Weaning From Tobacco with Nicotine or Varenicline in Severe and Mild Tobacco Dependence

Abstract

Background

Systematic reviews have documented the beneficial effects of bupropion, cytisine, nicotine, and varenicline as aids to permanent smoking cessation. We investigated the question whether the effect of treatment depends on the severity of tobacco dependence.

Methods

We systematically searched for relevant publications in bibliographic databases and trial registries, made inquiries to manufacturers, and consulted additional sources of information (last search on 1 September 2022). The smokers included in the present study were classified as mildly or severely tobacco-dependent on the basis of their scores on the Fagerström Test for Nicotine Dependence, with variable cut-off values (FTND score 4, 5, or 6). In this meta-analysis, we determined the results with respect to the endpoint of sustained smoking cessation at 6 and 12 months in each of the two subgroups and investigated the heterogeneity between them.

Results

No subgroup analyses that could enable us to answer the question posed in this meta-analysis were available for either bupropion or cytisine. Subgroup analyses were available for varenicline in 12 studies, involving a total of 9723 smokers, and for nicotine in 23, involving 15 003 smokers. No statistically significant heterogeneity (p > 0.05) between mildly and severely tobacco-dependent smokers was found for the effect of either drug on the endpoint sustained smoking cessation (at 6 and 12 months), and this was so independently of the FTND cut-off value that was used.

Conclusion

The benefit of varenicline and nicotine as aids to smoking cessation is independent of the severity of tobacco dependence.

Smoking tobacco is considered the most widespread preventable risk to health in Germany. Among the diseases promoted by tobacco consumption, cardiovascular disease, chronic obstructive pulmonary disease (COPD) and cancer, especially lung cancer, play a key role (1). In 2018, the death toll from smoking-related diseases was about 127 000 in Germany (1). According to March 2024 data from the German Study on Tobacco Use (DEBRA), 30.7% of the male and female German population aged 14 and over still smoked tobacco despite the associated health risks (2).

Since smoking tobacco is a combined physical and psychological dependence and nicotine in tobacco harbors a very high potential for dependency (3, 4), attempts to quit smoking frequently require some support. As aids to achieve sustained smoking cessation, various treatment options are available, including behavioral therapy interventions, such as brief advice by physicians and smoking cessation courses, and pharmacological treatments (5). The active substances approved in Germany for the pharmacotherapy of tobacco dependence are bupropion, cytisine, nicotine, and varenicline. Systematic reviews have repeatedly shown the advantageous effect of these substances compared to no pharmacological quit-smoking treatment with regard to the goal of achieving smoking cessation (6–11). In Germany, only insured persons diagnosed with severe tobacco dependence are eligible for reimbursement of these medications, pursuant to section 34 of Book V of the German Social Code (SGB V) (12). However, it remains unclear whether the beneficial effect of the medications used for smoking cessation is also present in the subgroup of persons with severe tobacco dependence. Thus, in 2022, the German Federal Joint Committee (G-BA, Gemeinsamer Bundesausschuss) commissioned the Institute for Quality and Efficiency in Health Care (IQWiG, Institut für Qualität und Wirtschaftlichkeit im Gesundheitswesen) to evaluate the benefits of the substances bupropion, cytisine, nicotine, and varenicline as aids to smoking cessation in smokers with severe tobacco dependence (13).

In order to answer this question, one of the aspects the IQWiG benefit assessment looked at was whether the treatment effect of the substances used as aids to smoking cessation is also observed in persons with severe tobacco dependence. The methodological approach used for the analysis was based on the following question:

Is there an effect modification (measured as heterogeneity between subgroups) associated with the characteristic “severity of tobacco dependence” when comparing the active substances to no pharmacological treatment for the endpoint “sustained smoking cessation”?

The Fagerström Test for Nicotine Dependence (FTND; a revision of the original Fagerström Tolerance Questionnaire [FTQ]) was used to assess the severity of tobacco dependence (3, 14). Various cut-off values (scores of 4, 5 and 6 points) were considered as a way of distinguishing between mildly and severely tobacco-dependent smokers.

Methods

The analyses performed are part of the IQWiG benefit assessment for the substances bupropion, cytisine, nicotine, and varenicline as aids to smoking cessation in severe tobacco dependence (13). The benefit assessment was performed in accordance with the General Methods of IQWiG (15). A detailed presentation of the pre-specified methodology (report plan) was published on the institute’s website prior to the preparation of the benefit assessment (www.iqwig.de/en/projects/a22–34.html).

