NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Cover of Smoking Cessation Interventions for Patients with Severe Mental Illnesses: A Review of Clinical Effectiveness and Guidelines

Smoking Cessation Interventions for Patients with Severe Mental Illnesses: A Review of Clinical Effectiveness and Guidelines

CADTH Rapid Response Report: Summary with Critical Appraisal

and .

Context and Policy Issues

While society as a whole has seen a drastic decline in smoking over the past few decades, people with a mental disorder have not had the same gains. Smoking among people with posttraumatic stress disorder (PTSD), for example, has been estimated at around 45%, with greater nicotine dependence and heavier smoking than the general population.1 Among people with schizophrenia, the likelihood of smoking is more than five times greater than that of the general population.2 In addition to the higher smoking prevalence, people with severe mental illness have premature mortality compared to the general population, a gap primarily driven by cardiovascular and respiratory diseases. These driving diseases have prompted calls for treating associated lifestyle factors including high smoking rates.3

Pharmacological interventions are available to assist in smoking cessation by reducing nicotine withdrawal. The main medications found to be effective are nicotine replacement therapy (NRT), varenicline and bupropion.4 The first comes in patch, lozenge, gum, nasal spray and oral inhaler form, and provides nicotine without tobacco while the user breaks smoking behavior. Varenicline by contrast is an orally ingested partial agonist that binds to the nicotinic receptor to reduce cravings. Finally, bupropion was an anti-depressant that was subsequently licensed for smoking cessation. It is believed to act by enhancing noradrenergic and dopaminergic release.4

While there is some evidence of predisposition and more severe nicotine dependence among those with schizophrenia,5,6 the observed low cessation rates among those with mental illness7 are thought to be partly due to lower availability and/or utilization of pharmacotherapy.8,9 People with mental illness are often excluded from large clinical trials,10 so effectiveness in this group is not evident, especially in the context of particularly high smoking rates, relapse rates and nicotine dependence.10 In addition, there is concern among smokers and clinicians that the medications are unsafe due to the potential for severe psychiatric reactions.11 For example, post-marketing reports have suggested varenicline and bupropion may increase the risk of suicidal thoughts.4

Treatment may be able to lessen the significant burden and associated risks of smoking among people with mental illness, but concerns about effectiveness and safety may limit its potential. To inform decisions about using pharmacotherapy among those with mental illness, specific evidence is thus required. As such, this rapid review summarizes the evidence on the effectiveness and evidence-based guidelines on smoking cessation intervention among those with severe mental illness.

Research Questions

  1. What is the clinical evidence regarding the effectiveness of smoking cessation interventions for patients with severe mental illnesses?
  2. What are the evidence-based guidelines regarding smoking cessation interventions for patients with severe mental illnesses?

Key Findings

Nicotine replacement therapy (NRT), varenicline and bupropion were all generally found to improve smoking cessation rates, most commonly measured as 7-day point prevalence abstinence against placebo, among people with severe mental illness. Varenicline and bupropion may be more effective than NRT. The longest follow up duration on treatment was six months, and sustained effects beyond this were unclear, except in one study that did not find an increase in suicide attempts/behaviours after one year.

There was no evidence that adverse events occurred more often than in those without mental illness, or that psychiatric symptoms worsened. Nausea and sleep disturbances were common with varenicline. All studies included some form of psychosocial support in addition to the pharmacological intervention, and only one study considered randomizing different dosages for NRT, and did not find a difference. In general, the quality of studies was low and sample sizes small.

One guideline was found, which did not favour any treatment, but suggested considering NRT, varenicline or bupropion for patients with severe mental illness, with safety caveats for the latter two that may require closer patient monitoring.

Methods

Literature Search Methods

A limited literature search was conducted on key resources including PubMed, The Cochrane Library, University of York Centre for Reviews and Dissemination (CRD) databases, Canadian and major international health technology agencies, as well as a focused Internet search. No filters were applied to limit the retrieval by study type. Where possible, retrieval was limited to the human population. The search was also limited to English language documents published between January 1,2006 and July 21, 2017.

Rapid Response reports are organized so that the evidence for each research question is presented separately.

Selection Criteria and Methods

One reviewer screened citations and selected studies. In the first level of screening, titles and abstracts were reviewed and potentially relevant articles were retrieved and assessed for inclusion. The final selection of full-text articles was based on the inclusion criteria presented in Table 1.

Table 1. Selection Criteria.

Table 1

Selection Criteria.

Exclusion Criteria

Articles were excluded if they did not meet the selection criteria outlined in Table 1, they were duplicate publications, were published prior to 2006 or already appeared in a systematic review identified through the search. Articles where the primary focus was on patients with mood disorders such as depression and anxiety were also excluded as this was not the target population of interest.

Critical Appraisal of Individual Studies

The included systematic reviews were critically appraised using the AMSTAR checklist,12 clinical studies were critically appraised using Downs and Black,13 and guidelines were assessed with the AGREE II instrument.14 Summary scores were not calculated for the included studies; rather, a review of the strengths and limitations of each included study were described.

Summary of Evidence

Quantity of Research Available

This report made use of a previous Rapid Response Reference List15 on the same subject to identify studies, which was supplemented by an updated search. The original search, which was completed in July 2016, identified 325 citations, of which 25 were ordered for full-text screening. A total of 86 additional citations were identified in the literature search update. Following screening of titles and abstracts of the 86 new citations, 65 citations were excluded and 21 potentially relevant reports from the electronic search were retrieved for full-text review. Forty-five potentially relevant publications were retrieved from the grey literature search and, in addition to the 25 from the previous reference list on the topic and the 21 added in the update, 91 full texts were reviewed. Of these potentially relevant articles, 68 publications were excluded for various reasons, while 23 publications met the inclusion criteria and were included in this report. Appendix 1 describes the PRISMA flowchart of the study selection.

Additional references of potential interest that didn’t meet the selection criteria are provided in Appendix 5.

Summary of Study Characteristics

Details of the characteristics of individual studies are provided in Appendix 2.

Study Design

Eight systematic reviews were found. Two reviews retrieved 22 and 13 studies respectively, including pharmacological interventions, though did not specify study types or specific interventions.16,17 The others included 17 studies representing 14 randomized controlled trials (RCTs) identified until December 2014,8 21 studies of seven RCTs until March 2009,6 eight RCT studies identified until September 2015,18 seven double-blinded, placebo controlled trials from a search up to August 2014,19 and 34 randomized trials with 10 focused on pharmacological interventions identified up to October 2012.20 Finally the most recent review with a search up to September 2016 included 28 RCTs of which 16 focused on pharmacological interventions.21

In addition to the systematic reviews, four double-blinded RCTs were found.10,2224 One of these was a pooled analysis of two RCTs with an open-label cessation phase, followed by a randomized, blinded maintenance phase.10 There was also one open label RCT25 and one secondary analysis of a blinded RCT.26

The remaining eight studies were non-randomized. In two cohort studies,27,28 patients self-selected into the treatment and comparator groups, and two retrospective cohort studies were based on chart reviews.29,30 Four pre/post studies where outcomes were compared over time in the same subjects before and after treatment were also included.3134

One guideline published by the National Institute for Health and Care Excellence (NICE)35 was found.

Country of Origin

The systematic reviews did not have country restrictions, though three were restricted to English language studies only.8,16,21 Of the randomized studies, one was situated in Australia,22 one in Taiwan,23 two in the U.S.,24,25 one was in Korea26 and one was a multinational trial.10

Among the non-randomized studies, two were from Spain,28,31 one took place in the UK,27 three from the U.S.,29,32,34 one from Brazil,30 and the remaining one was situated in Australia.33

The guideline was published by the United Kingdom’s NICE.35

Patient Population

Systematic reviews

Four of the systematic reviews specified a restriction to adults, usually defined as 18 or older,8,18,20,21 though in practice the studies included in the other reviews were also limited to adults. Three included only schizophrenia or schizoaffective disorder6,20 or schizophrenia spectrum disorders,17 while the other studies additionally included other disorders. One included people with bipolar disorder, delusional disorder and depressive psychoses, but in practice all but 3 of 17 studies were among people with schizophrenia or schizoaffective disorder.8 Another review also included schizophreniform disorder and delusional disorder, though similarly, 6 of 7 studies ended up being with schizophrenia patients, and one included both schizophrenia and bipolar disorder patients.19 The remaining two systematic reviews were more broad, with one specifying any severe mental illness18 and the other including schizophrenia or other psychotic disorders, bipolar disorder and depression with psychotic features.21 This last study specifically excluded those with PTSD, personality, anxiety disorders, major depression and autism. One study only required 50% or more of a study sample to have schizophrenia or schizoaffective disorder.16 Four systematic reviews specified the participants had to be smokers6,8,17,20 and one further specified the patients had to have motivation to quit or reduce smoking.8 While the others did not explicitly specify participants had to be smokers, in practice they were.

Clinical Studies

Five studies included only people with a diagnosis of schizophrenia or schizoaffective disorder.23,25,26,29,32 One of these included males only.25 People with bipolar disorder28 and chronic delusion31 were additionally included in two studies, and one study focused on schizophrenia or schizoaffective disorder depressed type.34 Two studies included people with and without mental illness, with one specifying the illness as schizophrenia or bipolar disorder,10 and the other basing it on a self-report of having any mental disorder.30 Two studies included inpatients of psychiatric facilities. One included patients meeting Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV for schizophrenia or schizoaffective disorders only,23 while the other included those with diagnoses of mood disorders, schizophrenia and related psychosis, or other.22

People with post-traumatic stress disorder were the focus of one study,24 while the remaining three studies were less specific. One included people with any psychotic disorder based on the Mini International Neuropsychiatric Interview,33 and the other included anyone with a primary mental illness diagnosis.27 This last study ultimately ended up including 64/111 (58%) with depression, 14/111(13%) with bipolar disorder, 7/111(6%) with psychosis, 24/111(22%) with psychosis and depression and 2/111(2%) with eating disorders.

All studies required participants to be smokers, but the definition varied. Most commonly, studies required smoking at least 10 cigarettes per day,24,25,31 with three additionally requiring an expired carbon monoxide (CO) concentration of at least 932,34 or 1026 parts per million. Two required smoking at least 15 cigarettes per day,28,33 and another two required a self-report of being a current or occasional smoker.22,23 Other studies implicitly restricted to smokers by recruiting from a tobacco specialist clinic,27 including those prescribed cessation medication29 or did not specify.10,30 In addition to being smokers, almost half of the clinical studies (n=6) only included patients with an expressed desire to quit or reduce their smoking.24,25,3032,34

There were several common exclusion criteria. Eight studies required having a clinically stable disorder,2326,3134 for example, no changes to psychopharmacological maintenance treatment in the six months before the study.31 People with other substance dependencies were also typically excluded,2426,3032,34 as were those with suicidal ideations or high suicide risk,25,26,28,3134 for example, one study stated that participants could not have had any hospitalizations for suicidal ideation in the preceding year.32

Guidelines

The guideline’s target population was adults aged 18 or older with schizophrenia or psychosis, as well as those involved in decisions concerning their care.35

Interventions and Comparators

Systematic reviews

Two systematic reviews included any smoking cessation intervention and comparison.16,17 The others were more specific. Two reviews included studies where the intervention was varenicline only,18,19 of which one included placebo comparisons19 and the other included placebo or other (unspecified) pharmacotherapy interventions.18 The third review included nicotine replacement therapy, bupropion and varenicline alone, or in combination, with or without psychological support, compared to placebo or each other.8 The fourth review included bupropion, alone or in combination with other interventions, with placebo comparators.6 The remaining two systematic reviews included any pharmacological or nonpharmacological intervention alone or in combination, compared to each other, placebo, or usual care (which was not defined),20,21 with one also adding no intervention as a comparator.21

Clinical studies

Three studies, all randomized, evaluated nicotine replacement therapy (NRT).2224 Of these, one compared a two-week supply of an unspecified form of NRT combined with motivational interview, nightly phone calls, an offer for a 12 week additional NRT supply, referral to quit service and smoking cessation groups, to treatment as usual which could consist of advice, NRT for three days and/or a smoking care plan.22 The second study compared high dose transdermal nicotine patch (31.2mg) to a low-dose patch (20.8mg). Both groups received group psycho-education sessions.23 The last study using an NRT intervention gave the intervention group two weeks of active nicotine patch loading (21mg) and the comparator was a placebo patch. Both groups also received individual cognitive behavioral therapy sessions.24

The standard dose for varenicline across six of seven studies with this intervention was 0.5 mg/day for 3 days, 0.5 mg/day for 4 days, then 1 mg twice daily until end of follow up at 11 weeks28,3134 or eight weeks.26 The remaining study randomized to maintenance treatment among abstinent smokers, providing only 1.0mg twice daily up to 24 weeks.10 In the randomized studies, the comparator was identical placebo.10,26 The comparator groups among the non-randomized studies included NRT patch (14, 21, 28 or 35 mg),28 NRT of variable dose/preparation,27 or consisted of the same people observed before and after the intervention.3134 All studies included some form of behavioral support sessions, usually in group format, that was administered to both treatment and comparator groups.

