons on occasions, though less extensively than in the US Guideline. For example, the conclusions about the efficacy of different forms of nicotine replacement are based on trials contrasting the results of trials of, say, nicotine nasal spray with placebo to trials of nicotine gum with placebo [8]. Very few trials contrast nasal spray with gum. Sometimes, however, the use of indirect comparisons can give results that differ from trials in which the comparisons are examined directly and it is not clear when indirect estimation is likely to mislead [25]. For example, the Cochrane review of individual behavioural support examined the question of intensity of behavioural support. The authors did so using only direct comparisons from trials. There are only four such trials included in this section of the review in comparison to the 43 studies used in the US review in the analysis of session length, 35 on contact time, and 46 on number of sessions. The four studies in the Cochrane review combined showed no evidence that intensity mattered, with a rate ratio of 0.96, 95% CI 0.74 to 1.25, which excludes substantial effects of more intensive support [26]. Given the direct evidence suggests no effect, it is surprising that the US guideline did not comment on or examine this direct evidence and surprising that they made such strong and unequivocal recommendations in this area.

We are not, however, suggesting that this recommendation that intensity of support matters is necessarily wrong. In the UK, where there is a national system of smoking cessation clinics, the clinics run by specialists, which have higher contact time and frequently more sessions, have higher success rates than clinics run by generalists trained to give smoking cessation support [27,28]. Whether these differences in success are down to patient differences, therapist skills, or the intensity of the support is unclear. However these results tend to support the guideline recommendation that intensity matters, but the methodological point about integrating direct and indirect evidence remains. Perhaps the US Guideline could take on the challenge of formal methods for indirect comparison, which would be helpful to examine the strength of evidence in these important areas of clinical practice.

Another area that deserves attention is the recommendations on combination treatment. The guideline draws on a meta-analysis to recommend that bupropion plus nicotine patch is more effective than nicotine patch alone to recommend this particular combination of treatments. This contrasts with the UK NICE guidance, which recommend against combination treatment[5], which draws on the Cochrane review [29]. Combining four studies to the US Guideline's three, the Cochrane review concludes that there is insufficient evidence that combining NRT and bupropion is better than either treatment alone. Further evidence will be needed to see which of these conclusions is correct. However, both the US and the UK NICE guidance recommend combinations of NRT, for which there is evidence from clinical trials that combinations are more effective than single forms alone.

Reduction to stop

One of the areas where the US Guideline makes similar recommendations to the UK is in the area of what has been called nicotine assisted reduction to stop (NARS) or reduce to quit. In NARS trials, smokers who did not want or intend to stop smoking in the near future but were prepared to reduce their smoking were randomised to either nicotine replacement or placebo. They were seen regularly and given advice and support in gradually reducing their smoking. Surprisingly, these studies showed that smokers who declared no intention to quit were twice as likely to maintain long-term abstinence when using NRT as compared to placebo and the absolute benefit of treatment was similar to that seen in smokers who were prepared to set a quit date [30]. Cost effectiveness analysis showed this was one of the most cost effective interventions in modern medicine [30]. However both the US and the UK guidelines recommend against incorporating NARS programmes into the health systems. Concerns were expressed that these trials might have inadvertently enrolled smokers who were motivated to quit but saw such a programme as their only opportunity to achieve this and enrolling such people accounted for the efficacy. The US team were also concerned that messages about reducing might become subverted to messages that reduction itself improved health, which is probably not the case [31]. Clearly it is imperative to help smokers who are trying to stop, but there are many more who are not immediately trying to stop but would like to reduce and this is an area of pubic health importance where we need research to reassure guideline committees that such programmes have a place in tobacco control.

Summary

The US Clinical Practice Guideline is a guideline that saves lives. As a consequence of the authors' comprehensive efforts to collate and synthesise research in this area, policy makers and clinicians will have an easier job of providing treatment for tobacco dependence. Compared with hypertension, efforts to tackle tobacco use dependence would save more lives [32,33], but these two disorders do not get equal attention from medical practitioners and other health care professionals and these guidelines should help redress this imbalance. Of course that will require the guideline recommendations to be implemented. For any researcher in this area, the guidelines also provide a valuable summary of the state of the research in the area of clinical treatment and policy. We welcome them.

