Systematic Review and Meta-Analysis to Assess the Safety of Bupropion and Varenicline in Pregnancy

Systematic Review and Meta-Analysis to Assess the Safety of Bupropion and Varenicline in Pregnancy Abstract Introduction Smoking in pregnancy is a substantial public health issue, but, apart from nicotine replacement therapy (NRT), pharmacological therapies are not generally used to promote cessation. Bupropion and varenicline are effective cessation methods in nonpregnant smokers and this systematic review investigates their safety in pregnancy. Methods We searched MEDLINE, EMBASE, CINAHL, and PsychINFO databases for studies of any design reporting pregnancy outcomes after bupropion or varenicline exposure. We included studies of bupropion used for smoking cessation, depression, or where the indication was unspecified. Depending on study design, quality was assessed using the Newcastle-Ottawa Scale or Cochrane Risk of Bias Tool. Most findings are reported narratively but meta-analyses were used to produce pooled estimates for the proportion of live births with congenital malformations and of the mean birthweight and gestational age at delivery following bupropion exposure. Results In total, 18 studies were included: 2 randomized controlled trials, 11 cohorts, 2 case– control studies, and 3 case reports. Study quality was variable. Gestational safety outcomes were reported in 14 bupropion and 4 varenicline studies. Meaningful meta-analysis was only possible for bupropion exposure, for which the pooled estimated proportion of congenital malformations amongst live-born infants was 1.0% (95% CI = 0.0%–3.0%, I2 = 80.9%, 4 studies) and the mean birthweight and mean gestational age at delivery was 3305.9 g (95% CI = 3173.2–3438.7 g, I2 = 77.6%, 5 studies) and 39.2 weeks (95% CI = 38.8–39.6 weeks, I2 = 69.9%, 5 studies), respectively. Conclusions There was no strong evidence that either major positive or negative outcomes were associated with gestational use of bupropion or varenicline. PROSPERO registration number CRD42017067064. Implications We believe this to be the first systematic review investigating the safety of bupropion and varenicline in pregnancy. Meta-analysis of outcomes following bupropion exposure in pregnancy suggests that there are no major positive or negative impacts on the rate of congenital abnormalities, birthweight, or premature birth. Overall, we found no evidence that either of these treatments might be harmful in pregnancy, and no strong evidence to suggest safety, but available evidence is of poor quality. Introduction Smoking in pregnancy is associated with increased risks of miscarriage, stillbirth, prematurity, low birth weight, perinatal morbidity, and mortality1 and is a significant problem in developed countries where rates vary between 8% and 23% of pregnant women smoking in pregnancy.2–4 Children of smoking mothers are twice as likely to become smokers themselves,5 so smoking in pregnancy and afterwards encourages the persistence of smoking across generations.6 Smoking in pregnancy is declining in developed countries but remains highest amongst younger, socially disadvantaged women4 and the annual costs of managing the smoking-attributable maternal and infant disease can be substantial.7 Studies have shown that pregnancy is the life event which most motivates smokers to attempt cessation, with around half of pregnant smokers attempting to quit.4 In addition, although the cost-efficacy of smoking cessation interventions in pregnancy is unclear,8 stopping smoking in pregnancy is likely to save healthcare resources, both with respect to the health of the infant and in the wider context of preventing the perseveration of smoking in the next generation. Thus, promoting smoking cessation during pregnancy will substantially improve the health not only of the infant and mother but of their extended family and, in the longer term, will contribute to reducing the substantial healthcare cost of smoking-related diseases. However, compared to those available for nonpregnant smokers, relatively few effective cessation interventions can be used in pregnancy and nicotine replacement therapy (NRT) is the only drug treatment used to any extent.9 The UK National Institute for Healthcare and Excellence (NICE) recommends using NRT, believing it safer that continued smoking in pregnancy. However, in pregnancy NRT has at best only a borderline significant effect on cessation (RR 1.28, 95% CI 0.99–1.66)10 and this lower efficacy, compared to use outside of pregnancy, is probably caused by poor adherence to NRT.11 If they were considered sufficiently safe, other effective cessation pharmacotherapies, varenicline and bupropion, could also be tried in pregnancy. Varenicline is well tolerated12 and probably more effective than other cessation treatments;13 animal research suggests it is not teratogenic.14 Similarly, bupropion is an effective smoking cessation aid which approximately doubles nonpregnant smokers chances of stopping.15 If varenicline or bupropion were to be proven effective for pregnant smokers, the health benefits which would accrue from stopping smoking would very likely outweigh any minor adverse effects. Consequently, to help assesses whether experimental studies might be ethical, we review evidence for the safety of varenicline and bupropion in pregnancy. Methods A study protocol was registered16 and the review adhered to Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines.17 Inclusion Criteria We included studies of any design that reported adverse pregnancy outcomes experienced by mothers, fetuses, or infants following use of varenicline or bupropion in pregnancy. Exclusion Criteria We excluded studies that presented no empirical data and those in which interventions combined bupropion or varenicline with other cessation pharmacotherapies. Search Strategy We searched the MEDLINE, EMBASE, CINAHL, and PsychINFO databases and hand-searched reference lists from reviews and included papers. We also searched for ongoing and unpublished studies at: www.gsk-clinicalstudyregister.com; clincialtrials.gov; www.who.int/trialsearch; www.controlled-trials.com/isrctn; and www.ukctg.nihr.ac.uk. As bupropion and varenicline were licensed and became available relatively recently, we sought studies published from 1990 until May 25, 2017 with no language restrictions. Search terms relating to pregnancy were developed from those used in a Cochrane review10 and were combined with qualitative terms relating to smoking, varenicline or bupropion. The protocol also states that we intended to include studies in which women used “dual” nicotine replacement therapy (NRT);16 however, only two such studies were identified, and therefore we retrospectively decided that they should not be reviewed separately from other NRT studies (eg, those investigating “mono” NRT). Data Extraction Titles and abstracts were screened by the lead reviewer, retrieving complete manuscripts if necessary to decide on inclusion. All articles were independently assessed by two reviewers to confirm inclusion in the review, with adjudication via a third reviewer