Two studies’ intervention was bupropion combined with forms of NRT.25,30 The first was an open-label, randomized study where buproprion was administered in combination with nicotine patch and lozenge with some receiving home visits from investigators to provide support, while the comparator group was treatment as usual consisting of a single first-line medication of either nicotine patch, bupropion or varenicline with group cognitive behavioural therapy sessions.25 The other bupropion study was non-randomized with three treatment arms consisting of nicotine patch plus either bupropion, gum or nortriptyline, while the comparator group received only the nicotine patch. All groups also received group cognitive behavioural therapy sessions.30

The final (non-randomized) study had varenicline, NRT (unspecified form) and bupropion treatment groups, and compared the latter two to the varenicline group.29

Guidelines

The guideline considered bupropion, varenicline and transdermal nicotine patch.35

Outcomes

Systematic reviews

One systematic review did not specify outcomes,17 and one specified, 'some smoking-related outcome variable was measured [self-reported smoking, breath carbon monoxide (CO), etc]." (pg 181)16 The rest of the reviews included smoking abstinence or cessation as the primary outcome, with one specifically requiring biochemically verified self-report of cessation.21 Four systematic reviews considered reduction in smoking6,18,20,21 and five stated psychiatric outcomes including mental state,20,21 positive, negative and depressive symptoms,6,19 psychiatric adverse events,18 or discontinuation due to psychiatric events.8 All studies but two16,17 collected information on adverse events in general.

Clinical studies

The clinical studies generally measured similar smoking cessation and reduction outcomes. The most common criteria for smoking cessation was 7-day point prevalence abstinence, measured in seven studies2225,28,30,32 while end of study abstinence was measured in three studies27,31,34 and weeks of continuous abstinence was measured in three studies.10,22,32 Thirty-day abstinence was measured in one study,24 and reduction in smoking was measured in eight studies.22,23,23,25,26,26,28,33

Nicotine dependence and withdrawal symptoms were also commonly measured. The Fagerstrom Test for Nicotine Dependence (FTND) was most common for this outcome.22,23,25,28,33 One study used a self-complete diary to assess smoking craving,24 while another used a tobacco withdrawal symptoms scale.27 Other scales used to assess dependence and/or withdrawal included the Glover-Nilsson Smoking Behavioural Questionnare,28 the Minnesota Nicotine Withdrawal Scale (MNWS),26,33 the Wisconsin Smoking Withdrawal Scale (WSWS),25,32,34 the Brief Questionnaire of Smoking Urge26 and the Modified Cigarette Evaluation Questionnaire (mCEQ).26

Fifteen different scales were used to measure psychiatric symptoms and/or safety across seven studies assessing these.23,25,26,29,3234 Most commonly, the Scale for Assessment of Negative Symptoms (SANS),25,26,32,34 the Brief Psychiatric Rating Scale (BPRS),25,3234 the Simpson-Angus Rating Scale (SARS),23,26 and the Positive and Negative Syndrome Scale (PANSS) were used.23,26 Two studies assessed depressive symptoms using the Hamilton Rating Scale for Depression (HADS),26 two used the Calgary Depression Scale for Schizophrenia (CDSS).32,34 and one used the Beck Depression Inventory (BDI).25,33 One study measured suicidal attempts/behaviours defined through International Classification of Disease (ICD) and E-codes.29 Other scales used included the Columbia- Suicide Severity Rating Scale (C-SSRS),25,33 the Abnormal Involuntary Movement Scale (AIMS),25 the Young Mania rating scale,33 the Kessler Psychological Distress Scale,22 the Barnes Akathisia Scale26 and the Clinical Global Impression (CGI) scale.25,26

Guidelines

The guideline did not state specific outcomes of interest.35

Summary of Critical Appraisal

Details of the appraisal of individual studies are provided in Appendix 3.

Systematic reviews

All the systematic reviews included at least two databases in their search, though one study stated Medline and PubMed as separate databases which was unclear as Pubmed is an interface for Medline.16 Two reviews used a limited number of key words for the search, and did not provide details on the included studies or on methods such as screening procedures, or consider study quality or publication bias.16,17 One of these also conducted some form of meta-analysis that was not described or interpretable.17

The remaining six systematic reviews were considered higher quality because they used comprehensive search terms, combined estimates using appropriate statistical methods such as random effects models, and assessed study quality using a study-design checklist,6 Grading of Recommendations Assessment, Development and Evaluation (GRADE)8 or the Cochrane Risk of Bias tool.1821 Publication bias was planned to be assessed using funnel plots in three studies,1921 though two of these ultimately found too few studies to do so.20,21 The remaining study did not find evidence of publication bias.19

Of the five systematic reviews considered higher quality, two did not report their findings regarding study bias despite stating that it was assessed,8,19 and only one study pointed to an a priori review protocol.20 Two studies were limited to English language and one only involved two abstract screeners for 10% of titles/abstracts rather than all titles and abstracts.21 Where estimates were combined via meta-analysis, heterogeneity was generally found to be about 0% using I2 for primary outcomes except one which had an I2 of 94%, though they conducted sensitivity analysis to explore this.19

Clinical Studies

One of the four blinded, randomized studies10,2224 provided enough description to judge that allocation concealment was likely successful.22 There was a similar lack of detail on blinding or randomization procedures in those studies claiming to have done so, though balanced baseline characteristics suggest randomization was adequate and the lack of description of blinding is primarily a concern in the one study without an identical comparator.22 This study had ill-defined treatment and control interventions that consisted of multiple components that may or may not have been given to all participants.22 One of the randomized controlled trials was open-label, which may subject the findings to bias induced by knowing the treatment group.25 This study also had an ill-defined comparator group, as some participants (n=4) received combined nicotine patch, bupropion and varenicline instead of a single medication as was initially specified.

The non-randomized studies were limited by their study design, in that their lacks inherent means (usually achieved by randomization) to balance potentially confounding characteristics that could influence the effect estimates. In place of randomization it would have been important for non-randomized studies2734 to consider other participant characteristics that could explain their results and/or use appropriate statistical methods. Three studies adjusted for several potential confounders, such as age, sex, indicators of general health status and baseline smoking profiles,27,29,30 while one accounted for within-person clustering of standard errors while adjusting for age started smoking.32 The remaining studies were either unclear or did not undertake an analysis that could minimize the biases induced by lack of randomization.28,31,33,34 In one of these, the sample size would have been too low for any sophisticated analysis (n=14),33 and one did have balanced baseline characteristics despite no randomization.28

Generalizability of several studies’ findings may be limited due to use of a select population based on the study’s setting and/or recruitment procedures. Two studies used single psychiatric hospitals,22,23 one used a specialist tobacco clinic,27 and two others recruited through flyers and advertising24,25 with unusually high, and potentially coercive, participant compensation in one ($650).24 Studies that recruited from multiple sites10,31,32 or administrative data29 were thought to be more generalizable. Two studies lacked enough detail to assess generalizability.26,28

There were other issues worth mentioning across the studies. First, there was lack of statistical power across several studies. Four studies had a power calculation and were found to be adequately powered27,29,32 while three10,23,30 are potentially adequately powered with sample sizes of n=184,23 n=267,30 and n=1272,10 and one potentially underpowered (n=60),26 though the latter four studies did not mention power calculations. Six studies acknowledged that they may be underpowered,22,24,25,28,31,33 one of which included 14 participants.33 Second, there was a lack of clarity in the presentation of results, for example, some studies provided no results tables or summarized form of results,26,31,34 and some contained post-hoc analyses that were unexplained.24,32,34

A strength across studies was the consistent use of biochemically-verified smoking abstinence through parts per million of expired carbon monoxide, except in one study which relied on self-report 7-day point prevalence abstinence.30 This study also relied on self-report for mental disorder diagnosis to define its population.30 Another strength was a common strategy to avoid the confounding effect of nicotine withdrawal when assessing adverse events and symptoms by starting treatment before a ‘target’ quit date.24,27,27,31,32,34

Guidelines

The guideline was considered high quality as it was developed based on an explicit review protocol, evidence was assessed using Grading of Recommendations Assessment, Development and Evaluation (GRADE), and the recommendations were based on an iterative interpretation of evidence by the Guideline Development Group which involved a large range of stakeholders. There was a detailed description of the search and development methods. The only limitation was that external validation was unclear.35

Summary of Findings

Details of the findings of individual studies are provided in Appendix 4.

What is the clinical evidence regarding the effectiveness of smoking cessation interventions for patients with severe mental illnesses?

Systematic reviews

The first review found evidence that NRT with psychosocial treatment can help reduce smoking, increase quit rates or maintain abstinence with cessation rates ranging from 23.1% to 66%, while bupropion cessation rates ranged from zero to 66% and did not seem to foster maintenance after abstinence was achieved. Depression or schizophrenic symptoms were not found to worsen.16 One other review found that the differences between groups in post-treatment proportions of abstinence ranging from 0.12 to 1.0 with NRT, though these were not described in sufficient detail to interpret.17 The last study attempting to study NRT (in the form of transdermal nicotine patch) could not conclude anything due to small sample sizes and few trials.20

Four of five meta-analyses comparing varenicline to placebo found a significantly higher likelihood of abstinence or smoking cessation in the varenicline group.8,18,20,21 The remaining study did not find a difference (relative risk [RR] = 0.79, 95% confidence interval [CI] 0.58 to 1.08), though it did find that varenicline was associated with increased nausea (RR = 1.79 [95% CI 1.20 to 2.67]). Other studies measuring psychiatric outcomes did not find a difference in side effects including suicidal ideation or depression.1821 When comparing varenicline to bupropion, one study found a ‘higher probability’ that bupropion outranks varenicline, though the conclusion was not firm.8

The findings were similar when comparing bupropion to placebo. There were significant higher likelihood of smoking abstinence and/or cessation in three reviews,6,8,20 and one review found mean differences in proportion abstinent ranging from 0 to 0.77.17 One review additionally found a significant change in cigarettes per day (mean difference = -10.77 [95% CI -16.52 to -5.01]) and two found significantly lowered expired CO at end of treatment (mean difference = -6.80 [95% CI -10.79 to -2.81],20 -6.84ppm [95% CI -11.11 to -2.56]6). There were no differences in mental state outcomes.6,20 When comparing bupropion to any intervention, there were also significant differences in medium (3.5 months) and long term (11.75 months) quit rates (RR = 2.93 [95% CI 1.61 to 5.34] and RR = 3.04 [95% CI 1.10 to 8.42], respectively), but not short-term. Similarly, there were no significant differences in changes in psychiatric symptoms, except one study found significant worsening of cognitive score in bupropion intervention group compared to placebo.21 Bupropion plus NRT compared to placebo plus NRT was not found to increase smoking cessation rates (odds ratio [OR] = 4.13 [95% CI 0.92 to 18.47]).8

Clinical Studies

The three randomized studies evaluating nicotine replacement therapy (NRT)2224 had different conclusions. One study did not find a difference in time to relapse or 6-week or 6-month abstinence among those randomized to active versus placebo NRT patches in a PTSD population.24 The remaining two studies were among inpatients. One also failed to find any difference in 7-day point prevalence abstinence, daily number of cigarettes, expired CO, nicotine craving or other symptom scales.23 The other inpatient study of NRT did find an almost three-fold increase in the likelihood of having a quit attempt (OR = 2.89 [95% CI 1.43 to 5.98]) and in the likelihood of reducing cigarettes smoked per day by 50% (OR = 5.90 [95% CI 2.89 to 15.25]), and a decrease in nicotine dependence (mean difference on FTND = -1.6 [95% CI -2.3 to -0.8]), though no effect on 7-day point prevalence abstinence at six months.22 It is important to note this last study had a more complex intervention that provided additional support to participants. There was limited reporting of adverse events or psychiatric safety outcomes, but in the study of high versus low dose NRT inpatients, five participants discontinued due to unspecified side effects.23 No group differences in psychological distress was found in the other inpatient study.22