Inclusion criteria

The IQWiG benefit assessment included randomized, controlled trials with a minimum study duration of 6 months, comparing bupropion, cytisine, nicotine, or varenicline based on their approvals as well as approved combinations of these substances to no pharmacological treatment for weaning from tobacco (including placebo). Since the target population of the benefit assessment comprised smokers with severe tobacco dependence, studies exclusively investigating smokers with mild tobacco dependence were excluded. Only those studies were included for which a study report or a full publication in German or English was available.

Information retrieval and data extraction

Detail on information retrieval are provided in eBox 1. Data extraction was performed by one person and reviewed by a second person. The bias potential of the study findings was rated by one person based on the criteria specified in the General Methods (15) and reviewed by a second person (eBox 2).

eBox 1. Information retrieval.

As part of the information retrieval process, we searched for pertinent studies in the databases MEDLINE, Embase and Cochrane Central Register of Controlled Trials, as well as ClinicalTrials.gov, International Clinical Trials Registry Platform (ICTRP) Search Portal and other trial registries. For a detailed description of the search strategies used refer to the eSupplement as well as the benefit assessment by the Institute for Quality and Efficiency in Health Care (IQWiG) (13). The last search was performed on 1 September 2022. Given the fact that the search in the trial registries did not identify any relevant ongoing or planned trials, it is reasonably safe to assume that no new findings have been published since the last search, even though this was conducted a longer time ago. In order to identify unpublished trials, we requested summaries of all studies carried out as well as clinical study reports of potentially relevant trials from the manufacturers of medicinal products approved as aids to smoking cessation in Germany. In addition, we screened the websites of the Federal Joint Committee (G-BA) and IQWiG as well as the reference lists of systematic reviews. Two persons carried out the selection of relevant studies independently of each other. Any discrepancies were resolved by means of discussion between the two.

eBox 2. Assessment of the bias potential of the results.

The bias potential of the results was assess for each included trial in an endpoint-specific manner. In particular, the following endpoint-spanning (A) and endpoint-specific (B) criteria were systematically extracted and assess:

A: Criteria for the endpoint-spanning assessment of the bias potential of the results

• Creation of the randomization sequence

• Masking of group assignment

• Blinding of participants and treating healthcare professionals

• Reporting independent of results

B: Criteria for the endpoint-specific assessment of the bias potential of the results

• Blinding of the persons collecting endpoint data

• Implementation of the intention-to-treat (ITT) principle

• Reporting independent of results

Summarizing, the bias potential of the results of the included trials was rated as low or high. If an endpoint-spanning high degree of bias potential was already identified with regard to the criteria listed under (A), the bias potential was considered to be high for all results of all endpoints, regardless of the assessment of endpoint-specific aspects. Otherwise, the criteria listed under (B) were subsequently taken into account for each endpoint.

Overall, the results of 2 of the 12 studies on varenicline were assessed across all endpoints as potentially highly biased due to unclear information about randomization and/or allocation concealment. The bias potential of the results for the endpoints “sustained smoking cessation at month 6“ and “sustained smoking cessation at month 12” was assessed as high for another five and six varenicline trials, respectively. The high level of bias potential was due to the fact that a high or unclear number of dropouts were noted at month 6 and/or month 12.

Of the 23 trials on nicotine, the results of six trials were assessed as having a high potential for bias across all endpoints, because the information about randomization and/or allocation concealment was unclear. The bias potential of the results for the endpoints “sustained smoking cessation at month 6“ and “sustained smoking cessation at month 12”was assessed as high for another eleven and ten nicotine trials, respectively. The reason for this was that the number of dropouts was very high or not reported at the respective points in time.

Based on the bias potential assessment, the qualitative certainty of results at endpoint level was rated as moderate (high bias potential) or high (low bias potential).

In our meta-analysis, we only considered trials in which the primary endpoint of sustained smoking cessation was reported. This endpoint was defined as follows: Biochemically validated, continuous smoking cessation without exceptions from day 1 of smoking cessation to the time of data colection at month 6/month 12. Studies using a different operationalization of the endpoint were not taken into account (for example, studies allowing occasional smoking according to the Russell Standard criteria [16] or studies only reporting 7-day/30-day point prevalence rates).