when agreement was not met. The following data were extracted by the lead reviewer and checked by a second reviewer, with any discrepancies resolved by a third reviewer: aims and design, numbers of participants, outcomes, data collection, analysis methods, and findings. Where studies reported interim analyses, further details were requested from authors. Quality Assessment Quality assessment was performed using Newcastle–Ottawa Scales18 for cohort or cross-sectional studies and the Cochrane Risk of Bias Assessment for RCTs.19 Initial assessment was made by the lead reviewer and checked by a second reviewer, and discrepancies resolved by a third reviewer. Where a trial had already been quality-assessed for a Cochrane review, we used this published assessment. Data Synthesis A priori, we anticipated that review studies might be so diverse that meaningful data synthesis could be challenging. Hence, we planned making final decisions on whether or not meta-analyses were possible once data extraction was finished, and outcomes of this deliberation are reported alongside review findings. If performed, we anticipated that meta-analyses would be conducted in Stata version 1420 using a random effects DerSimonian and Laird model to generate pooled means and 95% confidence intervals with heterogeneity quantified by the I2 statistic.21 Rather than not pool studies in the presence of a high I2 value, we planned to present this statistic alongside meta-analysis findings to inform the reader of the extent to which pooled estimates should be treated cautiously. Results We identified 772 studies (1053 including duplicates); no ongoing trials were identified from registries and no completed, unpublished studies were identified from pharmaceutical company databases. We identified 30 articles for retrieval in full, with 18 being included in the review (Figure 1); study details and outcomes are shown in Supplementary Material Table 1. Figure 1. View largeDownload slide PRISMA flowchart of included studies. Figure 1. View largeDownload slide PRISMA flowchart of included studies. Study Design, Outcome Measures, and Quality Assessment Included studies comprised two randomized controlled trials (RCTs),22,23 eleven cohort studies,24–34 two case–control studies,35,36 and three case reports.37–39 Maternal or fetal adverse outcomes were reported in fourteen bupropion22–31,36–38 and four varenicline studies.32–34,39 Although bupropion can be used as an anti-depressant or for smoking cessation, only two RCTs specified that bupropion had been prescribed for smoking cessation.22,23 The three observational studies evaluating varenicline exposure did not explicitly state that varenicline was used for smoking cessation, but all discussed its sole indication as a cessation pharmacotherapy.32–34 Congenital malformations were reported in eight studies (six following bupropion27–30,35,36 and two following varenicline33,34); reported malformation classification systems are described in Supplementary Material Table 1. Birthweight and gestational age at delivery were reported in five bupropion studies.22–24,26,27 Other outcomes included: fetal loss or stillbirth25,27,30,32,34; fetal length or head circumference22,23; preterm birth22–24; maternal medication adverse effects23,25,39; and pre-eclampsia.31 An overview of reported outcomes is shown in Table 1. Table 1. Overview of Reported Outcomes by Study   Fetal outcomes  Maternal outcomes  Paper  Drug of interest  Congenital malformations (incl CV)  Birth weight  GA at delivery  Pregnancy outcome  Length/Head circumference  Preterm birth  Other*  Pregnancy/Maternal adverse effects  Cotinine levels  Pre-eclampsia  Alwan  2010  Bupropion  ✓                    Berard  2016  Bupropion    ✓  ✓      ✓  ✓        Boshier  2003  Bupropion        ✓        ✓      Chan  2005  Bupropion    ✓  ✓                Chun-Fai-Chan  2005  Bupropion  ✓  ✓  ✓  ✓              Cole  2007  Bupropion  ✓                    Einarson  2009  Bupropion  ✓                    Gisslen  2011  Bupropion              ✓        GSK  2008  Bupropion  ✓      ✓              Leventhal  2010  Bupropion              ✓        Louik  2014  Bupropion  ✓                    Nanovskaya  2017  Bupropion    ✓  ✓    ✓  ✓  ✓    ✓    Palmsten  2013  Bupropion                    ✓  Stotts  2015  Bupropion    ✓  ✓    ✓    ✓  ✓  ✓    Harrison- Woolrych  2013  Varenicline        ✓      ✓        Kaplan  2014  Varenicline                ✓      Olsen  2015  Varenicline  ✓                    Richardson  2017  Varenicline  ✓      ✓                Fetal outcomes  Maternal outcomes  Paper  Drug of interest  Congenital malformations (incl CV)  Birth weight  GA at delivery  Pregnancy outcome  Length/Head circumference  Preterm birth  Other*  Pregnancy/Maternal adverse effects  Cotinine levels  Pre-eclampsia  Alwan  2010  Bupropion  ✓                    Berard  2016  Bupropion    ✓  ✓      ✓  ✓        Boshier  2003  Bupropion        ✓        ✓      Chan  2005  Bupropion    ✓  ✓                Chun-Fai-Chan  2005  Bupropion  ✓  ✓  ✓  ✓              Cole  2007  Bupropion  ✓                    Einarson  2009  Bupropion  ✓                    Gisslen  2011  Bupropion              ✓        GSK  2008  Bupropion  ✓      ✓              Leventhal  2010  Bupropion              ✓        Louik  2014  Bupropion  ✓                    Nanovskaya  2017  Bupropion    ✓  ✓    ✓  ✓  ✓    ✓    Palmsten  2013  Bupropion                    ✓  Stotts  2015  Bupropion    ✓  ✓    ✓    ✓  ✓  ✓    Harrison- Woolrych  2013  Varenicline        ✓      ✓        Kaplan  2014  Varenicline                ✓      Olsen  2015  Varenicline  ✓                    Richardson  2017  Varenicline  ✓      ✓              View Large Table 1. Overview of Reported Outcomes by Study   Fetal outcomes  Maternal outcomes  Paper  Drug of interest  Congenital malformations (incl CV)  Birth weight  GA at delivery  Pregnancy outcome  Length/Head circumference  Preterm birth  Other*  Pregnancy/Maternal adverse effects  Cotinine levels  Pre-eclampsia  Alwan  2010  Bupropion  ✓                    Berard  2016  Bupropion    ✓  ✓      ✓  ✓        Boshier  2003  Bupropion        ✓        ✓      Chan  2005  Bupropion    ✓  ✓                Chun-Fai-Chan  2005  Bupropion  ✓  ✓  ✓  ✓              Cole  2007  Bupropion  ✓                    Einarson  2009  Bupropion  ✓                    Gisslen  2011  Bupropion              ✓        GSK  2008  Bupropion  ✓      ✓              Leventhal  2010  Bupropion              ✓        Louik  2014  Bupropion  ✓                    Nanovskaya  2017  Bupropion    ✓  ✓    ✓  ✓  ✓    ✓    Palmsten  2013  Bupropion                    ✓  Stotts  2015  Bupropion    ✓  ✓    ✓    ✓  ✓  ✓    Harrison- Woolrych  2013  Varenicline        ✓      ✓        Kaplan  2014  Varenicline                ✓      Olsen  2015  Varenicline  ✓                    Richardson  2017  Varenicline  ✓      ✓                Fetal outcomes  Maternal outcomes  Paper  Drug of interest  Congenital malformations (incl CV)  Birth weight  GA at delivery  Pregnancy outcome  Length/Head circumference  Preterm birth  Other*  Pregnancy/Maternal adverse effects  Cotinine