The two randomized studies comparing varenicline to placebo both found significant effects on smoking abstinence. At week eight, the varenicline group had significantly lower expired CO compared to placebo (P=0.019), though there were no significant differences in measures of withdrawal, number of cigarettes smoked, or smoking urge.26 The other study using varenicline versus placebo for maintenance therapy found their schizophrenia/bipolar disorder (SBD) population had significantly lower abstinence rates overall compared to the non-SBD population (OR = 0.27, [95% CI 0.13 to 0.56]), except within the varenicline-assigned group, there was no difference (OR = 1.68, [95% C: 0.53 to 5.32]).10 Similarly, among varenicline-assigned participants, there were no significant differences between SBD and others in the rate of 7-day point-prevalence abstinences, though odds of achieving abstinence were lower for SBD patients in the placebo group compared to others (OR = 0.87; P = 0.011).10 Three participants with adverse events (3/60) including nausea (1/30 in each group), headache (1/30 in the varenicline group), and two participants with aggravated psychotic symptoms between weeks 2 and 4 withdrew from one of the trials.26

Two non-randomized studies of varenicline against NRT had opposing results. One study did not find any difference in 7-day point prevalence abstinence or in number attaining > 50% reduction in number of cigarettes per day, expired CO level, or nicotine dependence, against transdermal nicotine patch.28 The other did find an increased likelihood of continuous 2-week abstinence in the varenicline group at the eight week follow up in the total and mental illness population, though the effect was stronger in the mental illness population (OR = 2.88 [95% CI 1.08 to 7.63] in mental illness sample vs OR = 1.70 [95% CI 1.09 to 2.67] in total sample).27

Both non-randomized studies found higher rates of adverse events in the varenicline versus NRT groups. One study found more frequent nausea, disturbed sleep, vivid dreams, drowsiness, constipation, headache, dyspepsia, dry mouth, bad taste, low mood, diarrhea and disorientation in the varenicline group, and seven patients switched from varenicline to NRT due to unspecified adverse symptoms. The rates of adverse events were not different in those with and without mental illness.27 The other study found significant weight gain in both groups, and 10% more participants reporting at least one adverse event in the varenicline group (21/36 [58.3%] in NRT versus 27/39 [69.2%] in the varenicline group), most commonly abnormal/vivid dreams (n=9 and n=4, respectively), constipation (n=5 and n=9, respectively), and nausea/vomiting in the varenicline group only [n=12]). Four (10.2%) switched treatment groups due to adverse events and three (7.8%) reduced their varenicline dose.28

The cohort studies that compared outcomes pre and post treatment with varenicline were generally positive. A significant decrease in cigarettes per day at six months (Mean difference [B] = 13.61 [95% CI 6.58 to 20.75])33 and significant increase in 7-day point prevalence abstinence at 12 weeks were observed (B= 0.34, standard error [SE] = 0.03, P < 0.01).32 The latter study also found significant declines in the urge to smoke and WSWS scores.32 Two studies found that 41.1% (n=37) and 41.3% were abstinent at the end of the 12-week follow up.31,34 One of these further found declines in withdrawal scores among those who attained abstinence, but not among non-abstinent completers or those who dropped out,34 while the other had less conclusive findings regarding nicotine dependence and withdrawal symptoms. Two studies measuring expired CO level had opposite findings, though the one without significant results measured this outcome at six months,33 while the other measured it at twelve weeks (B =-0.03 [SE = 0.01], P < 0.01 ).32 Gastrointestinal issues such as nausea were found in all varenicline non-randomized studies. Sleep disturbance or abnormal dreams was also one of the most common adverse events in two studies,31,34 while suicidal ideation or psychiatric issues resulted in discontinuation in n=1/14,33 n=1/9031 and n=1/11032 participants. Two studies reported a significant decrease in symptom scores over time as measured by CDSS,32,34 and BPRS-Psychosis scale.32 There were no significant changes observed across the other symptom scales in any of the studies.

The two bupropion plus NRT studies also had positive findings. In one study, the two bupropion treatment plus NRT groups (with and without home visits) had significantly greater reduction in cigarettes per day and expired CO level compared to treatment as usual with any first-line medication. The treatment combined with home visits was associated with significant differences in FTND scores at 6 months (change of -2.2 vs -4.2, P < 0.05) and in 7-day point prevalence abstinence rates (X2(1) = 4.8, P = 0.03), compared to treatment as usual.25 The other bupropion study found a slightly increased likelihood of reporting 7-day abstinence when combined with nicotine patch, compared to nicotine patch alone (OR = 2.00 [95% CI 1.14 to 3.50]), and in nicotine patch plus gum compared to nicotine patch alone (OR = 2.10 [95% CI 1.04, 4.23]) though both the confidence intervals’ lower bounds approached the null value of one. No effect was observed for treatment retention, or for any outcome among participants without mental disorders.30

In the open label bupropion study,25 psychiatric and symptom measures were similarly not different across groups as measured by six scales, though 27%, 30% and 46% reported adverse events in treatment with home visits, treatment without home visits and treatment as usual groups respectively, most commonly insomnia (n = 4/34), with vivid dreams (n = 2/34), nausea (n = 2/34), rash (n = 2/34), agitation (n = 1/34).

The final study did not observe a difference in suicide attempts/behaviours, as per ICD-9 classification codes, between participants prescribed NRT or bupropion compared to varenicline after 12 months follow up (hazard ratio [HR] = 0.81 [95% CI 0.51 to 1.28] and HR = 0.37 [95% CI 0.05 to 2.70], respectively).29

What are the evidence-based guidelines regarding smoking cessation interventions for patients with severe mental illnesses?

To help stop or reduce smoking, the NICE guidelines recommended considering NRT or varenicline for people with psychosis or schizophrenia, and bupropion for people with schizophrenia. They also suggest warning patients of an increased risk of neuropsychiatric symptoms and to monitor patients more closely in the first 2-3 weeks. NRT can be offered to inpatients who do not express a desire to stop smoking.35

Limitations

Limitations of this review include the poor quality studies, and the small sample sizes which are evident from the generally large confidence intervals found in several studies. The included systematic reviews found similar quality issues of primary studies, commenting on the small sample sizes,6,18 high/unclear risk of bias due to poor reporting,6,8,21 and interstudy variability,17,18 which make their conclusions weaker.

There was also inconsistency across the clinical studies in terms of their definitions of ‘smoker’, the interventions and comparators, making generalizations about which treatment components are responsible for the observed effects difficult. Most studies focused on people with schizophrenia or schizoaffective disorder, so generalizability beyond this population may be limited, and the longest follow up duration on treatment was six months, except in one study that did not find an increase in suicidal attempts/behaviours after one year.29

Conclusions and Implications for Decision or Policy Making

While concerns about safety may prevent pharmacological options from being used more often among people with severe illness, the current review did not find evidence for an increased risk of adverse events compared to the general population, or worsening of psychiatric symptoms when using NRT, varenicline or bupropion with people with severe mental illness.

All three interventions were found to improve smoking cessation rates compared to placebo or no treatment, varenicline and bupropion may be more effective than NRT, and additional psychosocial support may enhance success. Nausea and sleep disturbances were common when using varenicline, but peaked and declined with time. Sustained effects beyond the treatment period were not generally measured, nor were varying dosages except for one study which found higher NRT dose did not improve outcomes.23

In general, the quality of studies was low and sample sizes small, suggesting the need for better reporting across these studies and more adequately powered analyses.