In order to answer the question of whether there is an effect modification for the endpoint of sustained smoking cessation related to the severity of tobacco dependence, the trial results were requested separately for the subgroups which had been created using the cut-off score values of 4, 5 and 6 in the FTND or FTQ; the information was requested from the manufacturers of the respective medicinal products or, in the case of cytisine, from the study authors.

The test comprises 8 (FTQ) or 6 (FTND) questions about smoking behavior. The test includes in particular responses with regard to the mean number of cigarettes smoked per day and the time until smoking the first cigarette after waking. While the test allows statements to be made about the severity of the dependence, it does not take any potential clinical consequences of tobacco consumption into account. Points are assigned to the possible answers. In this way, a total of 0–11 (FTQ) or 0–10 (FTND) points can be achieved, with a higher score denoting a more severe dependence. For the purpose of this analysis, we made no distinction between FTND and FTQ and used identical cut-off values. Given that the test does not specify the score at which a severe dependence is diagnosed and the absence of a uniform scientific consensus on a suitable FTND/FTQ cut-off value, three cut-off values were evaluated.

Statistical analysis

The study results were calculated for each of the subgroups using the cut-off values 4, 5 and 6 points in the FTND/FTQ; furthermore, the estimated effects (relative risk) and 95% confidence intervals from the studies were summarized, using forest plots. Heterogeneity between subgroups was used to identify an effect modification related to the “severity of tobacco dependence” characteristic. This was determined by heterogeneity testing (at a significance level of α = 0.05) (17). In order to detect signs of heterogeneity between the subgroups with the greatest possible sensitivity, no correction for multiple testing was performed. The results for each subgroup were summarized in a meta-analysis by means of a random effects model based on the Knapp-Hartung method, using the Paule-Mandel heterogeneity estimator (18). The statistical software SAS, version 9. 4 was used for all calculations.

Results

Results of information gathering

A total of 136 studies met the inclusion criteria (eFigure 1). These included three studies on cytisine, 38 studies on varenicline and 100 studies on nicotine. Five studies evaluated both varenicline and nicotine. In addition, 51 studies on bupropion as well as six studies on combinations of active ingredients were identified, with bupropion as a combination partners in each of these studies. The questions with regard to bupropion or drug combinations with bupropion could not be answered, because neither a full publication nor the study reports requested from the manufacturer were available for a relevant portion of the studies.

eFigure 1.

Flowchart for information retrieval

For cytisine, an effect modification by the characteristic “severity of tobacco dependence“ could not be assessed, because data on subgroup analysis by FTND/FTQ cut-off values for the endpoint “sustained smoking cessation” were available from only one of the three studies identified.

Thus, the findings detailed below only refer to varenicline and nicotine.

Characteristics of the studies on varenicline and nicotine

The endpoint “sustained smoking cessation“ was reported in 20 of the 38 varenicline studies identified. Data on subgroup analyses for the cut-off values 4, 5 and 6 points in the FTND/ FTQ in respect to the endpoint “sustained smoking cessation“ were provide for all twelve manufacturer-sponsored studies on varenicline. A total of 9723 smokers were included in the twelve studies for which subgroup analysis data were available.

The endpoint “sustained smoking cessation“ was reported in 43 of the 100 identified studies on nicotine. The requested manufacturer data for all three FTNZ/FTQ cut-off values were provide for 22 of the 24 manufacturer-sponsored trials. For one further study on nicotine, only data for the 6-point cut-off value in respect to the endpoint “sustained smoking cessation” at month 6 were available. A total of 15 003 smokers were included in the 23 studies for which subgroup analysis data were available. The studies evaluated various dosage forms of nicotine or combinations of a transdermal patch and another dosage form.

In most trials, the majority of smokers had made at least one attempt to quit smoking prior to inclusion in the study. All of the studies on varenicline evaluated a course of treatment over twelve weeks. In the studies on nicotine, the study duration varied between six weeks and 18 months. Additional characteristics of the study populations included in our meta-analysis are provided in Table 1 (refer to [13] for a more comprehensive overview of the study and population characteristics)

Table 1. Population characteristics of the trials on varenicline and nicotine included in the subgroup analyses.