levels  Pre-eclampsia  Alwan  2010  Bupropion  ✓                    Berard  2016  Bupropion    ✓  ✓      ✓  ✓        Boshier  2003  Bupropion        ✓        ✓      Chan  2005  Bupropion    ✓  ✓                Chun-Fai-Chan  2005  Bupropion  ✓  ✓  ✓  ✓              Cole  2007  Bupropion  ✓                    Einarson  2009  Bupropion  ✓                    Gisslen  2011  Bupropion              ✓        GSK  2008  Bupropion  ✓      ✓              Leventhal  2010  Bupropion              ✓        Louik  2014  Bupropion  ✓                    Nanovskaya  2017  Bupropion    ✓  ✓    ✓  ✓  ✓    ✓    Palmsten  2013  Bupropion                    ✓  Stotts  2015  Bupropion    ✓  ✓    ✓    ✓  ✓  ✓    Harrison- Woolrych  2013  Varenicline        ✓      ✓        Kaplan  2014  Varenicline                ✓      Olsen  2015  Varenicline  ✓                    Richardson  2017  Varenicline  ✓      ✓              View Large Study quality was variable (Table 2) with seven of the eleven observational studies assessed as of low methodological quality (score of <7);25,26,29,30,32–34 the three case reports were considered to provide only low-quality evidence.37–39 Table 2. Quality Assessment of Included Studies Cohort studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Outcome (max 3*)  Total (max 9*)  Berard, 2016  ☆☆☆  ☆☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Palmsten, 2013  ☆☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Chun-Fai-Chan, 2005  ☆☆☆  ☆☆  ☆☆  ☆☆☆☆☆☆☆  Cole, 2007  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Einarson, 2009  ☆☆☆  ☆  ☆☆  ☆☆☆☆☆☆  Boshier, 2004  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Chan, 2005  ☆☆  ☆  ☆☆  ☆☆☆☆☆  Harrison-Woolrych, 2013  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Olsen, 2015  ☆☆☆  0  ☆☆  ☆☆☆☆☆  GlaxoSmithKleine, 2008  ☆  0  ☆☆  ☆☆☆  Richardson, 2017  ☆☆  0  ☆  ☆☆☆  Case–control studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Exposure (max 3*)  Total (max 9*)  Alwan, 2010  ☆☆☆☆  ☆☆  ☆  ☆☆☆☆☆☆☆  Louik, 2014  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Randomized controlled trials—Cochrane Risk of Bias    Random sequence generation  Allocation concealment  Blinding of participants and personnel  Blinding of outcome assessment  Incomplete outcome data  Selective outcome reporting  Nanovskaya, 2017  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Stotts, 2015*  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Case reports  Gisslen, 2011  High risk by design, unblinded assessments  Kaplan, 2014  High risk by design, unblinded assessments  Leventhal, 2010  High risk by design, unblinded assessments  Cohort studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Outcome (max 3*)  Total (max 9*)  Berard, 2016  ☆☆☆  ☆☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Palmsten, 2013  ☆☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Chun-Fai-Chan, 2005  ☆☆☆  ☆☆  ☆☆  ☆☆☆☆☆☆☆  Cole, 2007  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Einarson, 2009  ☆☆☆  ☆  ☆☆  ☆☆☆☆☆☆  Boshier, 2004  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Chan, 2005  ☆☆  ☆  ☆☆  ☆☆☆☆☆  Harrison-Woolrych, 2013  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Olsen, 2015  ☆☆☆  0  ☆☆  ☆☆☆☆☆  GlaxoSmithKleine, 2008  ☆  0  ☆☆  ☆☆☆  Richardson, 2017  ☆☆  0  ☆  ☆☆☆  Case–control studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Exposure (max 3*)  Total (max 9*)  Alwan, 2010  ☆☆☆☆  ☆☆  ☆  ☆☆☆☆☆☆☆  Louik, 2014  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Randomized controlled trials—Cochrane Risk of Bias    Random sequence generation  Allocation concealment  Blinding of participants and personnel  Blinding of outcome assessment  Incomplete outcome data  Selective outcome reporting  Nanovskaya, 2017  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Stotts, 2015*  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Case reports  Gisslen, 2011  High risk by design, unblinded assessments  Kaplan, 2014  High risk by design, unblinded assessments  Leventhal, 2010  High risk by design, unblinded assessments  *Quality assessment for Stotts 2015 as assessed in the Cochrane Review “Pharmacological interventions for promoting smoking cessation during pregnancy” (2015)10 View Large Table 2. Quality Assessment of Included Studies Cohort studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Outcome (max 3*)  Total (max 9*)  Berard, 2016  ☆☆☆  ☆☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Palmsten, 2013  ☆☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Chun-Fai-Chan, 2005  ☆☆☆  ☆☆  ☆☆  ☆☆☆☆☆☆☆  Cole, 2007  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Einarson, 2009  ☆☆☆  ☆  ☆☆  ☆☆☆☆☆☆  Boshier, 2004  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Chan, 2005  ☆☆  ☆  ☆☆  ☆☆☆☆☆  Harrison-Woolrych, 2013  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Olsen, 2015  ☆☆☆  0  ☆☆  ☆☆☆☆☆  GlaxoSmithKleine, 2008  ☆  0  ☆☆  ☆☆☆  Richardson, 2017  ☆☆  0  ☆  ☆☆☆  Case–control studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Exposure (max 3*)  Total (max 9*)  Alwan, 2010  ☆☆☆☆  ☆☆  ☆  ☆☆☆☆☆☆☆  Louik, 2014  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Randomized controlled trials—Cochrane Risk of Bias    Random sequence generation  Allocation concealment  Blinding of participants and personnel  Blinding of outcome assessment  Incomplete outcome data  Selective outcome reporting  Nanovskaya, 2017  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Stotts, 2015*  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Case reports  Gisslen, 2011  High risk by design, unblinded assessments  Kaplan, 2014  High risk by design, unblinded assessments  Leventhal, 2010  High risk by design, unblinded assessments  Cohort studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Outcome (max 3*)  Total (max 9*)  Berard, 2016  ☆☆☆  ☆☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Palmsten, 2013  ☆☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Chun-Fai-Chan, 2005  ☆☆☆  ☆☆  ☆☆  ☆☆☆☆☆☆☆  Cole, 2007  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Einarson, 2009  ☆☆☆  ☆  ☆☆  ☆☆☆☆☆☆  Boshier, 2004  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Chan, 2005  ☆☆  ☆  ☆☆  ☆☆☆☆☆  Harrison-Woolrych, 2013  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Olsen, 2015  ☆☆☆  0  ☆☆  ☆☆☆☆☆  GlaxoSmithKleine, 2008  ☆  0  ☆☆  ☆☆☆  Richardson, 2017  ☆☆  0  ☆  ☆☆☆  Case–control studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Exposure (max 3*)  Total (max 9*)  Alwan, 2010  ☆☆☆☆  ☆☆  ☆  ☆☆☆☆☆☆☆  Louik, 2014  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Randomized controlled trials—Cochrane Risk of Bias    Random sequence generation  Allocation concealment  Blinding of participants and personnel  Blinding of outcome assessment  Incomplete outcome data  Selective outcome reporting  Nanovskaya, 2017  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Stotts, 2015*  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Case reports  Gisslen, 2011  High risk by design, unblinded assessments  Kaplan, 2014  High risk by design, unblinded assessments  Leventhal, 2010  High risk by design, unblinded assessments  *Quality assessment for Stotts 2015 as assessed in the Cochrane Review “Pharmacological interventions for promoting smoking cessation during pregnancy” (2015)10 View Large Potential for Meta-Analyses From the distribution of outcomes across studies and study designs (Table 1), it was clear that the few studies investigating varenicline were so different in design and outcome measurement that meaningful meta-analyses were not possible. For bupropion studies, meta-analyses investigating effects on congenital malformation, mean birthweight, and mean gestational age at birth were considered feasible and were undertaken by combining studies or study arms with sufficiently similar designs. For the meta-analysis investigating effects of bupropion exposure on congenital malformations, we included only cohort studies.27–30 For birthweight and gestational age at birth meta-analyses, we pooled data from bupropion-exposed arms in cohort studies24,26,27 and RCTs22,23 to determine the mean value associated with each outcome. Bupropion Congenital Malformations Six studies, four cohort27–30 and two case control studies,35,36 reported congenital malformations. Cohort studies included 3376 pregnancies (Figure 2) and from these studies the pooled estimate for the percentage of congenital malformations amongst live-born infants exposed to bupropion at any point during gestation was 1.0% (95% CI = 0.0%–3.0%, I2 = 80.9%) (Figure 3a). As individual studies classified congenital malformations in different ways (Supplementary Material Table 1), we accepted the presence or absence of malformations was as defined within each study and no attempt was made to derive a single classification system applied to all studies. Pregnancies which ended in stillbirth, miscarriage, intra-uterine fetal death or termination were excluded from the analysis, so we defined the proportion of pregnancies with congenital malformations as follows:  Proportion with any congenital malformation=number of live born infants with a malformationtotal number of live born infants exposed in utero Figure 2. View largeDownload slide Overview of pregnancies included in meta-analysis of congenital malformations following bupropion exposure. Figure 2. View largeDownload slide Overview of pregnancies included in meta-analysis of congenital malformations following bupropion exposure. Figure 3. View largeDownload slide (A) Proportion of congenital malformations following bupropion exposure. (B) Mean birthweight following bupropion exposure. (C) Mean gestational age at delivery following bupropion exposure. Figure 3. View largeDownload slide (A) Proportion of congenital malformations following bupropion exposure. (B) Mean birthweight following bupropion exposure. (C) Mean gestational age at delivery following bupropion exposure. The two case–control studies had conflicting results.35,36 Both used National Birth Defects Prevention Study (NBDPS) criteria40 to classify congenital cardiac defects; overall, Alwan found no evidence that maternal bupropion exposure in pregnancy increased infants’ risks of developing congenital cardiac defects (adjusted odds ratio 1.4, 95% CI = 0.8–2.5).35 Alwan did however, report an increased risk of left outflow tract cardiac defects (adjusted odds ratio 2.6, 95% CI = 1.2–5.7) which was not found by Louik36 (adjusted odds ratio 0.4, 95% CI = 0–2.4). Louik investigated the risk of developing eight different cardiac defects following bupropion exposure but did not attempt to estimate the overall risk of any cardiac defect and reported an increased risk of ventricular septal defects (adjusted odds ratio 2.9, 95% CI = 1.5–5.5) which was not found by Alwan (adjusted odds ratio 1.2, 95% CI = 0.5–3.4). Louik found no increased risks for other sub-categories of cardiac defects following bupropion exposure. Birthweight Two RCTs (combined n = 35) found no significant differences in birthweight between bupropion or placebo groups (Supplementary Material Table 1) but both will likely have been under-powered to detect clinically significant differences.22,23 One controlled cohort study conducted in smokers found significantly higher mean birthweights amongst infants born after exposure to bupropion (3315.9 g, SD = 553.3, n = 72) compared to those who smoked and used no treatment (2943.5 g, SD = 733.5, n = 900, p < .05).24 However, this finding was not replicated in two other bupropion cohort studies.26,27 Meta-analysis of the 262 pregnancies in bupropion-exposed arms from cohorts and RCTs gives a pooled estimate for mean birthweight amongst infants exposed to bupropion of 3305.9 g (95% CI = 3173.2–3438.7 g, I2 = 77.6%, n = 262) (Figure 3b). Gestational Age at Delivery Of five studies reporting gestational age at birth, two were RCTs22,23and three cohort studies.24,26,27 No significant differences were found between trial groups within the likely under-powered RCTs or between exposure groups in two of the cohort studies.26,27 However, one cohort study found that infants born after bupropion exposure had a significantly higher mean gestational age at birth (39.1 weeks, SD = 1.3, n = 72) compared to non-exposed infants born to smokers (37.5 weeks, SD = 3.3, n = 900, p < .05).24 The pooled estimate for mean gestational age at delivery in the five studies which included 260 pregnancies was 39.2 weeks (95% CI = 38.8–39.6, I2 = 69.9%) (Figure 3c). Foetal Loss Three cohort studies reported fetal loss following bupropion exposure,25,27,30 with only one study including control group data.27 The GlaxoSmithKline “Bupropion Pregnancy Registry” cohort reported data from 994 prospectively registered pregnant women (featuring 1005 monitored fetuses) following gestational bupropion exposure. Following first trimester bupropion exposure, there were 669 live births, 3 fetal deaths occurring at or later than 20 weeks gestation, 38 induced abortions, and 96 spontaneous pregnancy losses occurring before 20 weeks. Following second trimester exposure, there were 145 live births, 1 induced abortion and 1 spontaneous pregnancy loss and after bupropion exposure in the third trimester there were 51 live births and 1 fetal death. In prospectively registered pregnancies, 603 were either lost to follow-up or pending delivery when the register closed, resulting in a loss of outcome data.30 One study found significantly higher rates of spontaneous abortions (p = .009) and therapeutic abortions (p = .015) in a bupropion cohort compared with those exposed to “non-teratogenic” agents.27 A small uncontrolled, cohort study of 12 pregnancies, reported 5 live births, 2 therapeutic terminations (no explanation of reasons for termination given), 1 intra-uterine death, and 4 cases that were lost to follow-up following gestational bupropion exposure.25 Varenicline Four varenicline studies reported relevant adverse outcomes; three cohort studies32–34 and one case report study.39 No study explicitly stated that such exposures were unintentional, though this was probably the case as there were no smoking cessation studies and varenicline has no therapeutic indications in pregnancy. Richardson reported outcomes and congenital