References

1.
Foa EB, Asnaani A, Rosenfield D, Zandberg LJ, Gariti P, Imms P. Concurrent varenicline and prolonged exposure for patients with nicotine dependence and PTSD: a randomized controlled trial. J Consult Clin Psychol. 2017 Jun 1. Epub ahead of print. [PMC free article: PMC5578908] [PubMed: 28569519]
2.
de Leon J, Diaz FJ. A meta-analysis of worldwide studies demonstrates an association between schizophrenia and tobacco smoking behaviors. Schizophr Res. 2005 Jul 15;76(2-3):135–57. [PubMed: 15949648]
3.
Lawrence D, Kisely S, Pais J. The epidemiology of excess mortality in people with mental illness. Can J Psychiatry. 2010 Dec;55(12):752–60. [PubMed: 21172095]
4.
Rigotti NA. Pharmacotherapy for smoking cessation in adults. In: Post TW, editor. UpToDate [Internet]. Waltham (MA): UpToDate; 2017 [cited 2017 Aug 10]. Available from: www​.uptodate.com Subscription required.
5.
Moran LV, Sampath H, Kochunov P, Hong LE. Brain circuits that link schizophrenia to high risk of cigarette smoking. Schizophr Bull [Internet]. 2013 Nov [cited 2017 Aug 10];39(6):1373–81. Available from: http://www​.ncbi.nlm.nih​.gov/pmc/articles/PMC3796082 [PMC free article: PMC3796082] [PubMed: 23236076]
6.
Tsoi DT, Porwal M, Webster AC. Efficacy and safety of bupropion for smoking cessation and reduction in schizophrenia: systematic review and meta-analysis. Br J Psychiatry [Internet]. 2010 May [cited 2017 Jul 27];196(5):346–53. Available from: http://bjp​.rcpsych.org​/content/bjprcpsych/196/5/346.full.pdf [PubMed: 20435957]
7.
Cook BL, Wayne GF, Kafali EN, Liu Z, Shu C, Flores M. Trends in smoking among adults with mental illness and association between mental health treatment and smoking cessation. JAMA. 2014 Jan 8;311(2):172–82. [PMC free article: PMC5555156] [PubMed: 24399556]
8.
Roberts E, Eden EA, McNeill A, Robson D. Efficacy and tolerability of pharmacotherapy for smoking cessation in adults with serious mental illness: a systematic review and network meta-analysis. Addiction [Internet]. 2016 Apr [cited 2017 Jul 27];111(4):599–612. Available from: https://www​.ncbi.nlm​.nih.gov/pmc/articles/PMC4801667 [PMC free article: PMC4801667] [PubMed: 26594837]
9.
Smith AL, Carter SM, Chapman S, Dunlop SM, Freeman B. Why do smokers try to quit without medication or counselling? A qualitative study with ex-smokers. BMJ Open [Internet]. 2015 Apr 30 [cited 2017 Aug 10];5(4):e007301. Available from: http://www​.ncbi.nlm.nih​.gov/pmc/articles/PMC4420973 [PMC free article: PMC4420973] [PubMed: 25933811]
10.
Evins AE, Hoeppner SS, Schoenfeld DA, Hoeppner BB, Cather C, Pachas GN, et al. Maintenance pharmacotherapy normalizes the relapse curve in recently abstinent tobacco smokers with schizophrenia and bipolar disorder. Schizophr Res. 2017 May;183:124–9. [PMC free article: PMC5432398] [PubMed: 27956009]
11.
Anthenelli RM, Benowitz NL, West R, St Aubin L, McRae T, Lawrence D, et al. Neuropsychiatric safety and efficacy of varenicline, bupropion, and nicotine patch in smokers with and without psychiatric disorders (EAGLES): a double-blind, randomised, placebo-controlled clinical trial. Lancet. 2016 Jun 18;387(10037):2507–20. [PubMed: 27116918]
12.
Shea BJ, Grimshaw JM, Wells GA, Boers M, Andersson N, Hamel C, et al. Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews. BMC Med Res Methodol [Internet]. 2007 [cited 2017 Aug 10];7:10. Available from: http://www​.ncbi.nlm.nih​.gov/pmc/articles/PMC1810543/ [PMC free article: PMC1810543] [PubMed: 17302989]
13.
SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health [Internet]. 1998 Jun [cited 2017 Aug 10];52(6):377–84. Available from: http://www​.ncbi.nlm.nih​.gov/pmc/articles/PMC1756728/ [PMC free article: PMC1756728] [PubMed: 9764259]
14.
Brouwers M, Kho ME, Browman GP, Burgers JS, Cluzeau F, Feder G, et al. AGREE II: advancing guideline development, reporting and evaluation in healthcare. CMAJ [Internet]. 2010 Dec [cited 2017 Aug 10];182(18):E839–E842. Available from: http://www​.ncbi.nlm.nih​.gov/pmc/articles/PMC3001530/ [PMC free article: PMC3001530] [PubMed: 20603348]
15.
Smoking cessation interventions for patients with severe mental illnesses: clinical effectiveness and guidelines [Internet]. Ottawa: CADTH; 2016 Aug 4. [cited 2017 Aug 23]. (Rapid response report: reference list). Available from: https://www​.cadth.ca​/sites/default/files​/pdf/htis/aug-2016/RA0858​%20Smoking%20Cessation​%20Mental%20Illness%20Final.pdf
16.
Bennett ME, Wilson AL, Genderson M, Saperstein AM. Smoking cessation in people with schizophrenia. Curr Drug Abuse Rev. 2013 Sep;6(3):180–90. [PubMed: 23721094]
17.
Ferron JC, Altermann AI, McHugo GJ, Brunette MF, Drake RE. A review of research on smoking cessation interventions for adults with schizophrenia spectrum disorders. Ment Health Subst Use. 2009;2(1):64–79.
18.
Wu Q, Gilbody S, Peckham E, Brabyn S, Parrott S. Varenicline for smoking cessation and reduction in people with severe mental illnesses: systematic review and meta-analysis. Addiction. 2016 Sep;111(9):1554–67. [PubMed: 27043328]
19.
Kishi T, Iwata N. Varenicline for smoking cessation in people with schizophrenia: systematic review and meta-analysis. Eur Arch Psychiatry Clin Neurosci. 2015 Apr;265(3):259–68. [PubMed: 25283510]
20.
Tsoi DT, Porwal M, Webster AC. Interventions for smoking cessation and reduction in individuals with schizophrenia. Cochrane Database Syst Rev. 2013 Feb 28;(2):CD007253. [PMC free article: PMC6486303] [PubMed: 23450574]
21.
Peckham E, Brabyn S, Cook L, Tew G, Gilbody S. Smoking cessation in severe mental ill health: what works? An updated systematic review and meta-analysis. BMC Psychiatry [Internet]. 2017 Jul 14 [cited 2017 Aug 10];17(1):252. Available from: http://www​.ncbi.nlm.nih​.gov/pmc/articles/PMC5513129 [PMC free article: PMC5513129] [PubMed: 28705244]
22.
Stockings EA, Bowman JA, Baker AL, Terry M, Clancy R, Wye PM, et al. Impact of a postdischarge smoking cessation intervention for smokers admitted to an inpatient psychiatric facility: a randomized controlled trial. Nicotine Tob Res. 2014 Nov;16(11):1417–28. [PubMed: 24939916]
23.
Chen HK, Lan TH, Wu BJ. A double-blind randomized clinical trial of different doses of transdermal nicotine patch for smoking reduction and cessation in long-term hospitalized schizophrenic patients. Eur Arch Psychiatry Clin Neurosci. 2013 Feb;263(1):75–82. [PubMed: 22729212]
24.
Dennis PA, Kimbrel NA, Dedert EA, Beckham JC, Dennis MF, Calhoun PS. Supplemental nicotine preloading for smoking cessation in posttraumatic stress disorder: results from a randomized controlled trial. Addict Behav. 2016 Aug;59:24–9. [PMC free article: PMC4852372] [PubMed: 27046670]
25.
Brody AL, Zorick T, Hubert R, Hellemann GS, Balali S, Kawasaki SS, et al. Combination extended smoking cessation treatment plus home visits for smokers with schizophrenia: a randomized controlled trial. Nicotine Tob Res. 2017 Jan;19(1):68–76. [PMC free article: PMC5157714] [PubMed: 27613888]
26.
Jeon DW, Shim JC, Kong BG, Moon JJ, Seo YS, Kim SJ, et al. Adjunctive varenicline treatment for smoking reduction in patients with schizophrenia: a randomized double-blind placebo-controlled trial. Schizophr Res. 2016 Oct;176(2-3):206–11. [PubMed: 27543252]
27.
Stapleton JA, Watson L, Spirling LI, Smith R, Milbrandt A, Ratcliffe M, et al. Varenicline in the routine treatment of tobacco dependence: a pre-post comparison with nicotine replacement therapy and an evaluation in those with mental illness. Addiction. 2008 Jan;103(1):146–54. [PubMed: 18028247]
28.
Garcia-Portilla MP, Garcia-Alvarez L, Sarramea F, Galvan G, Diaz-Mesa E, Bobes-Bascaran T, et al. It is feasible and effective to help patients with severe mental disorders to quit smoking: an ecological pragmatic clinical trial with transdermal nicotine patches and varenicline. Schizophr Res. 2016 Oct;176(2-3):272–80. [PubMed: 27237598]
29.
Wu IH, Chen H, Bordnick P, Essien EJ, Johnson M, Peters RJ, et al. Comparison of suicide attempts/behaviors following smoking cessation treatments among schizophrenic smokers. Arch Psychiatr Nurs. 2017 Feb;31(1):62–7. [PubMed: 28104060]
30.
Loreto AR, Carvalho CFC, Frallonardo FP, Ismael F, Andrade AG, Castaldelli-Maia JM. Smoking cessation treatment for patients with mental disorders using CBT and combined pharmacotherapy. J Dual Diagn. 2017 May 10;1–9. Epub ahead of print. [PubMed: 28489501]
31.
Raich A, Ballbe M, Nieva G, Cano M, Fernandez T, Bruguera E, et al. Safety of varenicline for smoking cessation in psychiatric and addicts patients. Subst Use Misuse. 2016;51(5):649–57. [PubMed: 27050328]
32.
Pachas GN, Cather C, Pratt SA, Hoeppner B, Nino J, Carlini SV, et al. Varenicline for smoking cessation in schizophrenia: safety and effectiveness in a 12-week, open-label trial. J Dual Diagn [Internet]. 2012 [cited 2017 Jul 27];8(2):117–25. Available from: https://www​.ncbi.nlm​.nih.gov/pmc/articles/PMC3414422/ [PMC free article: PMC3414422] [PubMed: 22888309]
33.
Castle D, Baker AL, Richmond R, Filia SL, Harris D, Pirola-Merlo AJ. Varenicline plus healthy lifestyle intervention for smoking cessation in psychotic disorders. Ann Clin Psychiatry. 2012 Nov;24(4):285–91. [PubMed: 23145385]
34.
Cather C, Hoeppner S, Pachas G, Pratt S, Achtyes E, Cieslak KM, et al. Improved depressive symptoms in adults with schizophrenia during a smoking cessation attempt with varenicline and behavioral therapy. J Dual Diagn. 2017;13(3):168–78. [PubMed: 28414583]
35.
Psychosis and schizophrenia in adults: prevention and management [Internet]. London (GB): National Institute for Health and Care Excellence (NICE); 2014. [cited 2017 Aug 10]. (Clinical guideline; no.178). Available from: https://www​.nice.org​.uk/guidance/cg178/resources​/psychosis-and-schizophrenia-in-adults-prevention-and-management-pdf-35109758952133

Appendix 1. Selection of Included Studies

Image app1f1

Appendix 2. Characteristics of Included Publications

Table 1Characteristics of Included Systematic Reviews

First Author, Publication YearTypes and numbers of primary studies includedPopulation CharacteristicsInterventionComparator(s)Clinical Outcomes
Bennett, 201316

Types of studies not described

29 studies included, of which 22 studies included pharmacological intervention (NRT = 2, bupropion = 11, varenicline = 2, NRT + psychosocial intervention = 5, bupropion + psychosocial intervention = 2)

50% or more individuals in sample with schizophrenia spectrum diagnosis"Some individual or group intervention for smoking cessation was provided" (pg 181)None stated"Some smoking-related outcome variable was measured [self-reported smoking, breath carbon monoxide (CO), etc]." (pg 181)
Ferron, 20091713 prospective studies that did not use a single-subject designPeople who smoke, with schizophrenia spectrum disordersNone stated ("an intervention for smoking cessation", pg 66)None statedEliminating or reducing smoking
Kishi, 201519

Double-blinded RCTs only

Seven doubled-blind, randomized, placebo-controlled trials included comparing varenicline to placebo selected

Individuals with schizophrenia, schizoaffective disorder, schizophreniform disorder or delusional disorderVareniclinePlacebo

Smoking abstinence rate

Positive, negative, and depressive symptoms, discontinuation rate, and individual side effects.

Peckham, 201721RCTs only (28 studies included in total, of which 16 focused on pharmacological interventions including n=9 burpoprion, n=6 varenicline, n=1 NRT)Aged 18 or older and ICD or DSM IV diagnosis of schizophrenia or other psychotic disorders, bipolar disorder and depression with psychotic features, but not personality disorder, severe anxiety disorder, PTSD, major depression or autism.Behavioural or pharmacological (any products licensed for smoking cessation) as monotherapy or in combinationEach other, placebo, usual care or no intervention

Biochemically-verified self-reported smoking cessation

Smoking reduction, change in body weight, change in psychiatric symptoms (any validated scale)

Adverse events

Roberts, 20168

Included RCTs only

17 studies representing 14 RCTs were included (bupropion to placebo = 6, varenicline to placebo = 5, bupropion to varenicline and placebo = 1, bupropion plus NRT to placebo plus NRT = 2)

Adults who currently smoke with motivation to quit or reduce smoking, with severe mental illness including schizophrenia, schizoaffective disorder, bipolar disorder, delusional disorder or depressive psychosesNRT, bupropion or varenicline alone or in combination, with or without psychological supportPlacebo or other intervention

Sustained smoking cessation (6 months or longest reported time point)

7-day point prevalence abstinence rate

Discontinuation due to adverse events

Tsoi, 20106RCTs only (included 21 reports of 7 trials)Smokers with a current diagnosis of schizophrenia according to either the ICD-10 or the DSM-IV were included.Bupropion, including bupropion combined with other pharmacological or non-pharmacological interventionsPlacebo

Abstinence

Change in severity of smoking dependence

Change in mental state (positive, negative and depressive symptoms using validated tools)

Adverse events

Tsoi, 20132034 randomised trials, of which 10 focused on pharmacological interventions (n=7 bupropion to placebo, n=2 varenicline to placebo, n=1 contingent reinforcement vs NRT)Adult smokers with ICD or DSM diagnosis of schizophrenia or schizoaffective disorder

4 trials recruited inpatients only

Pharmacological or non-pharmacological interventions (alone or in combination) for smoking cessation/reduction.Another pharmacological or non-pharmacological intervention, placebo or usual care

Smoking abstinence at longest follow-up

Change in mental state

Smoking abstinence at end of intervention

Reduction of smoking behaviour or dependence

Other adverse events

Wu, 201618Included RCTs and quasi-randomized controlled trials (8 RCT studies comparing varenicline to placebo selected)Adults (over age 18) with any type of severe mental illnessVareniclinePlacebo or other pharmacotherapy interventionSmoking cessation, reduction in number of cigarettes per day and safety (number of psychiatric adverse events)

RCT = randomized controlled trial; NRT = nicotine replacement therapy; ICD = International Classification of Disease; DSM = Diagnostic and Statistical Manual of Mental Disorders

Table 2Characteristics of Included Clinical Studies

First Author, Publication Year, Country, Study NameStudy DesignPatient CharacteristicsIntervention(s)Comparator(s)Clinical Outcomes
Brody, 2017, U.S. Combination extended smoking cessation treatment plus home visits for smokers with schizophrenia: a randomized controlled trial25Open label RCTAdult male smokers with DSM IV criteria diagnosis of schizophrenia who smoke 10 to wu cigarettes per day and a desire for cessation treatment living independently or in a supervised arrangement.Bupropion plus nicotine patch, nicotine lozenge, + group CBT (with and without home visits by study investigator)Treatment as usual (group CBT plus single first-line smoking cessation medication)

7-day point prevalence abstinence

Number of cigarettes

Dependence, withdrawal, urge (FTND, WSWS)

Safety measures: BPRS, SANS, CGI, BDI, C-SSRS, AIMS

Castle, 2012, Australia, Varenicline plus healthy lifestyle intervention for smoking cessation in psychotic disorders33Pre/post studyPatients at least 18 years old with psychotic disorder based in Mini International Neuropsychiatric Interview, with stable medication for last 3 months and smoking at least 15 cigarettes per day.Varenicline 0.5 mg/d for days 1 to 3; 1 mg/d for days 4 to 7; and 2 mg/d days 8 to 84 + Healthy Lifestyles programNone

Expired CO

Opiate Treatment Index

Dependence, withdrawal, urge (FTND, MNWS)

Psychiatric symptoms (BPRS, BDI, Young Mania Rating Scale)

Safety check using C-SSRS

Cather, 2017, U.S., Improved depressive symptoms in adults with schizophrenia during a smoking cessation attempt with varenicline and behavioral therapy34Pre/post studyParticipants with schizophrenia or schizoaffective disorder, depressed type (SSD), aged 18 to 70 years, reported smoking 10 cigarettes per day for at least the past year with expired CO concentration >= ppm, taking stable dose of antipsychotic medication for 30+ days, with a desire to quit smoking.Varenicline 0.5 mg daily for three days, 0.5 mg twice daily for four days, and 1 mg twice daily for 11 weeks + group CBT tailored for smokers with severe mental illnessNone