	Varenicline trials N = 12* included smokers (n = 9723)	Nicotine trials N = 23* included smokers (n = 15 003)
Age (years)	39–57	39–51
Sex Women (%) Men (%)	3.5–62.5 37.5–96.5	3.5–100 0–96.5
FTND/FTQ score ≥ 6 points (%) < 6 points (%)	5.1–6.0 44–63 37–56	3.7–7.4 41–99 1–59
Mean number of cigarettes per day	16.7–24.3	17.9–29.8
Years of smoking	20.0–40.5	20.9–29.4

* Range of the means/percentages of the studies

FTQ, Fagerström Tolerance Questionnaire;

FTND, Fagerström Test for Nicotine Dependence;

n, number of smokers;

N, number of studies analyzed

Subgroup analyses for varenicline and nicotine

For varenicline, the comparison of varenicline and no pharmacological treatment found advantages for varenicline in both the subgroup of severely tobacco-dependent smokers and the subgroup of mildly tobacco-dependent smokers, and that was so independently of the cut-off value used (Table 2). For example, for the endpoint “sustained smoking cessation at month 6”, the pooled effect estimate of the relative risk in the subgroup FTND score ≥ 6 points was 2.72 (95% confidence interval: [2.14; 3.44]) and in the subgroup FTND score < 6 points 2.36 [1.96; 2.83) (eFigure 2). Heterogeneity of statistical significance (p>0.05) between the subgroups was not found for any of the cut-off values. The results were consistent for the endpoints “sustained smoking cessation at month 6” and “sustained smoking cessation at month 12”.

Table 2. Effects of the substances varenicline and nicotine on achieving sustained smoking cessation at month 6 and month 12 by severity of tobacco dependence^*1.

	Effect size in severe tobacco dependence*2	Effect size in mild tobacco dependence*2	Heterogeneity*4
	RR [95% CI]*3	RR [95% CI]*3
Varenicline vs. placebo/no pharmacological treatment
Sustained smoking cessation at month 6
FTND ≥ 6 vs. < 6	2.72 [2.14; 3.44]	2.36 [1.96; 2.83]	p = 0.293; I2 = 9.4%
FTND ≥ 5 vs. < 5	2.57 [2.02; 3.27]	2.50 [1.96; 3.18]	p = 0.845; I2 = 0%
FTND ≥ 4 vs. < 4	2.66 [2.15; 3.29]	2.18 [1.74; 2.73]	p = 0.158; I2 = 49.7%
Sustained smoking cessation at month 12
FTND ≥ 6 vs. < 6	2.81 [2.22; 3.55]	2.39 [1.80; 3.17]	p = 0.299; I2 = 7.2%
FTND ≥ 5 vs. < 5	2.54 [2.00; 3.24]	2.69 [1.83; 3.95]	p = 0.768; I2 = 0%
FTND ≥ 4 vs. < 4	2.68 [2.00; 3.59]	2.37 [1.68; 3.34]	p = 0.512; I2 = 0%
Nicotine vs. placebo/no pharmacological treatment
Sustained smoking cessation at month 6
FTND/FTQ ≥ 6 vs. < 6	1.81 [1.53; 2.14]	1.71 [1.41; 2.07]	p = 0.634; I2 = 0%
FTND/FTQ ≥ 5 vs. < 5	1.84 [1.57; 2.14]	1.65 [1.29; 2.12]	p = 0.454; I2 = 0%
FTND/FTQ ≥ 4 vs. < 4	1.90 [1.56; 2.32]	1.71 [1.24; 2.35]	p = 0.539; I2 = 0%
Sustained smoking cessation at month 12
FTND/FTQ ≥ 6 vs. < 6	1.60 [1.36; 1.87]	1.58 [1.22; 2.06]	p = 0.954; I2 = 0%
FTND/FTQ ≥ 5 vs. < 5	1.63 [1.40; 1.89]	1.49 [1.06; 2.10]	p = 0.624; I2 = 0%
FTND/FTQ ≥ 4 vs. < 4	1.62 [1.36; 1.92]	1.59 [1.02; 2.49]	p = 0.941; I2 = 0%

^*1 In each case, the meta-analytical effect estimate (relative risk) for each subgroup and the result of the test for heterogeneity between the subgroups are shown.

^*2 Smokers included in the trials were assigned to subgroups on the basis of the FTND/FTQ scores (severe tobacco dependence vs. mild tobacco dependence). Three different cut-off values were used to define the subgroups in the FTND/FTQ (6, 5 and 4 points).