malformations following exposure to varenicline in pregnancy (n = 89) in a study which compared pregnant women exposed to non-teratogenic agents (n = 267) with those exposed to either NRT or bupropion (combined group, n = 267). As determined by the EUROCAT classification system for congenital malformations,41 seven infants (7.87%) in the varenicline group were reported to have a congenital malformation; two of which were “major” and five “minor.” No significant between-group differences were found in malformation rates.34 Another cohort study reported malformation rates in infants both exposed (4.3%, n = 254) and not exposed to varenicline in utero and also in those exposed to maternal smoking during gestation (4.2%, n = 5296), and those exposed to neither varenicline nor smoking in utero (4.2%, n = 656139). Rates appeared similar but no statistical comparison of groups was undertaken.33 One uncontrolled cohort of 23 varenicline-exposed pregnancies reported 14 live full-term births (61%), two live preterm (<37 weeks gestation) births (9%), four terminations of pregnancy (17%), two spontaneous or missed abortions (9%), and one ectopic pregnancy (4%). Within this cohort five fetal adverse events were reported, as shown in Supplementary Material Table 1.32 One case report described a normal pregnancy, delivery, and infant health until six months following gestational varenicline exposure for 4 weeks from the last menstrual period.39 Discussion We believe that this is the first systematic review investigating the safety of bupropion and varenicline in pregnancy. We found no evidence that either of these treatments might be harmful in pregnancy but available evidence is of poor quality and there is also no strong evidence to suggest safety. Most studies investigated outcomes following bupropion exposure and pooled estimates for birthweight, gestation at birth, and congenital abnormality rates do not suggest that any of these outcomes are adversely affected. However, estimates’ confidence intervals were relatively wide and more data would be required to improve precision. Far fewer studies investigated outcomes following varenicline exposures and overall there is probably insufficient evidence to make firm conclusions about the safety on any of either therapy in pregnancy. This study has some limitations. Relatively few studies were eligible for inclusion reflecting a paucity of relevant data and the majority of those in the review were small and observational, with only two RCTs.22,23 This restricted the assessment of potential causal relationships. Included studies generally had low methodological quality; some are case reports37–39 or cohort studies which lack control groups.25,30,32 Relatively few studies reported similar outcomes, restricting the potential for meta-analysis and, where these were conducted, heterogeneity was high, presumably due to differences between study designs and comparator groups; this heterogeneity means that this study’s calculated pooled estimated need to be treated with caution. As there were so few studies investigating bupropion specifically for smoking cessation purposes,22,23 we combined those which stated explicitly that bupropion had been used for smoking cessation and also those where the purpose of bupropion use was unspecified and this could have been prescribed for either smoking cessation or depression. Consequently, some of the data in our meta-analyses will have come from nonsmokers who would be expected to have better birth outcomes than smokers. As the evidence regarding the safety of varenicline during pregnancy is sparser than that for bupropion, with safety data identified in only four studies,32–34,39 this review predominantly focuses on bupropion. Because of the low-quality designs (eg, uncontrolled32) and complex comparator groups (eg, pregnant women exposed to non-teratogenic agents or those using either NRT or bupropion as a single group), no conclusion as to the safety of varenicline can be made.34 A strength of this review is its novelty and systematic approach. By including studies with any design, it is likely we have identified the majority of available safety evidence. Additionally, the rigorous quality assessment indicates that further investigation of the safety of pharmacotherapy during pregnancy is required. We have aimed to maximize use of available data and believe we have made the best use of this whilst also being sensitive to the limitations inherent in empirical studies’ designs. Our pooled estimate for the proportion of congenital malformations in live-born infants following gestational bupropion exposure (1%) is similar to those reported in comparable populations. From 2011–2015 EUROCAT, a European network of population-based registries, reported a congenital abnormalities rate of 2.5% amongst live births, fetal deaths, stillbirths, and terminations for fetal abnormalities.42 In addition, the MACDP (Metropolitan Atlanta Congenital Defects Program) determined the rate in five central counties of Atlanta between 1968 and 2003, to be 2.67%.43 Included papers reported only abnormalities within live-born infants; however, population-based registries generally include pregnancies in which the fetus dies before term. As abnormalities are less likely to be present in live-born infants, the 1% review-derived rate may underestimate prevalence in all pregnancies and the interpretation of our data is not completely straightforward. Despite this, it is reassuring that the upper 95% CI for the estimate (3%) is close to population estimates; further studies would increase the precision of this estimate and possibly provide further reassurance that congenital abnormality rates are not higher after bupropion exposure. Although 95% confidence intervals are consistent with wide range of values, the meta-analysis derived point estimate for mean birthweight following bupropion exposure 3305.9 g (95% CI = 3173.2–3438.7 g) was similar to the population average of the countries in which the studies reporting this outcome were conducted. Studies included in the birthweight meta-analysis were predominantly North American and only one was UK-based.27 Population-based data show that the average birthweight for those born between 37 and 41 weeks of gestation in the United States in 2005 was 3389 g (SD = 466),44 and in 2009 the mean birthweight of Canadian babies was 3364 g.45 Calculating the effects of bupropion on birthweight is also complicated by the known reduction in birthweight associated with maternal smoking; for example, one large study of 3338 mothers reported an adjusted birthweight deficit within babies born to active smokers averaging 226 g.46 Four of the studies contributing to the pooled estimate for birthweight following bupropion exposure included only pregnant smokers22–24,26 and the remaining study controlled for the effects of smoking by matching study groups by smoking status.27 None of the review studies reported a mean birthweight within the bupropion-exposed groups that was significantly less than their control groups22–24,26,27; in four studies, birthweights were