End-of-study abstinence

Depressive symptoms, withrawal (CDSS, WSWS)

Chen, 2013, Taiwan, A double-blind randomized clinical trial of different doses of transdermal nicotine patch for smoking reduction and cessation in long-term hospitalized schizophrenic patientsBlinded RCTPatients from chronic wards of psychiatric hospital meeting DSM-IV criteria for schizophrenia or shizoaffective disorders, who were regular smokers.High dose NRT (31.2mg TNP) for 4 weeks, then 20.8mg for 4 weeks + group psycho-educationLow dose NRT (20.8mg TNP) for 8 weeks + group psycho-education

Number of cigarettes

7-day smoking abstinence

Expired CO

Dependence (FTND)

Psychiatric symptoms (PANSS, SARS)

Dennis, 2016, US, Supplemental nicotine preloading for smoking cessation in posttraumatic stress disorder: results from a randomized controlled trial24Blinded RCTIndividuals with PTSD (Clinician Administered PTSD Scale) aged 18-70, with a desire to quit smoking and smoking at least 10 cigarettes per day in the past year.2 weeks active nicotine patch loading (21mg/24 h patch) prior to quit date + individual CBT + six weeks 21mg/24 h patch post-quit and one form of rescue nicotine replacement (e.g. gum)2 weeks placebo nicotine patch prior to quit date + individual CBT + six weeks placebo patch post-quit

7-day point prevalence abstinence

Thirty-day smoking abstinence

Diary assessments for smoking frequency and craving, and PTSD symptoms

Evins, 2017, Multiple. Maintenance pharmacotherapy normalizes the relapse curve in recently abstinent tobacco smokers with schizophrenia and bipolar disorder10Blinded pooled analysis of two RCTsParticipants with and without schizophrenia or bipolar disorder with a 14 or 7 day point prevalence abstinence at week 12 of the initial cessation treatmentVarenicline treatment (1.0mg twice per day) + tapering behavioral supportIdentical placebo + tapering behavioral support

Four-week continuous abstinence rates after 12 weeks of maintenance therapy

Time to first relapse

Garcia-Portilla, 2016, Spain, It is feasible and effective to help patients with severe mental disorders to quit smoking: An ecological pragmatic clinical trial with transdermal nicotine patches and varenicline28Cohort studyPatients with DSM-IV diagnoses of schizophrenia, schizoaffective or bipolar disorder, smoking >= 15 cigarettes/day for at least 1 year without any abstinence periods longer than 1 month, FTND score >=4, expired CO > 9 ppm, aged 18-65Varenicline 0.5 mg/day for the first 3 days, 0.5 mg twice daily on days 4-7, and 1 mg twice daily for the remaining 11 weeks. (preceded by 4-12 weeks individual motivational therapy)TNP 24 hour at 14, 21, 28 or 35 mg. (preceded by 4-12 weeks individual motivational therapy)

7-day point prevalence abstinence

Proportion with >= 50% reduction in number of cigarettes per day in the last week

Dependence, withdrawal, urge (FTND, Glover-Nilsson Smoking Behavioural Questionnaire)

Jeon, 2016, Korea, Adjunctive varenicline treatment for smoking reduction in patients with schizophrenia: a randomized double-blind placebo-controlled trial26Blinded RCT secondary analysisAged 18 to 60 years, score 75 or less on PANSS and no medication changes for last 3 months, with DSM-IV diagnosis of schizophrenia receiving antipsychotic medication. 'Smoker' = smoking > 10 cigarettes daily for at least 1 year with expired CO level >10ppm.Varenicline, 0.5 mg for days 1-3, 0.5 mg twice per day for days 4-7, and 1 mg twice daily for weeks 2-8 + self-help booklet and weekly telephone visits.Identical placebo + self-help booklet and weekly telephone visits.

Expired CO

Number of cigarettes

Dependence, withdrawal, urge (MNWS, Brief Questionnaire of Smoking Urge, m-CEQ)

Psychiatric symptom/safety (HADS, PANSS, SANS, SARS, CGI, Barnes Akathisia Scale)

Loreto, 2017, Brazil, Smoking cessation treatment for patients with mental disorders using CBT and combined pharmacotherapy30Retrospective cohortAged 18 to 65, attended health centre staff at least once during treatment protocol, voluntarily participate in the smoking cessation treatment, with a prescription for nicotine patch alone or in combination with another medication.

Nicotine patch plus bupropion

Nicotine patch plus gum

Nicotine patch plus nortriptyline All interventions included group CBT

Nicotine patch + group CBT

7-day point prevalence abstinence

Retention

Pachas, 2012, US, Varenicline for smoking cessation in schizophrenia: safety and effectiveness in a 12-week, open-label trial32Pre/post studyAged 18-70 with DSM-IV-text revision diagnosis of schizophrenia or schizoaffective disorder, smokied >= 10 cigarettes per day, clinically stable (stable antipsychotic medication for >= 1 month), expired CO > 9ppm, with desire to quit smoking. Excluded those with dementia, other substance use disorder in last 6 months and hospitalization for suicidal ideation in past 12 monthsVarenicline 0.5 mg/day for 3 days, 0.5 mg/day for 4 days, then 1 mg twice daily for 11 weeks and weekly, + group CBT based on Freedom From Smoking programNone

7-day point prevalence abstinence

Weeks of continuous abstinence

Dependence, withdrawal, urge (FTND, WSWS)

Psychiatric symptoms (SANS, CDSS, BPRS)

Raich, 2016, Spain, Safety of varenicline for smoking cessation in psychiatric and addicts patients31Pre/post studyAged 18 or older with a diagnosis of nicotine dependence, who smoked more than 10 cigarettes/day in the past year with a desire to quit, with psychotic disorder (schizophrenia, schizoaffective disorder or chronic delusion)Varenicline 0.5 mg per day for 3 days, 0.5 mg twice a day from days 4 to 7, and 1 mg twice a day during the following 11 weeks + initial visit with CBTNone

End-of-study abstinence

Stapleton, 2008, UK, Varenicline in the routine treatment of tobacco dependence: a pre-post comparison with nicotine replacement therapy and an evaluation in those with mental illness27Cohort studyNo exclusions except pregnant or breastfeeding, trying to conceive, under 18 years old, severe renal function impairment (3) were excluded routinely from 111 (67%) had mental illness with primary diagnosis: 64/111 (58%) depression, 14/111(13%) bipolar disorder, 7/111(6%) h psychosis, 24/111(22%) psychosis and depression and 2/111(2%) eating disorders.Varenicline (12-week course) plus group support sessionsVariable dose NRT plus group support sessions 1.5 hours (60% nicotine patch, 25% nasal spray, 11% gum or lozenge and 5% inhalator/microtab, + offer of 2nd NRT product to be used in combination)

End-of-study abstinence

Self-completion tobacco withdrawal symptoms scale

Stockings, 2014, Australia, Impact of a postdischarge smoking cessation intervention for smokers admitted to an inpatient psychiatric facility: a randomized controlled trial22Blinded RCTPatients of psychiatric facility over 12 months (May 2010-11) who are at least 18 years old and self-reported smoker; diagnoses (reduced to mood disorders, schizophrenia and related psychosis, other)Treatment as usual + motivational interview and 2-week supply of NRT, nightly phone calls with offer for 12-week additional supply of NRT, referral to quit service, smoking cessation groupsTreatment as usual (any of: brief advice to quit, NRT during admission and for 3 days postdischarge, smoking care plan in discharge summary)

Continuous abstinence (from date of discharge)

7-day point prevalence abstinence

Prevalence of quit attempts (not smoking for at least 24 hours)

Number of cigarettes

dependence (FTND)

Psychological distress (K10))

Wu, 2017, U.S., Comparison of suicide attempts/behaviors following smoking cessation treatments among schizophrenic smokers29Retrospective cohortEnrolled in database between 12/13/1995 and 10/31/2011 with ICD-9 code diagnosis of schizophrenia or schizoaffective disorder, above age 18, not prescribed Bupropion for depression in 6 months prior, not prescribed other medication on same day as index medication and newly initiated cessation medicationVarenicline NRT BupropionVarenicline used as referenceSuicidal Attempts/Behaviors

RCT = randomized controlled trial; NRT = nicotine replacement therapy; TNP = transdermal nicotine patch; CO = carbon monoxide; ppm = parts per million; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders; CBT = cognitive behavioural therapy; FTND = Fagerstrom Test for Nicotine Dependence;MNWS = Minnesota Nicotine Withdrawal Scale; WSWS = Wisconsin Smoking Withdrawal Scale; mCEQ = Modified Cigarette Evaluation Questionnaire; BPRS = Brief Psychiatric Rating Scale; SARS = Simpson-Angus Rating Scale; PANSS = Positive and Negative Syndrom Scale; HADS = amilton Rating Scale for Depression; CDSS = Calgary Depression Scale for Schizophrenia; BDI = Beck Depression Inventory; C-SSRS = Columbia- Suicide Severity Rating Scale; AIMS = Abnormal Involuntary Movement Scale; K10 = Kessler Psychological Distress Scale; CGI = Clinical Global Impression

Table 3Characteristics of Included Guidelines

CitationIntended users/Target popIntervention and Practice ConsideredOutcomesEvidence collection, selection and synthesisEvidence Quality and StrengthRecommendations development and Evaluation
NICE35Target population is adults with schizophrenia (including schizophrenia-related disorders such as schizoaffective disorder and delusional disorder) or psychosis, and intended users are primary, community, secondary, tertiary and other healthcare professionals who make decisions concerning their careBupropion, varenicline and transdermal nicotine patchAny

Literature search involved 15 databases and based on Cochrane review with a new search to update

Meta-analysis complete where possible

For bupropion vs placebo, three studies ranked 'moderate' and remaining 4 were low or very low

For varenicline vs placebo, both studies graded 'low'

Nicotine patch study was from a conference paper

Guideline Development Group drafted recommendations based on evidence summaries, with assistance of special advisors

RCT = randomized controlled trial

Appendix 3. Critical Appraisal of Included Publications

Table 4Strengths and Limitations of Systematic Reviews and Meta-Analyses using AMSTAR12

StrengthsLimitations
Bennett16

Three databases searched (Pubmed, Medline, Psycinfo)

No dates specified

Limited key words used (only “smoking cessation”, “nicotine dependence” “schizophrenia”, “serious mental illness”, and “smoking treatment outcomes”.)

Research question vague

No critical appraisal or table of study descriptions

No flow chart of study selection or procedures for data extraction or abstract review

Limited to English language

Ferron17

Hand searching of references

Only two databases searched (PsychInfo and PubMed)

Limited number of search terms

No flow chart or details on extraction procedures

Research question vague

No risk of publication or study bias assessed

Unclear statistical methods to determine effect sizes

Kishi19

PubMed, the Cochrane Library databases, and PsycINFO up to 2014

No language restriction

Hand searched reference lists

Trials identified by two authors

Quality assessed using Cochrane risk of bias assessment

Combined results using random effects model

Reported relative, absolute and measures of impact

Funnel plot to assess publication bias

Double-blinded RCTs only

Limited databases

Did not report quality findings

Peckham21

Registered review protocol

Searched 6 databases (MEDLINE (PubMed), EMBASE, PsycINFO, CINAHL, Health Management Information Consortium (HMIC) and CENTRAL)

Search strategy adapted from terms developed by Cochrane group

Hand-searched reference lists

2 authors independently decided which studies to include and extracted the data

Cochrane risk of bias tool used to assess study quality

Appropriate statistical methods for meta-analysis

Assessed publication bias risk through funnel plots

Limited to English

Did not have 2 independent screeners for all studies (validated in 10% of studies only)

Roberts8

Embase, Medline, PsychINFO and the Cochrane Central Register of Controlled Trials searched until 2014

Abstracts assessed by two authors + hand searching for additional citations

Quality assessed using GRADE

Imputation of outcomes for missing participants as continuing smokers

Appropriate methods for meta-analysis, i.e. random effects model for pairwise comparisons, and Bayesian random effects for network linking

Limited to RCTs and English language studies

No exploration of publication bias and did not report quality findings

Tsoi6

Included Cochrane Central Register of Controlled Trials of the Cochrane Library, MEDLINE, EMBASE and PsycINFO, as well as unpublished studies, conference abstracts, trial records and reference lists

No date or language restrictions

2 authors independently decided which studies to include and extracted the data

Assessed study quality based on allocation concealment, masking, completeness of follow-up and ITT analysis