References

Guyatt GH, Oxman AD, Kunz R, Falck-Ytter Y, Vist GE, Liberati A et al.: Going from evidence to recommendations. BMJ 2008, 336: 1049-1051.
Fiore MC, Jaen CR, Baker TB, Bailey WC, Benowitz NL, Curry SJ et al.: Treating tobacco use and dependence: 2008 update. U S Department of Health and Human Services 2008, 1-256.
Raw M, McNeill A, West RJ: Smoking cessation guidelines for health care professionals. Thorax 1998, 53 (Suppl 5, Part 1): S1-S19.
West R, McNeill A, Raw M: Smoking cessation guidelines for health professionals: an update. Thorax 2000, 55: 987-999.
National Institute for Health and Clinical Excellence. Guidance on smoking cessation. 2008. Ref Type: Report
Vogt F, Hall S, Marteau TM: General practitioners' and family physicians' negative beliefs and attitudes towards discussing smoking cessation with patients: a systematic review. Addiction 2005, 100: 1423-1431.
Steinberg MB, Schmelzer AC, Richardson DL, Foulds J: The Case for Treating Tobacco Dependence as a Chronic Disease. Ann Intern Med 2008, 148: 554-556.
Stead LF, Perera R, Bullen C, Mant D, Lancaster T: Nicotine replacement therapy for smoking cessation. Cochrane Database of Systematic Reviews: Reviews 2008 Issue 1 John Wiley & Sons, Ltd Chichester, UK DOI: 10 1002/14651858 CD000146 pub3 2008.
Hajek P, Stead LF, West R, Jarvis M, Lancaster T: Relapse prevention interventions for smoking cessation. Cochrane Database of Systematic Reviews: Reviews 2005 Issue 1 John Wiley & Sons, Ltd Chichester, UK DOI: 10 1002/14651858 CD003999 pub2 2005.
Moore L, Campbell R, Whelan A, Mills N, Lupton P, Misselbrook E et al.: Self help smoking cessation in pregnancy: cluster randomised controlled trial. BMJ 2002, 325: 1383-1386.
Hughes JR, Keely JP, Niaura RS, Ossip-Klein DJ, Richmond RL, Swan GE: Measures of abstinence in clinical trials: issues and recommendations. Nicotine & Tobacco Research 2003, 5: 13-25.
Cahill K, Stead LF, Lancaster T: Nicotine receptor partial agonists for smoking cessation. Cochrane Database of Systematic Reviews: Reviews 2007 Issue 1 John Wiley & Sons, Ltd Chichester, UK DOI: 10 1002/14651858 CD006103 pub2 2007.
Shiffman S, Ferguson SG, Gwaltney CJ: Immediate hedonic response to smoking lapses: relationship to smoking relapse, and effects of nicotine replacement therapy. Psychopharmacology (Berl) 608, 2006.
Stapleton JA, Russell MA, Feyerabend C, Wiseman SM, Gustavsson G, Sawe U et al.: Dose effects and predictors of outcome in a randomized trial of transdermal nicotine patches in general practice. Addiction 1995, 90: 31-42.
Aveyard P, Brown K, Saunders C, Alexander A, Johnstone E, Munafo MR et al.: A randomised controlled trial of weekly versus basic smoking cessation support in primary care. Thorax 2007, In press.
Stapleton J, Stapleton J: Cigarette smoking prevalence, cessation and relapse. Stat Methods Med Res 1998, 7: 187-203.
Etter JF, Stapleton JA: Nicotine replacement therapy for long-term smoking cessation: a meta-analysis. Tob Control 2006, 15: 280-285.
Keely JP, Hughes JR, Carpenter MJ: A method to convert prolonged abstinence and point prevalence quit rates. Nicotine & Tobacco Research 2001, 3: 272-273.
Hughes JR, Keely JP, Niaura RS, Ossip-Klein DJ, Richmond RL, Swan GE: Measures of abstinence in clinical trials: issues and recommendations. Nicotine & Tobacco Research 2003, 5: 13-25.
West R, Hajek P, Stead L, Stapleton J: Outcome criteria in smoking cessation trials: proposal for a common standard. Addiction 2005, 100: 299-303.
West R, McNeill A, Raw M: Smoking cessation guidelines for health professionals: an update. Thorax 2000, 55: 987-999.
National Institute for Health and Clinical Excellence. Brief interventions and referral for smoking cessation in primary care and other settings. PH001, 1-36. 2006. London, National Institute for Health and Clinical Excellence. Ref Type: Report
Ministry of Health. Smoking cessation guidelines. 1-54. 2007. Wellington, New Zealand, Ministry of Health. Ref Type: Report
Song F, Altman DG, Glenny A, Deeks JJ: Validity of indirect comparison for estimating efficacy of competing interventions: empirical evidence from published meta-analyses. BMJ 2003, 326: 472.
Song F, Altman DG, Glenny A, Deeks JJ: Validity of indirect comparison for estimating efficacy of competing interventions: empirical evidence from published meta-analyses. BMJ 2003, 326: 472.
Lancaster T, Stead LF: Individual behavioural counselling for smoking cessation. The Cochrane Database of Systematic Reviews: Reviews 2005 Issue 2 John Wiley & Sons, Ltd Chichester, UK DOI: 10 1002/14651858 CD001292 pub2 2005.
Aveyard P, Johnson C, Fillingham S, Parsons A, Murphy M: A pragmatic randomised controlled trial of nortriptyline plus nicotine replacement versus placebo plus nicotine replacement for smoking cessation. BMJ 2008, 336: 1223-1227.
McEwen A, West R, McRobbie H: Effectiveness of specialist group treatment for smoking cessation vs. one-to-one treatment in primary care. Addict Behav 1650, 2006.
Hughes JR, Stead LF, Lancaster T: Antidepressants for smoking cessation. Cochrane Database of Systematic Reviews: Reviews 2007 Issue 1 John Wiley & Sons, Ltd Chichester, UK DOI: 10 1002/14651858 CD000031 pub3 2007.
Wang D, Connock M, Barton PM, Fry-Smith A, Aveyard P, Moore D: Cut down to quit with nicotine replacement therapies (NRT) in smoking cessation: Systematic review of effectiveness and economic analysis. Health Technol Assess 2008, 12.
Tverdal A, Bjartveit K: Health consequences of reduced daily cigarette consumption. Tob Control 2006, 15: 472-480.
Kiiskinen U, Vartiainen E, Puska P, Aromaa A: Long-term cost and life-expectancy consequences of hypertension. J Hypertens 1998, 16: 1103-1112.
Peto R, Doll R: The hazards of smoking and the benefits of stopping. In Understanding nicotine and tobacco addiction. Edited by Bock G, Goode J. Chichester: John Wiley & Sons; 2006:3-28.