higher in the bupropion cohorts,22,23,26,27 and in one study this finding was statistically significant.24 The latter study reported increasingly heavier birthweights between pregnant smokers who used no cessation pharmacotherapy, who used a nicotine patch, and who used bupropion, with rates of smoking cessation during pregnancy of 0%, 79%, and 81%, respectively. The high rates of smoking cessation in the bupropion-exposed cohort in this study may have been the driving factor behind the higher birthweight within the group, rather than being associated with bupropion pharmacotherapy itself, but this nonetheless is a beneficial outcome. We calculated the pooled mean gestational age at delivery following bupropion exposure as 39.2 weeks (95% CI = 38.8–39.6), as shown in Figure 3. This is comparable to the normal 40-week gestation and clinically insignificant. When assessing the studies that compared bupropion-exposed infants to pregnant smokers not using bupropion, there was also no evidence of a significant negative effect. One study found that the mean gestational age at birth for infants born to pregnant smokers using bupropion was significantly later than that of pregnant smokers using nicotine patch or no cessation pharmacotherapy, which may be in some part associated with higher smoking cessation rates within the bupropion-exposed group.24 The remainder of the studies either found no significant differences in mean gestational age at delivery22,26 or reported similar findings between exposed and non-exposed groups with no determination of significance levels.23,27 Whilst this review demonstrates the paucity of safety evidence, the authors are aware of several ongoing studies which will provide further insight. These include the Australian “Smoking MUMS Study,” a population-based investigation to further assess the safety of these agents in pregnancy,47 two investigating bupropion,48,49 and one of varenicline.50 Conclusion This review finds no conclusive evidence for the safety of gestational use of bupropion or varenicline. Pooling the limited available evidence suggests that bupropion has no major positive or negative impacts on the rates of congenital abnormalities, birthweight, or premature birth. Supplementary Material Supplementary data is available at Nicotine & Tobacco Research online. Funding No funding was sought to undertake this review. Declaration of Interests The authors declare no conflicts of interest. Acknowledgments The authors would like to thank April McCambridge for her assistance with this review. Professor Coleman is a National Institute for Health Research (NIHR) Senior Investigator. The views expressed in this article are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care. ET, MJ, LC, and TC were involved in the development of the research question. ET performed the electronic searches and initial screening by title and abstract. Articles were reviewed independently by ET and one of MJ, LC, or TC, and agreement was sought on whether or not these met inclusion criteria. If required, consensus was achieved by consulting a third author. Data extraction was completed by ET and checked by MJ, LC, or TC, with discrepancies resolved by consensus or by involving a third researcher, where necessary. ET was responsible for conducting the qualitative review. ET, MJ, LZ, and TC all contributed to the drafting of the final manuscript. References 1. Clarke S, Woodcock A, Bewley B. Smoking and the young - summary of a report of a working party of the Royal College of Physicians. J R Coll Physicians of Lond . 1992; 26( 4): 352– 356. 2. Curtin SC, Matthews TJ. Smoking prevalence and cessation before and during pregnancy: Data from the birth certificate, 2014. Natl Vital Stat Rep . 2016; 65( 1): 1– 14. Google Scholar PubMed  3. Cui Y, Shooshtari S, Forget EL, Clara I, Cheung KF. Smoking during pregnancy: Findings from the 2009-2010 Canadian Community Health Survey. PLoS One . 2014; 9( 1): e84640. Google Scholar CrossRef Search ADS PubMed  4. McAndrew F, Thompson J, Fellows L, et al.   Infant Feeding Survey 2010 . The NHS HSE Info Centre; 2012. https://digital.nhs.uk/catalogue/PUB08694. Accessed July 06, 2017. 5. Leonardi-Bee J, Jere ML, Britton J. Exposure to parental and sibling smoking and the risk of smoking uptake in childhood and adolescence: a systematic review and meta-analysis. Thorax . 2011; 66( 10): 847– 855. Google Scholar CrossRef Search ADS PubMed  6. Roberts KH, Munafò MR, Rodriguez D, et al.   Longitudinal analysis of the effect of prenatal nicotine exposure on subsequent smoking behavior of offspring. Nicotine Tob Res . 2005; 7( 5): 801– 808. Google Scholar CrossRef Search ADS PubMed  7. Godfrey C. Estimating the Costs to the NHS of Smoking in Pregnancy for Pregnant Women and Infants . PHRC, University of York; 2010. http://phrc.lshtm.ac.uk/papers/PHRC_A3-06_Final_Report.pdf. Accessed July 06, 2017. 8. Ruger JP, Weinstein MC, Hammond SK, Kearney MH, Emmons KM. Cost-effectiveness of motivational interviewing for smoking cessation and relapse prevention among low-income pregnant women: A randomized controlled trial. Value Health . 2008; 11( 2): 191– 198. Google Scholar CrossRef Search ADS PubMed  9. Dhalwani NN, Szatkowski L, Coleman T, Fiaschi L, Tata LJ. Prescribing of nicotine replacement therapy in and around pregnancy: A population-based study using primary care data. Br J Gen Pract . 2014; 64( 626): e554– e560. Google Scholar CrossRef Search ADS PubMed  10. Coleman T, Chamberlain C, Davey MA, et al.   Pharmacological interventions for promoting smoking cessation during pregnancy. Cochrane Database Syst Rev . 2015( 12): CD010078. doi:10.1002/14651858.CD010078.pub2 11. Vaz LR, Aveyard P, Cooper S, Leonardi-Bee J, Coleman T; SNAP Trial Team. The association between treatment adherence to nicotine patches and smoking cessation in pregnancy: A secondary analysis of a randomized controlled trial. Nicotine Tob Res . 2016; 18( 10): 1952– 1959. Google Scholar CrossRef Search ADS PubMed  12. Kotz D, Viechtbauer W, Simpson C, van Schayck OC, West R, Sheikh A. Cardiovascular and neuropsychiatric risks of varenicline: A retrospective cohort study. Lancet Respir Med . 2015; 3( 10): 761– 768. Google Scholar CrossRef Search ADS PubMed  13. Cahill K, Stead LF, Lancaster T. Nicotine receptor partial agonists for smoking cessation. Cochrane Database Syst Rev . 2012; 18( 4). CD006103. doi:10.1002/14651858.CD006103.pub6. 14. Agency EM. Champix: EPAR - Scientific Discussion . Pub European Medicines Agency; 2006. http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Scientific_Discussion/human/000699/WC500025254.pdf. Accessed July 06, 2017. 15. Wu P, Wilson K, Dimoulas P, Mills EJ. Effectiveness of smoking cessation therapies: A systematic review and meta-analysis. BMC Public Health . 2006; 6: 300. Google Scholar CrossRef Search ADS PubMed  16. Turner E, Jones M, Vaz L, et al.   A systematic review to assess the safety of drug treatments which are used rarely for smoking cessation in pregnancy: Dual nicotine replacement therapy, varenicline and bupropion. PROSPERO  2017: CRD42017067064. http://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42017067064. Accessed May 30, 2017. 