Assessment of heterogeneity and appropriate methods for meta-analysis

Did not use external quality assessment tool

No assessment of publication bias

Tsoi20

Clear PICO question

A priori protocol

Inclusion of both RCTs and quasi-RCTs

Searched several databases including Cochrane Tobacco Addiction Group Specialised Register, CENTRAL (the Cochrane Central Register of Controlled Trials), MEDLINE, EMBASE, PsycINFO, CINAHL, BIOSIS Previews and Web of Science, in addition to trial registry platforms

No date or language restrictions

Three authors independent screening and two authors extracted data

Risk of bias assessed using Cochrane methodology, heterogeneity and reporting biases assessed using funnel plots

Appropriate meta-analysis methods

Lack of summary of risk of bias across studies

Wu18

Searched MEDLINE, EMBASE, PsycINFO, CINAHL and the Cochrane Library in September 2015

No language restrictions

Hand searched reference lists

Trials identified by two authors

Quality assessed using Cochrane risk of bias assessment

Combined results using random effects model

Report both risk ratio and odds ratio

Unclear how 'smoker' was defined

No exploration of publication bias

RCT = randomized controlled trial

Table 5Strengths and Limitations of Clinical Studies using Downs and Black13

StrengthsLimitations
Evins10

Blinded, randomized controlled trial

International patient population improves generalizability

Appropriate statistical analysis with covariates including gender, race, age, and severity of nicotine dependence

Exploration of missing data impact using imputation models, and appropriate combining of imputed datasets

Large sample size (n=1293)

Sensitivity analysis to assess the impact of decisions on missing data handling and baseline differences

Lack of detail on blinding or allocation concealment procedures, or inclusion and exclusion criteria

No table to compare characteristics of placebo versus treatment groups, or to present regression results

Present odds ratios despite using survival analysis (which is usually presented as hazard ratios)

Chen23

Blinded, randomized trial

Appropriate statistical analysis

ITT analysis

Successful randomization as balanced baseline characteristics achieved

Unclear definition of smoker

No detail on allocation concealment, blinding, randomization or recruitment procedures

Public psychiatric hospital setting only

Lack of power calculation despite conducting pilot study, however sample size relative large (n=184)

Did not provide exact p-values

Dennis24

Blinded, randomized trial

Blinding achieved through making the list unavailable to study coordinators or investigators

Appropriate multi-level statistical modeling to account for repeat measures on each person

Baseline characteristics balanced

Recruitment through flyers/letters could induce selection bias

High compensation ($650) potentially coercive

Unclear whether rescue NRT (e.g. gum) was also given in placebo format, or how this was used

Lack of detail on allocation concealment, study setting or power calculation

Post-hoc statistical analysis without mention in methods

Unclear mixing of ITT and as-treated analysis

No PRISMA diagram

Stockings22

Blinded, randomized trial

Evidence of successful allocation concealment from all project and clinical staff and follow up interviewers

Patients approached upon admission to determine eligibility, minimizing selection bias

Successful randomization except higher occurrence of bipolar disorders as primary diagnosis (mean 9.9 in control vs 18.3 in intervention group)

ITT analysis using generalized mixed modelling to examine differences over time, or odds ratios otherwise

Single site may limit generalizability

Treatment and control interventions not well-defined

Jeon26

Blinded, randomized trial

Successful blinding through identical capsules

Balanced baseline characteristics achieved, though slightly elevated average daily neuroleptic dose in placebo group (mean = 624.17mg versus 457.93 mg in varenicline group)

No description of randomization procedures

Recruitment procedures not described

Statistical analysis questionable due to use of repeated-measures ANOVA for time trajectories, as only results in a p-value

Unclear presentation of results, and difficult to decipher the specific contrasts

Garcia-Portilla28

Measured baseline characteristics were generally balanced despite no randomization

Statistical analysis was appropriate except if there were unmeasured, unbalanced characteristics predicting the outcome, which should have been adjusted for

'Real world' setting

Non-randomized; treatment given according to patient and provider preferences, and somatic comorbidities (which can also be a consequence of the treatment, so adverse events may be underestimated)

No details on recruitment procedures so generalizability unclear

Open label

Comparison not well-defined as nicotine patches were given at several different doses without stratification

Last observation carried forward (LOCF) method for missing data may not be appropriate if participant relapses after loss to follow up

4 patients changed from treatment to comparator due to adverse event (elevated liver enzymes, nausea) or at their own request

Did not adjust for site clustering in statistical analysis

Stapleton27

Power calculation demonstrates sufficiently powered study

Baseline characteristics across groups were similar

Appropriate statistical model, adjusting for sex, race, education, age, on benefits, other health disorder, tobacco smoking severity, though unclear accounting for correlation within individuals

Large sample size overall (n=404)

Recruitment for specialist tobacco clinic may limit generalizability

Mental illness participants not specifically recruited, and only a secondary analysis

Treatment not randomized, based on patient preferences

Not feasible to compare adverse event occurrence because of low sample sizes in mental illness group across individual symptoms

Brody25

Balanced randomization achieved, except treatment group with home visits had higher average scores on Brief Psychiatric rating Scale (49.4 versus 45.3 and 47.1 in TAU and treatment group with no home visits, respectively) and Beck Depression Inventory (14.2 versus 10.2 and 10.3 in TAU and treatment group with no home visits, respectively)

Generalized linear mixed model used to estimate time trajectories of cigarettes per day, while controlling for within-person effects

Pilot study so potentially underpowered

Recruited via flyer advertisements from the smoking and schizophrenia treatment programs = select population

Small sample size (n=34) reduces interpretability of statistical model results, especially due to the complexity of the model choice (including an unstructured trajectory over time)

Possibility of spill-over effects as some participant in TAU (n=4) received all three medications as well

Castle33

Appropriate statistical analysis (t-test) given the low sample size

No objective measure of smoking

Small sample size (n=14)

No randomization or control, so cannot rule out other factors explaining the observed differences

No important characteristics of individuals reported

Cather34

Lead-in period of one month prior to the quit date used to distinguish adverse events associated with pharmacotherapy from nicotine withdrawal

Not possible to differentiate effect of medication versus behavioural therapy on the outcomes, though stronger case for medication effect as effect is seen immediately after initiation

No confidence intervals or results table from statistical models available, as well as several comparisons that were not stated a priori, limiting interpretability of the results

Unclear design: authors call the study a 'trial', but design would suggest this is more like a pre/post study as the comparison is the subjects' own baseline score; if a case-crossover design, statistical analysis was not typical of matched study (i.e. conditional logistic)

Unclear whether model was adjusted for covariates

Non-randomized study with unclear comparator group

Raich31

Multi-centre study: 11 specialized tobacco addiction units of the Substance Abuse Treatment Network of Catalonia

'Real life' setting may improve generalizability

High drop-out rate (53/90) = 59%

May be underpowered because could only recruit 90 patients versus 129 calculated as required

Treatment not randomized

No comparator group

Statistical analysis did not adjust for any other factors that may have changed - no parametric model

Unclear presentation of results; no table and single report that all were abstinent

Loreto30

Primary outcomes not stated a priori

Power calculation demonstrated sufficient power

Adjusted p-values for multiple testing

Strong statistical analysis account for correlation between repeat observations and adjusted for baseline demographic descriptors, biological descriptors, and indices of nicotine dependence

Investigated confounding potential of retention, though arguably unnecessary for ITT analysis

Missing measurements considered smokers

Multi-centre study (major hospitals and universities in partnership with local mental health centers)

Unclear how treatment was assigned, randomization or recruitment procedures

No comparison group except self, so cannot rule out other explanatory factors

Wu29

Use of large EMR database with over 7.4 million patients, and large sample size meeting eligibility criteria (n=3925)

Appropriate use of Cox regression model to analyse time to suicide/censoring

Adjusted for mostly appropriate covariates (age, race, gender, region, BMI, payment type (government or non-government insurance), specialty group, comorbidity index, and severity of mental disorder

Representative population

Use of prescription as exposure does not mean patients filled prescription

Censoring of patients who switched smoking cessation medication could exclude those at highest risk of suicide (bias effect towards the null)

Adjusting for previous suicide behaviors/attempts or depression may induce confounding if there are common causes of the initial suicide attempt/behaviour and the observed one

Potential for misclassification using ICD-9 codes for suicidal/self-injurious behavior

Non-randomized study with no comparator group

Not possible to discern whether participants actually smoked or the actual indication for the medication

Lack of description of missing data and potential impact

BMI = body mass index; RCT= randomized controlled trial; ICD = International Classification of Disease; NRT = nicotine replacement therapy; TAU = treatment as usual; ITT = intention-to-treat; EMR = electronic medical record

Table 6Strengths and Limitations of Guidelines using AGREE II14

StrengthsLimitations
National Institutes for Health and Care Excellence35

Review protocol for question related to smoking cessation interventions stated clearly

Assessed strength of evidence using GRADE

Interpretation of evidence by the Guideline Development Group, which involves a wide range of stakeholders including patients

Detailed description of search and development methods

Benefits and harm trade off considered by Guidelines Development Group in making the recommendations

Updated 1 existing Cochrane on the topic review but did not consider other sources or systematic reviews

No evidence of external validation

Appendix 4. Main Study Findings and Author’s Conclusions

Table 7Summary of Findings of Included Studies

Main Study FindingsAuthor’s Conclusion
National Institute for Health and Care Excellence35
From page 183:
“7.3.8.1 Offer people with psychosis or schizophrenia who smoke help to stop smoking, even if previous attempts have been unsuccessful. Be aware of the potential significant impact of reducing cigarette smoking on the metabolism of other drugs, particularly clozapine and olanzapine. [new 2014]
7.3.8.2 Consider one of the following to help people stop smoking:
  • nicotine replacement therapy (usually a combination of transdermal patches with a short-acting product such as an inhalator, gum, lozenges or spray) for people with psychosis or schizophrenia or
  • bupropion for people with a diagnosis of schizophrenia or
  • varenicline for people with psychosis or schizophrenia.

Warn people taking bupropion or varenicline that there is an increased risk of adverse neuropsychiatric symptoms and monitor them regularly, particularly in the first 2-3 weeks. [new 2014]
7.3.8.3 For people in inpatient settings who do not want to stop smoking, offer nicotine replacement therapy to help them to reduce or temporarily stop smoking. [new 2014]”
N/A
Bennett, 201316

Three studies of NRT combined with psychosocial treatment suggest "can help smokers with schizophrenia reduce or quit smoking or maintain abstinence following quitting."- cessation rates ranged from 23.1% to 66% (pg 181)

Eight studies of buproprion alone or with NRT or a psychosocial intervention suggest "bupropion is associated with greater reduction and cessation than placebo while treatment is active but does not generally foster maintenance of gains once medication is removed" (pg 184), though it also does not make depression or schizophrenia symptoms worse - cessation rates ranged from no impact to 66%

Only two studies with varenicline are "promising" (pg 184)

"Overall, both pharmacologic and psychosocial smoking cessation treatments have been found to be useful in helping individuals" (pg 186)
Ferron, 200917
NRT vs treatment as usual at 1 or 3 months follow up

Effect sizes ranged from 0.12 to 1.0 for follow up proportions (presumably abstinent) following arcsine transformations Bupropion vs placebo at 3 months follow up

Effect sizes ranged from 0 to 0.77 for follow up proportions (presumably abstinent) following arcsine transformations

Treatments not found to be toxic

"Preliminary data show modest efficacy of nicotine replacement therapy, psychosocial interventions and bupropion"(pg 64)
Kishi, 201519
Varenicline vs Placebo (5 papers; I2 = 94%)
Smoking cessation: RR = 0.79 (95% CI 0.58, 1.08)
Abnormal dreams: RR = 0.47 (95 % CI 0.22, 0.99, p = 0.05, I2 = 0 %) Nausea: RR = 1.79 (95 % CI 1.20,2.67, p = 0.004, I2 = 10 %)

No significant difference in the discontinuation rates, suicidal ideation, depression, other side effects

"Varenicline adjuvant therapy was not more efficacious than placebo for smoking cessation in individuals with SZ."(pg 265)
Peckham, 201721
Addition of buproprion to some intervention (8 trials; I2 = 0%)

Quit rate short term (median 4 weeks): RR = 6.42 (95% CI 0.82, 50.1)

Quit rate medium term (median 3.5 months): RR = 2.93 (95% CI 1.61, 5.34)

Quit rate long term (median 11.75 months): RR = 3.04 (95% CI 1.10, 8.42)


Varenicline versus placebo (5 trials, I2 = 0%)

Quit rate medium term (median 6 months): RR = 4.13 (95% CI 1.36, 12.53)

No significant differences in changes in psychiatric symptoms across 22 studies, except one study found significant worsening of cognitive score in bupropion intervention group compared to placebo