^*3 Effect estimate and 95% CI from the meta-analysis with random effects

^*4 Heterogeneity between subgroups: p-value from Q test. The I 2 value is a measure for quantifying heterogeneity between subgroups; it indicates what proportion of the variability is explained by heterogeneity.

FTQ, Fagerström Tolerance Questionnaire;

FTND, Fagerström Test for Nicotine Dependence; CI, confidence interval;

RR, relative risk, vs., versus

eFigure 2.

Forest Plot of subgroup analyses for the comparison of varenicline vs. no pharmacological treatment. The results for the endpoint “sustained smoking cessation at month 6” are shown. A cut-off value of 6 points in the Fagerström test for Nicotine Dependence (FTND) was used for the subgroup analyses. The effect estimate (relative risk [RR]) with 95% confidence interval (CI) is shown for each individual study. Effects >1 indicate an advantage for varenicline. The pooled effect estimates in the respective bottom rows were calculated using a random effects model. The information on heterogeneity provided directly below the trials refers to the respective subgroups. The information on heterogeneity between the subgroups can be found at the very bottom and is additionally provided in Table 2.

n, number of smokers with event; N, number of analyzed smokers

Similarly, nicotine showed advantages over no pharmacological treatment for the endpoint “sustained smoking cessation” in the subgroups of severely and less severely tobacco-dependent smokers (Table 2). For example, for the endpoint “sustained smoking cessation at month 6”, the pooled effect estimate of the relative risk in the subgroup FTND score ≥ 6 points was 1.81 [1.53; 2.14] and in the subgroup FTND score < 6 points 1.71 [1.41; 2.07) (eFigure 3). There was no heterogeneity of statistical significance (p>0.05) between the subgroups. This applied to both the end point “sustained smoking cessation at month 6” and the end point “sustained smoking cessation at month 12”.

eFigure 3.

Forest Plot of subgroup analyses for the comparison of nicotine vs. no pharmacological treatment. The results for the endpoint “sustained smoking cessation at month 6” are shown. A cut-off value of 6 points in the Fagerström Test for Nicotine Dependence (FTND)/ Fagerström Tolerance Questionnaire (FTQ) was used for the subgroup analyses. The effect estimate (relative risk [RR]) with 95% confidence interval (CI) is shown for each individual study. Effects >1 indicate an advantage for nicotine. The pooled effect estimates in the respective bottom rows were calculated using a random effects model. The information on heterogeneity provided directly below the trials refers to the respective subgroups. The information on heterogeneity between the subgroups can be found at the very bottom and is also provided in Table 2.

*¹ Only an unpublished study report is available for this study; there is no published information available on whether the FTND or FTQ was used in this study.

*² In this study, the FTND was used.

*³ In this study, the FTQ was used.

n, number of smokers with event; N, number of analyzed smokers

Discussion

With the amendment of section 34 SGB V as part of the German Healthcare Advancement Act (GVWG, Gesundheitsversorgungsweiterentwicklungsgesetz), the entitlement to a one-off provision of medication for smoking cessation as part of evidence-based smoking cessation programs was established. While the S3 clinical practice guideline “Smoking and tobacco dependence: screening, diagnosis and treatment” recommends the use of medication regardless of the severity of tobacco dependence (5), the legislator restricted the eligibility for reimbursement to insured persons with existing severe tobacco dependence (12). Yet, it remains undetermined how the severity of tobacco dependence is to be defined. In clinical research, the FTND/FTQ scores are commonly used to grade the severity of tobacco dependence (3, 14).

The benefit of varenicline and nicotine as aids to smoking cessation independent of FTND/FTQ scores

The available subgroup analyses for various FTND/FTQ cut-off values show no influence of the severity of tobacco dependence on the effect of treatment with varenicline or nicotine (Box). This applies to the achievement of both sustained smoking cessation at month 6 and at month 12.

Box. Important for clinical practice.

• Smoking is a mayor risk factor for numerous diseases, in particular cardiovascular disease, chronic obstructive pulmonary disease and cancer (1).

• The severity of tobacco dependence can be measured using the Fagerström test (FTND) (3). The six questions include responses with regard to the mean number of cigarettes smoked per day and the time until smoking the first cigarette after waking. A higher score (scale 0–10) denotes a more severe dependence.