17. Liberati A, Altman DG, Tetzlaff J, et al.   The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: Explanation and elaboration. PLoS Med . 2009; 6( 7): e1000100. Google Scholar CrossRef Search ADS PubMed  18. Wells G, Shea B, O’Connell D, et al.   The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-Analyses. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed May 7, 2017. 19. Higgins JPT, Altman DG, Gøtzsche PC, et al.   The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ . 2011; 343: d5928. Google Scholar CrossRef Search ADS PubMed  20. StataCorp. 2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP. https://www.stata.com/support/faqs/resources/citing-software-documentation-faqs/ 21. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials . 1986; 7( 3): 177– 188. Google Scholar CrossRef Search ADS PubMed  22. Nanovskaya TN, Oncken C, Fokina VM, et al.   Bupropion sustained release for pregnant smokers: A randomized, placebo-controlled trial. Am J Obstet Gynecol . 2017; 216( 4): 420.e1– 420.e9. Google Scholar CrossRef Search ADS   23. Stotts AL, Northrup TF, Cinciripini PM, et al.   Randomized, controlled pilot trial of bupropion for pregnant smokers: Challenges and future directions. Am J Perinatol . 2015; 32( 4): 351– 356. Google Scholar PubMed  24. Bérard A, Zhao JP, Sheehy O. Success of smoking cessation interventions during pregnancy. Am J Obstet Gynecol . 2016; 215( 5): 611.e1– 611.e8. Google Scholar CrossRef Search ADS   25. Boshier A, Wilton LV, Shakir SA. Evaluation of the safety of bupropion (Zyban) for smoking cessation from experience gained in general practice use in England in 2000. Eur J Clin Pharmacol . 2003; 59( 10): 767– 773. Google Scholar CrossRef Search ADS PubMed  26. Chan B, Einarson A, Koren G. Effectiveness of bupropion for smoking cessation during pregnancy. J Addict Dis . 2005; 24( 2): 19– 23. Google Scholar CrossRef Search ADS PubMed  27. Chun-Fai-Chan B, Koren G, Fayez I, et al.   Pregnancy outcome of women exposed to bupropion during pregnancy: A prospective comparative study. Am J Obstet Gynecol . 2005; 192( 3): 932– 936. Google Scholar CrossRef Search ADS PubMed  28. Cole JA, Modell JG, Haight BR, Cosmatos IS, Stoler JM, Walker AM. Bupropion in pregnancy and the prevalence of congenital malformations. Pharmacoepidemiol Drug Saf . 2007; 16( 5): 474– 484. Google Scholar CrossRef Search ADS PubMed  29. Einarson A, Choi J, Einarson TR, Koren G. Incidence of major malformations in infants following antidepressant exposure in pregnancy: Results of a large prospective cohort study. Can J Psychiatry . 2009; 54( 4): 242– 246. Google Scholar CrossRef Search ADS PubMed  30. GlaxoSmithKline. The Bupropion Pregnancy Registry: Final Report . 2008. http://pregnancyregistry.gsk.com/documents/bup_report_final_2008.pdf. Accessed May 27, 2017. 31. Palmsten K, Huybrechts KF, Michels KB, et al.   Antidepressant use and risk for preeclampsia. Epidemiology . 2013; 24( 5): 682– 691. Google Scholar CrossRef Search ADS PubMed  32. Harrison-Woolrych M, Paterson H, Tan M. Exposure to the smoking cessation medicine varenicline during pregnancy: A prospective nationwide cohort study. Pharmacoepidemiol Drug Saf . 2013; 22( 10): 1086– 1092. Google Scholar PubMed  33. Olsen M, Petronis KR, Froslev T, et al.   Maternal use of varenicline and risk of congenital malformations. Pharmacoepidemiol Drug Saf . 2015; 24: 244. 34. Richardson JL, Stephens S, Yates LM, et al.   Pregnancy outcomes after maternal varenicline use; analysis of surveillance data collected by the European Network of Teratology Information Services. Reprod Toxicol . 2017; 67: 26– 34. Google Scholar CrossRef Search ADS PubMed  35. Alwan S, Reefhuis J, Botto LD, Rasmussen SA, Correa A, Friedman JM; National Birth Defects Prevention Study. Maternal use of bupropion and risk for congenital heart defects. Am J Obstet Gynecol . 2010; 203( 1): 52.e1– 52.e6. Google Scholar CrossRef Search ADS   36. Louik C, Kerr S, Mitchell AA. First-trimester exposure to bupropion and risk of cardiac malformations. Pharmacoepidemiol Drug Saf . 2014; 23( 10): 1066– 1075. Google Scholar CrossRef Search ADS PubMed  37. Gisslen T, Nathan B, Thompson T, Rao R. Hyperinsulinism associated with gestational exposure to bupropion in a newborn infant. J Pediatr Endocrinol Metab . 2011; 24( 9-10): 819– 822. Google Scholar CrossRef Search ADS PubMed  38. Leventhal K, Byatt N, Lundquist R. Fetal cardiac arrhythmia during bupropion use. Acta Obstet Gynecol Scand . 2010; 89( 7): 980– 981. Google Scholar CrossRef Search ADS PubMed  39. Kaplan YC, Olgac Dündar N, Kasap B, Karadas B. Pregnancy outcome after varenicline exposure in the first trimester. Case Rep Obstet Gynecol . 2014; 2014: 263981. Google Scholar PubMed  40. Reefhuis J, Gilboa SM, Anderka M, et al.  ; National Birth Defects Prevention Study. The National Birth Defects Prevention Study: A review of the methods. Birth Defects Res A Clin Mol Teratol . 2015; 103( 8): 656– 669. Google Scholar CrossRef Search ADS PubMed  41. EESoC. Guidelines for Registration. http://www.eurocat-network.eu/aboutus/datacollection/guidelinesforregistration/guide1_4. Accessed July 06, 2017. 42. EESoC. Prevalence Tables. http://www.eurocat-network.eu/accessprevalencedata/prevalencetables. Accessed July 06, 2017. 43. Program MACD. Metropolitan atlanta congenital defects program: Executive summary. Birth Defects Res A Clin Mol Teratol . 2007; 79( 2): 66– 93. CrossRef Search ADS   44. Donahue SM, Kleinman KP, Gillman MW, Oken E. Trends in birth weight and gestational length among singleton term births in the United States: 1990-2005. Obstet Gynecol . 2010; 115( 2 Pt 1): 357– 364. Google Scholar CrossRef Search ADS PubMed  45. Statistics Canada CVS. Birth Database (CANSIM table 102–4510). http://www5.statcan.gc.ca/cansim/a26?lang=eng&id=1024510. Accessed July 06, 2017. 46. Ellard GA, Johnstone FD, Prescott RJ, Ji-Xian W, Jian-Hua M. Smoking during pregnancy: The dose dependence of birthweight deficits. Br J Obstet Gynaecol . 1996; 103( 8): 806– 813. Google Scholar CrossRef Search ADS PubMed  47. Havard A, Jorm LR, Preen D, et al.   The Smoking MUMS (Maternal Use of Medications and Safety) Study: protocol for a population-based cohort study using linked administrative data. BMJ Open . 2013; 3( 9): e003692. Google Scholar CrossRef Search ADS PubMed  48. Kranzler H. Placebo-controlled Trial of Bupropion for Smoking Cessation in Pregnant Women (BIBS). NLM identifier: NCT02188459. https://clinicaltrials.gov/ct2/show/NCT02188459. Accessed July 06, 2017. 49. Miller H. Bupropion for Smoking Cessation in Pregnancy. NLM identifier: NCT01875172. https://clinicaltrials.gov/ct2/show/NCT01875172. Accessed July 06, 2017. 50. Pfizer. Varenicline Pregnancy Cohort Study. NLM identifier: NCT0 1875172. https://clinicaltrials.gov/ct2/show/NCT01290445. Accessed July 06, 2017. © The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nicotine and Tobacco Research Oxford University Press

Systematic Review and Meta-Analysis to Assess the Safety of Bupropion and Varenicline in Pregnancy

Loading next page...