"In line with the results of our previous review, this updated review indicates that people with [severe mental illness(SMI)] can quit smoking and the same interventions that work for people in the general population work for people with SMI e.g. the use of varenicline, bupropion or NRT to support a quit attempt" (pg 13)
Roberts, 20168
Network meta-analysis for smoking cessation

Bupropion vs placebo: OR = 4.51 (95% CI 1.45, 14.04)

Varenicline vs placebo: OR = 5.17 (95% CI 1.78, 15.06)

Bupropion vs varenicline: OR = 1.15 (95% CI 0.24, 5.45)


Direct pairwise meta-analysis for smoking cessation

Bupropion plus NRT compared to placebo plus NRT: OR = 4.13 (95% CI 0.92, 18.47)


Tolerability

No significant differences in drop-out rate in any of the comparisons

"Bupropion and varenicline are effective and tolerable for smoking cessation in adults with SMI. Both varenicline and bupropion had superior treatment efficacy to placebo and were not different from each other." (pg 7)
Tsoi, 20106
Bupropion vs placebo
End-of-treatment abstinence (6 trials, I2 = 0%):
RR = 2.57 (95% CI 1.35, 4.88)
6-month abstinence (5 trials):
RR = 2.78 (95% CI 1.02, 7.58)
End-of-treatment expired CO (3 trials, I2 = 0%):
Mean difference = -6.84 ppm (95% CI -11.11, -2.56)
6-month expired CO (3 trials, I2 = 83%):
Mean difference = -5.73 ppm (95% CI -18.09, 6.63)

No significant worsening of positive, negative and depressive symptoms Safety

Significantly higher dry month, concentration, jitteriness, light-headedness, muscle stiffness and frequent nocturnal awakening reported in one study

Discontinuation from bupropion + NRT (n=2) due to insomnia and dizziness in one study

"Smokers with schizophrenia who used bupropion to aid smoking cessation had a two and a half times higher rate of abstinence at the end of the drug therapy compared with those who did not use bupropion....Although some side-effects of treatment that might be important to individuals were noted, there were no significant serious adverse clinical events such as seizure." (pg 349)
Tsoi, 201320
Bupropion vs placebo
Smoking abstinence at 6-month follow up (5 trials, I2 = 0.00)
RR = 2.78 (95% CI 1.02,7.58)
Smoking abstinence at end of treatment (7 trials, I2 = 0.00)
RR = 3.03 [1.69, 5.42]
Change in Expired CO by end of treatment (4 trials, I2 = 0%):
Mean difference = -6.80 ppm (95% CI -10.79, -2.81)
Change in Expired CO at six months (3 trials, I2 = 83%):
Mean difference =-5.55 ppm (95% CI -17.89, 6.78)
Change in cigarettes per day at end of treatment in abstinence studies (3 trials, I2 = 40%):
Mean difference = -10.77 (95% CI -16.52, -5.01)
Change in cigarettes per day at end of treatment among reduction studies (2 trials):
Mean difference = -2.61 (95% CI -7.99, 2.77)
Change in cigarettes per day at six months (2 trials, I2 = 0%):
Mean difference = 0.40 (95% CI -5.72, 6.53)

No significant differences between positive, negative or depressive symptoms


Varenicline vs placebo (2 trials)

Abstinence at end of treatment (2 trials, I2 = 0%): RR = 4.74 (95% CI 1.34, 16.71)

Abstinence at 6-month follow up (1 trial): RR = 5.06 (95% CI 0.67, 38.24)

No significant differences between positive, negative or depressive symptoms

Change in cigarettes per day (1 trial): Mean difference = 3 (95% CI 0.4, 6.1)


TNP

"Unclear whether transdermal nicotine patch (TNP) helped smoking cessation in this group of patients, as it was tested in only a few trials with small sample sizes" (pg 23)

"Our review supports the effectiveness of bupropion for smoking cessation in patients with schizophrenia…[we found] no evidence of any significant deterioration of mental state secondary to use of bupropion in people with schizophrenia…The evidence for bupropion as an aid to smoking reduction in people with schizophrenia is inconclusive."
"We also found some evidence in support of varenicline for smoking cessation among individuals with schizophrenia…although there is no evidence that varenicline worsens symptoms in schizophrenia, there is some concern about serious adverse events such as suicidal ideation or behaviour among schizophrenia patients on varenicline"
"For other drug treatments (including NRT) and psychosocial interventions, we did not find sufficient convincing evidence in to support their use in clinical practice." (pg 26-7)
Wu, 201618
Varenicline versus placebo
Abstinence rates at 12 weeks (four studies I2 = 0%, p = 0.91)
RR = 4.33 (95% CI 1.96, 9.56)
Reduction in number of cigarettes smoked at treatment end (five studies I2=89.2%):
Mean difference = 6.39 (95% CI = 2.22, 10.46)
Psychiatric symptoms (low number of events)
Suicidal ideation (4 studies - I2 = 0%):
RR = 1.06 (95% CI 0.40, 2.82)
Depressed mood (3 studies - I2 = 28.6%):
RR = 1.45 (95% CI 0.45,4.64)
Anxiety (4 studies - I2 = 33.7%):
RR = 0.77 (95% CI 0.28,2.17)
Other side effects
No significant differences between the groups across 35 types of adverse events
Most common were nausea (n=46/158 varenicline vs n=25/114 placebo), abnormal dreams(n = 27/158 vs n = 24/114), abdominal pain (n = 25/158 vs n = 18/114), insomnia (n = 31/158 vs n = 21/114), fatigue/lethargy (n = 26/158 vs n = 17/114)
"The results of our meta-analysis suggest that varenicline reduced smoking significantly in people with SMI compared with placebo. Given that estimates of the rate of quitting vary markedly among studies, in terms of length of time quit, length of follow-up and means of measurement, the use of changes in daily cigarette consumption as a measurement of smoking behaviour change facilitates direct comparison across studies" (pg 111)
Evins, 201710
SBD vs general population continuous abstinence rate at week 24
Overall: OR = 0.27, 95% CI: 0.13, 0.56, p<0.001
With varenicline: OR = 1.68, 95% CI: 0.53, 5.32, p = 0.38
With placebo: OR = 0.26, 95% CI: 0.13, 0.52, p < 0.001
SBD versus general population point-prevalence abstinence hazard ratios [presented by author as OR]
With varenicline: OR = 0.99; p = 0.897
With placebo: OR = 0.87; p = 0.011
Time to first relapse quartile 1 at 12 weeks[X2 test p<0.0001]
SBD on placebo: Q1 = 12 days
No SBD on placebo: Q1 = 17 days
SBD on varenicline: Q1 > 95 days
No SBD on varenicline: Q1 = 88
"Among those assigned to placebo, those with SBD were more likely to relapse and lapsed sooner than smokers without psychiatric illness…Among those on maintenance varenicline, the 6-month abstinence rates and time to first lapse was no different in those with SBD than for those without psychiatric illness."(pg 127)
Chen, 201323
Mean differences from baseline high vs low NRT (SD)

Daily number of cigarettes: -3.2 (7.1) vs -0.15 (7.5)

CO level: -0.1 (6.4) vs -0.6 (6.5)

FTND: -1.2 (2.3) vs -0.6 (2.4)

Positive and negative syndrome scale: -2.5 (11.0) vs -2.5 (11.0)

Simpson-Angus rating scale score: -0.02 (0.2) vs -0.02 (0.3)

7-day point prevalence abstinence: 1.1% (1/92) vs 4.3% (4/92)


Safety

5 discontinued due to side effects (unspecified)

"In summary, among a cohort of chronic institutionalized schizophrenic patients who took part in smoking cessation programs, smoking cessation and reduction outcomes were not correlated with NRT dose, and the cessation rate was much lower than those in similar studies."(pg 80)
Dennis, 201624
Active versus placebo nicotine patch

Pre-quit phase (2 weeks) nicotine craving: t(59) = -1.17, p = 0.25

Pre-quit phase (2 weeks) smoking: t(1340) = -0.74, p = 0.46

Pre-quit phase (2 weeks) PTSD symptoms: F(5,230) = 0.47, p=0.80

Post-quit phase time to relapse: HR=0.97, p=0.91

Abstinence 6-weeks: OR = 1.54 (95% CI 0.23,10.15)

Abstinence 6-months: n=26 (100%) reported smoking

"We found that supplemental nicotine patch-preloading did not lead to reductions in craving and smoking during the preloading phase, nor was it associated with reductions in smoking-associated relief from PTSD symptoms and negative affect."(pg 28)
Stockings, 201422
NRT+ vs treatment as usual
Fischer's exact tests for differences in continuous abstinence at 6 months:p = 0.26
Odds ratio for control (reference) vs intervention group at 6 months
Point prevalence abstinence: OR = 1.32 (95% CI 0.47,4.36)
Quit attempts: OR = 2.89 (95% CI 1.43, 5.98)
50% reduction in cigarettes per day: OR = 5.90 (95% CI 2.89, 15.25)
Mean difference between control and intervention group at 6 months
Cigarettes per day: -7.1 (95% CI -10.7 ,-3.5) p < .0001
Nicotine dependence (FTND): -1.6 (95% CI -2.3, -0.8) p < .0001
Psychological distress (K10): -0.7 (95% CI -3.7, 2.3) p=.642
"For smokers with a mental disorder, cessation support provided post-hospitalization was effective in reducing cigarette consumption and nicotine dependence, and encouraging quit attempts at 6 months."
Jeon, 201626
Varenicline vs placebo (baseline to week 8)
mNWS scores: p=0.391
QSU-brief: p=0.083
mCEQ: p=0.355 (time x group interaction p=0.002)
Expired CO: p=0.019 (time x group interaction p= 0.046)
Amount of cigarette: p=0.063 (time x group interaction p=0.007 )
PANSS total: p = 0.893
SANS: p = 0.170
HAM-D: p = 0.805
Safety

Three adverse events (3/60) including nausea (1/30 in each group), headache (1/30 in varenicline group) resulting in discontinuation, and two aggravated psychotic symptoms resulting in withdrawal between weeks 2 and 4

"Our results suggest that varenicline is effective for smoking reduction and is generally well-tolerated and safe in combination with antipsychotics for patients with schizophrenia" (pg 210)
Garcia-Portilla, 201628
Varenicline vs TNP mean differences (Week 12, 24, 36)

[Week 12] 7-day point prevalence abstinence: Mean difference = 1.4% (chisquare = 0.015, p=1.000)

[Week 12] >= 50% reduction in the number of cigarettes per day: Mean difference = 2.9%, chi-square= 0.100, p=0.776)

[Week 24] 7-day point prevalence abstinence: Mean difference = 7.9% (chisquare = 0.475, p = 0.639)

[Week 24] >= 50% reduction in the number of cigarettes per day: Mean difference = 1.8%, chi-square = 0.030, p = 1.000)

[Week 36] 7-day point prevalence abstinence: Mean difference = 16.4% (chi-square = 2.153, p=0.159)

[Week 36] >= 50% reduction in the number of cigarettes per day: Mean difference = 0.7%, chi-square = 0.005, p = 1.000)


Other effects

No observed group differences over time in breath CO level, FTND scores, GN-SBQ scores or proportion of mild, moderate, heavy smokers


Safety

Significant weight gain in both groups

Varenicline group had significantly lower cholesterol levels

21/36 (58.3%) in TNP and 27/39 (69.2%) in varenicline group experienced at least 1 AE, most commonly abnormal/vivid dreams (n=9 and n=4, respectively), constipation (n=5 and n=9, respectively), and nausea/vomiting in varenicline group only (n=12; p < 0.0005))

4/39 (10.2%) switched to TNP due to adverse event and 3/39 (7.8%) reduced their varenicline dose

"After 12 weeks of treatment with TNP or varenicline, combined with group therapy, a smoking cessation rate of 50% was achieved…As expected, this rate decreased with time, but 6 months after the end of the acute-treatment phase, 37% of patients in the trial remained abstinent. There were no differences in the dropout rates between the two drugs at any point in the study. Both pharmacological treatments were safe and generally well tolerated." (pg 276)
Stapleton, 200827
Total population including those without mental illness at 8 weeks

2-week abstinence prevalence NRT vs varenicline: OR = 1.70 (95% CI 1.09, 2.67) Mean Difference=10.8% (95% CI 1.8%, 19.9%)


Population with mental illness only at 8 weeks

2-week abstinence prevalence NRT vs varenicline: OR = 2.88 (95% CI 1.08, 7.63) Mean difference=16.5% (95% CI -0.01%, 34.2%)