• The effect of nicotine replacement therapy (NRT) is independent of the severity of tobacco dependence. Among severely tobacco-dependent smokers (FTND score ≥ 6), the number needed to treat (NNT) to achieve sustained smoking cessation at month 6 was 22. Among mildly tobacco-dependent smokers (FTND score < 6), the NNT was 25. Common side effects include oral and throat irritation, stomach upset, hiccups, and skin irritation. At times, no distinction can be made between withdrawal symptoms and side effects. The duration of treatment varies between NRT products.

• The effect of varenicline treatment is independent of the severity of tobacco dependence. Among severely tobacco-dependent smokers (FTND score ≥ 6), the NNT to achieve sustained smoking cessation at month 6 is 17, among mildly tobacco-dependent smokers (FTND score < 6) 12. The most common side effects (not always distinguishable from withdrawal symptoms) include nausea, headache, sleep disorders, and abnormal dreams. Varenicline treatment is initiated 1 to 2 weeks prior to the smoking cessation and lasts 12 weeks. Prolonged use up to 24 weeks is possible.

• At present, no statement can be made as to whether bupropion, cytisine, and combinations of various substances are also effective in severe tobacco dependence (13).

Thus, the effects calculated in the benefit assessment by IQWiG and in systematic reviews for the total population of smokers can be applied to smokers with severe tobacco dependence (6–11, 13).

In light of the available findings, it should be discussed whether FTND/FTQ scores should be used to define the target population of severely tobacco-dependent smokers pursuant to section 34 SGB V.

In their comments on the IQWiG benefit assessment, the experts noted that further aspects should be taken into account in order to adequately define the target population intended by the legislator. The purpose of these aspects was to ensure that, on the one hand, smokers with existing tobacco-associated disease (e.g., cardiovascular disease, COPD) are included and, on the other hand, that it is also possible to offer primary prevention. One possible criterion being discussed for defining the target population of severely tobacco-dependent smokers is the inability to abstain from smoking, which can, for instance, show as repeated failed attempts at abstinence or as continued tobacco consumption despite an existing tobacco-associated disease or being pregnant (19).

The question of whether the treatment effect depends on the severity of tobacco dependence was already addressed in a systematic review by the National Institute for Health Research (NIHR) (9). The analysis used a meta-regression approach to assess the effect of the severity of tobacco dependence on achieving sustained smoking cessation. It found that the treatment effect depended on the severity of tobacco dependence. In addition to nicotine and varenicline, the review looked at bupropion and combinations of these substances as well as e-cigarettes.

However, it remains open whether all treatments studied were associated with effect modification and whether the extent of effect modification differs between the various treatments. No conclusive evaluation can therefore be made as to how far these findings conflict with the results of our meta-analysis

Another aspect addressed in the NIHR analysis was the influence of the mean number of cigarettes smoked on achieving sustained smoking cessation. No effect modification was found between smokers who consumed > 20 cigarettes per day on average and smokers who consumed ≤ 20 cigarettes per day. A Cochrane review on varenicline contains analyses that looked at studies including smokers consuming ≤ 10 cigarettes per day separately (11). In this group too, a treatment effect on smoking cessation in favor of varenicline was found.

Limitations

There are some limitations to our study. We investigated, for example, effect modification by tobacco dependence severity for the endpoint “sustained smoking cessation“, but not for other endpoints. In addition, our subgroup analyses included only studies in which the reported endpoint “sustained smoking cessation“ was defined as continuous smoking cessation without exemptions at month 6 and/or month 12. Studies using different definitions of the endpoint were not included.

The reason behind this approach is that the primary goal of treating tobacco dependence is to achieve sustained smoking cessation. Furthermore, we requested the data on varenicline and nicotine we needed for our subgroup analyses only for studies sponsored by pharmaceutical companies, as we considered the chances of success of author requests to be low for the majority of studies, about half of which were conducted more than 20 years ago. Nevertheless, in light of twelve studies on varenicline (including 9723 smokers) and 23 studies on nicotine (including 15 003 smokers), the study pool of our meta-analysis appears to be large enough to exclude, with reasonable certainty, an effect modification attributable to the severity of tobacco dependence.

Conclusion

Based on the findings of the available studies, it can be assumed that the beneficial effect of the active ingredients varenicline and nicotine with regard to achieving sustained smoking cessation is independent of the severity of tobacco dependence.

For bupropion and cytisine, by contrast, the influence of the severity of tobacco dependence on the effect of treatment could not be analyzed due to a lack of data.