Safety

Significantly higher nausea, disturbed sleep, vivid dreams, drowsiness, constipation, headache, dyspepsia, dry mouth, bad taste, low mood, diarrhoea and disorientation in varenicline group

7 patients switched from varenicline to NRT due to adverse symptoms (unspecified)

Adverse symptoms not found to be higher or more severe in those with mental illness

"The results suggest that, with routine psychological and behavioural group support, varenicline is more effective than NRT in aiding short-term smoking cessation…The results also indicate that varenicline is similarly effective in those with mental illness, supporting the regulatory decision to allow varenicline treatment in these patients" (pg 152)
Brody, 201725
7-day point prevalence abstinence at 6 months

TAU vs treatment (no home visits): X2(1) = 0.7, p = .4

TAU vs treatment (with home visits): X2(1) = 4.8, p = .03


Change in cigarettes per day at week 0 versus week 26

TAU vs treatment (no home visits): -7.5 vs -14 (p < 0.05)

TAU vs treatment (with home visits): -7.5 vs -16.1 (p < 0.05)


Change in exhaled carbon monoxide level at week 0 versus week 26

TAU vs treatment (no home visits): -0.8 vs -7.3 (p < 0.05)

TAU vs treatment (with home visits):-0.8 vs -7.0 (p < 0.05)


Change in FTND score at week 0 versus week 26

TAU vs treatment (no home visits): -2.2 vs -3.0

TAU vs treatment (with home vists): -2.2 vs -4.2 (p < 0.05)


Safety

No significant changes in safety measures from baseline scores in BPRS, SANS, CGI, BDI, C-SSRS, AIMS

27%, 30% and 46% reported adverse events in treatment with home visits, treatment without home visits and TAU groups respectively, most commonly insomnia (n = 4), with vivid dreams (n = 2), nausea (n = 2), rash (n = 2), agitation (n = 1)

"In conclusion, rapidly initiated combination and extended treatment improves smoking reduction/cessation outcomes compared to TAU in smokers with schizophrenia. In addition, [home visits] appear to be a promising adjunct to encourage smoking reduction and abstinence, and may be worthy of future research." (pg 74)
Castle, 201233
Baseline vs 6-months mean differences
Cigarettes per day: B = 13.61 (95% CI 6.58, 20.75), p= 0.001
Expired CO: B = 9.23 (95% CI -5.04, 23.51), p = 0.02
Dependence (FTND): B = 2.1 (95% CI 0.48, 3.62), p = 0.01
Withdrawal (other-rated): B = -0.32 (95% CI -0.63, -0.001), p = 0.05
Withdrawal (self-rated): B = 1.3 (95% CI 0.06, 0.61), p = 0.02
Safety

Most common side effects were sleep disturbance and nausea (exact number unclear)

3/14 patients discontinued due to psyciatric issues (n=1) or nausea (n=2)

No significant changes in MNWS, BPRS, BDI from baseline to follow up (BDI (pre: 9.2 [SD 7.0], post: 8.1 [SD 8.1]); YMRS (pre: 3.8 [SD 5.5], post 4.9 [SD 6.0]); or BPRS (pre: 35.6 [SD 5.0], post: 39.8 [SD 8.9]).

"This open study demonstrated that varenicline, in association with a comprehensive healthy lifestyle intervention, was associated with a substantial decrease in cigarette smoking among a heterogeneous group of patients with psychotic disorders. Abstinence was achieved in 42% of the participants at the 6-month mark. Side effects were mostly nonpsychiatric (ie, sleep disturbance, nausea) and transient; 1 patient with [bipolar disorder] dropped out because of a severe worsening of depression with suicidality." (pg 288)
Cather, 201734

74 participants (41.3%) attained 2+ weeks of continuous abstinence at the end of 12-week varenicline treatment period (mean = 42.7 +/- 18.6 days)


Other symptoms

CDSS scores decreased over time with varenicline treatment in the abstinent-achieving group (F(13, 816) = 6.22, p < .001) and non-abstinent completers (F(13, 841) = 2.48, p = .003_ but not , study dropouts (F(12, 1349) = 1.48, p = .125)

WSWS scores decreased over time significantly in those who attained abstinence F(13,820) = 1.76, p = .046, but not among non-abstinent completers (F(13, 850) = 1.29, p = .210), or those who dropped out (F(12, 1340) = 1.09, p = .368.)

"We conclude that smokers with schizophrenia and schizoaffective disorder who have significant depressive symptoms may be successful in smoking cessation attempts with varenicline while maintaining stable psychiatric symptoms." (pg 8)
Raich, 201631
Abstinence at week 12

All remaining subjects (n=37 (41.1%)) were abstinent at Week 12


Safety

53/90 discontinued; 4/53 discontinued due to adverse events (unspecified)

Most common adverse events were dry mouth (n=26; 28.9%), flatulence (n=25; 27.8%), abnormal dreams (n=25; 27.8%) and nausea (n=20; 22.2%)

2 patients with 'moderate suicidal ideation' during weeks 2 and 6, one of whom discontinued

"The present study shows smoking cessation with varenicline presents an acceptable safety level in patients with psychiatric disorders (psychotic disorder, alcohol dependence, and opioid dependence). Gastrointestinal adverse events are the most prevalent, although treatment dropout rates with varenicline are very low" (pg 652)
Pachas, 201232
Mean change per week from baseline to 12 weeks

CO: B =-0.03 (SE = 0.01), p < 0.01

7-day abstinence: B= 0.34 (SE = 0.03), p < 0.01

WSWS: B=-0.65 (SE =0.15), p < 0.01

Urge to Smoke: B= -0.29 (SE =0.04), p < 0.01

Calgary Depression Scale: B = -0.14 (SE =0.01), p < 0.01

BPRS -Psychosis: B= -1.34 (t=2.815, p< 0.01)

SANS Total: B= -1.0 (t=-0.914)


Safety

Most frequent adverse event was transient nausea

n=3/110 psychiatric hospitalizations (with 1 for paranoia and suicidal ideation)

n=12/110 discontinued study (nausea (5), anxiety (2), weight gain (1), depressed mood (1), paranoia (1), suicidal ideation (1), and substance use (1))

Significant weight gain was observed on average (B=202.59 (SD=44.35) at baseline to B=207.6 (SD=45.4) pounds at 12 weeks)

"Over 12 weeks, participants demonstrated increased abstinence rates, and decreased withdrawal symptoms, depressive symptoms and psychosis. The most common AE was transient nausea"(pg 6)
Loreto, 201730
Among patients with mental disorder
OR for abstinence compared to nicotine patch only [95% CI]:
Nicotine patch plus bupropion 2.00 [1.14, 3.50]
Nicotine patch plus gum 2.10 [1.04, 4.23]
Nicotine patch plus nortriptyline 2.07 [0.53, 8.08]
HR for treatment retention compared to nicotine patch only [95% CI]:
Nicotine patch plus bupropion 0.87 [0.60, 1.26]
Nicotine patch plus gum 0.70 [0.42, 1.14]
Nicotine patch plus nortriptyline 0.68 [0.27, 1.71]
Among patients without mental disorder
OR for abstinence compared to nicotine patch only [95% CI]:
Nicotine patch plus bupropion 1.51 [0.97, 2.35]
Nicotine patch plus gum 1.17 [0.62, 2.21]
Nicotine patch plus nortriptyline 1.96 [0.65, 5.96]
HR for treatment retention compared to nicotine patch only [95% CI]:
Nicotine patch plus bupropion 0.77 [0.57, 1.05]
Nicotine patch plus gum 0.96 [0.62, 1.47]
Nicotine patch plus nortriptyline 0.79 [0.36, 1.72]
"The use of CBT plus combined pharmacotherapy (NRT patch plus gum or bupropion) could be a powerful smoking cessation intervention in patients with MD, more so than in patients without MD." (pg 7)
Wu, 201729
Hazard ratio for suicide attempts/behaviours (relative to Varenicline group)
NRT: HR = 0.81 (95% CI 0.51, 1.28)
Bupropion: HR = 0.37 (95% CI 0.05, 2.70)
"Our study was the first to examine suicide behaviors or attempts among this minority population and we did not find any differences between the medications." (pg 67)

95% CI = 95% confidence interval; OR = odds ratio; RR = risk ratio; HR = hazard ratio, AE = adverse event, SMI = severe mental illness; SBD = schizophrenia or biopolar disorder; TAU = treatment as usual; NRT = nicotine replacement therapy; TNP = transdermal nicotine patch; CO = carbon monoxide; ppm = parts per million; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders; CBT = cognitive behavioural therapy; FTND = Fagerstrom Test for Nicotine Dependence; mNWS = Minnesota Nicotine Withdrawal Scale; WSWS = Wisconsin Smoking Withdrawal Scale; mCEQ = Modified Cigarette Evaluation Questionnaire; BPRS = Brief Psychiatric Rating Scale; SARS = Simpson-Angus Rating Scale; PANSS = Positive and Negative Syndrome Scale; HAM-D = Hamilton Rating Scale for Depression; CDSS = Calgary Depression Scale for Schizophrenia; BDI = Beck Depression Inventory; C-SSRS = Columbia- Suicide Severity Rating Scale; AIMS = Abnormal Involuntary Movement Scale; K10 = Kessler Psychological Distress Scale; CGI = Clinical Global Impression; QSU = Questionnaire of Smoking Urges

Appendix 5. Additional References of Potential Interest

About the Series

CADTH Rapid Response Report: Summary with Critical Appraisal
ISSN: 1922-8147

Version: 1.0

Funding: CADTH receives funding from Canada’s federal, provincial, and territorial governments, with the exception of Quebec.

Suggested citation:

Smoking cessation interventions for patients with severe mental illnesses: a review of clinical effectiveness and guidelines. Ottawa: CADTH; 2017 Aug (CADTH rapid response report: summary with critical appraisal).

Disclaimer: The information in this document is intended to help Canadian health care decision-makers, health care professionals, health systems leaders, and policy-makers make well-informed decisions and thereby improve the quality of health care services. While patients and others may access this document, the document is made available for informational purposes only and no representations or warranties are made with respect to its fitness for any particular purpose. The information in this document should not be used as a substitute for professional medical advice or as a substitute for the application of clinical judgment in respect of the care of a particular patient or other professional judgment in any decision-making process. The Canadian Agency for Drugs and Technologies in Health (CADTH) does not endorse any information, drugs, therapies, treatments, products, processes, or services.

While care has been taken to ensure that the information prepared by CADTH in this document is accurate, complete, and up-to-date as at the applicable date the material was first published by CADTH, CADTH does not make any guarantees to that effect. CADTH does not guarantee and is not responsible for the quality, currency, propriety, accuracy, or reasonableness of any statements, information, or conclusions contained in any third-party materials used in preparing this document. The views and opinions of third parties published in this document do not necessarily state or reflect those of CADTH.

CADTH is not responsible for any errors, omissions, injury, loss, or damage arising from or relating to the use (or misuse) of any information, statements, or conclusions contained in or implied by the contents of this document or any of the source materials.

This document may contain links to third-party websites. CADTH does not have control over the content of such sites. Use of third-party sites is governed by the third-party website owners’ own terms and conditions set out for such sites. CADTH does not make any guarantee with respect to any information contained on such third-party sites and CADTH is not responsible for any injury, loss, or damage suffered as a result of using such third-party sites. CADTH has no responsibility for the collection, use, and disclosure of personal information by third-party sites.

Subject to the aforementioned limitations, the views expressed herein are those of CADTH and do not necessarily represent the views of Canada’s federal, provincial, or territorial governments or any third party supplier of information.

This document is prepared and intended for use in the context of the Canadian health care system. The use of this document outside of Canada is done so at the user’s own risk.

This disclaimer and any questions or matters of any nature arising from or relating to the content or use (or misuse) of this document will be governed by and interpreted in accordance with the laws of the Province of Ontario and the laws of Canada applicable therein, and all proceedings shall be subject to the exclusive jurisdiction of the courts of the Province of Ontario, Canada.

Copyright © 2017 Canadian Agency for Drugs and Technologies in Health.

The copyright and other intellectual property rights in this document are owned by CADTH and its licensors. These rights are protected by the Canadian Copyright Act and other national and international laws and agreements. Users are permitted to make copies of this document for non-commercial purposes only, provided it is not modified when reproduced and appropriate credit is given to CADTH and its licensors.

Except where otherwise noted, this work is distributed under the terms of a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International licence (CC BY-NC-ND), a copy of which is available at http://creativecommons.org/licenses/by-nc-nd/4.0/

Bookshelf ID: NBK525600PMID: 30222276

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (876K)

Other titles in this collection

Related information

Similar articles in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...