 
/lp/ou_press/systematic-review-and-meta-analysis-to-assess-the-safety-of-bupropion-Dnu3IygYyM
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
1462-2203
eISSN
1469-994X
D.O.I.
10.1093/ntr/nty055
Publisher site
See Article on Publisher Site

Abstract

Abstract Introduction Smoking in pregnancy is a substantial public health issue, but, apart from nicotine replacement therapy (NRT), pharmacological therapies are not generally used to promote cessation. Bupropion and varenicline are effective cessation methods in nonpregnant smokers and this systematic review investigates their safety in pregnancy. Methods We searched MEDLINE, EMBASE, CINAHL, and PsychINFO databases for studies of any design reporting pregnancy outcomes after bupropion or varenicline exposure. We included studies of bupropion used for smoking cessation, depression, or where the indication was unspecified. Depending on study design, quality was assessed using the Newcastle-Ottawa Scale or Cochrane Risk of Bias Tool. Most findings are reported narratively but meta-analyses were used to produce pooled estimates for the proportion of live births with congenital malformations and of the mean birthweight and gestational age at delivery following bupropion exposure. Results In total, 18 studies were included: 2 randomized controlled trials, 11 cohorts, 2 case– control studies, and 3 case reports. Study quality was variable. Gestational safety outcomes were reported in 14 bupropion and 4 varenicline studies. Meaningful meta-analysis was only possible for bupropion exposure, for which the pooled estimated proportion of congenital malformations amongst live-born infants was 1.0% (95% CI = 0.0%–3.0%, I2 = 80.9%, 4 studies) and the mean birthweight and mean gestational age at delivery was 3305.9 g (95% CI = 3173.2–3438.7 g, I2 = 77.6%, 5 studies) and 39.2 weeks (95% CI = 38.8–39.6 weeks, I2 = 69.9%, 5 studies), respectively. Conclusions There was no strong evidence that either major positive or negative outcomes were associated with gestational use of bupropion or varenicline. PROSPERO registration number CRD42017067064. Implications We believe this to be the first systematic review investigating the safety of bupropion and varenicline in pregnancy. Meta-analysis of outcomes following bupropion exposure in pregnancy suggests that there are no major positive or negative impacts on the rate of congenital abnormalities, birthweight, or premature birth. Overall, we found no evidence that either of these treatments might be harmful in pregnancy, and no strong evidence to suggest safety, but available evidence is of poor quality. Introduction Smoking in pregnancy is associated with increased risks of miscarriage, stillbirth, prematurity, low birth weight, perinatal morbidity, and mortality1 and is a significant problem in developed countries where rates vary between 8% and 23% of pregnant women smoking in pregnancy.2–4 Children of smoking mothers are twice as likely to become smokers themselves,5 so smoking in pregnancy and afterwards encourages the persistence of smoking across generations.6 Smoking in pregnancy is declining in developed countries but remains highest amongst younger, socially disadvantaged women4 and the annual costs of managing the smoking-attributable maternal and infant disease can be substantial.7 Studies have shown that pregnancy is the life event which most motivates smokers to attempt cessation, with around half of pregnant smokers attempting to quit.4 In addition, although the cost-efficacy of smoking cessation interventions in pregnancy is unclear,8 stopping smoking in pregnancy is likely to save healthcare resources, both with respect to the health of the infant and in the wider context of preventing the perseveration of smoking in the next generation. Thus, promoting smoking cessation during pregnancy will substantially improve the health not only of the infant and mother but of their extended family and, in the longer term, will contribute to reducing the substantial healthcare cost of smoking-related diseases. However, compared to those available for nonpregnant smokers, relatively few effective cessation interventions can be used in pregnancy and nicotine replacement therapy (NRT) is the only drug treatment used to any extent.9 The UK National Institute for Healthcare and Excellence (NICE) recommends using NRT, believing it safer that continued smoking in pregnancy. However, in pregnancy NRT has at best only a borderline significant effect on cessation (RR 1.28, 95% CI 0.99–1.66)10 and this lower efficacy, compared to use outside of pregnancy, is probably caused by poor adherence to NRT.11 If they were considered sufficiently safe, other effective cessation pharmacotherapies, varenicline and bupropion, could also be tried in pregnancy. Varenicline is well tolerated12 and probably more effective than other cessation treatments;13 animal research suggests it is not teratogenic.14 Similarly, bupropion is an effective smoking cessation aid which approximately doubles nonpregnant smokers chances of stopping.15 If varenicline or bupropion were to be proven effective for pregnant smokers, the health benefits which would accrue from stopping smoking would very likely outweigh any minor adverse effects. Consequently, to help assesses whether experimental studies might be ethical, we review evidence for the safety of varenicline and bupropion in pregnancy. Methods A study protocol was registered16 and the review adhered to Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines.17 Inclusion Criteria We included studies of any design that reported adverse pregnancy outcomes experienced by mothers, fetuses, or infants following use of varenicline or bupropion in pregnancy. Exclusion Criteria We excluded studies that presented no empirical data and those in which interventions combined bupropion or varenicline with other cessation pharmacotherapies. Search Strategy We searched the MEDLINE, EMBASE, CINAHL, and PsychINFO databases and hand-searched reference lists from reviews and included papers. We also searched for ongoing and unpublished studies at: www.gsk-clinicalstudyregister.com; clincialtrials.gov; www.who.int/trialsearch; www.controlled-trials.com/isrctn; and www.ukctg.nihr.ac.uk. As bupropion and varenicline were licensed and became available relatively recently, we sought studies published from 1990 until May 25, 2017 with no language restrictions. Search terms relating to pregnancy were developed from those used in a Cochrane review10 and were combined with qualitative terms relating to smoking, varenicline or bupropion. The protocol also states that we intended to include studies in which women used “dual” nicotine replacement therapy (NRT);16 however, only two such studies were identified, and therefore we retrospectively decided that they should not be reviewed separately from other NRT studies (eg, those investigating “mono” NRT). Data Extraction Titles and abstracts were screened by the lead reviewer, retrieving complete manuscripts if necessary to decide on inclusion. All articles were independently assessed by two reviewers to confirm inclusion in the review, with adjudication via a third reviewer when agreement was not met. The following data were extracted by the lead reviewer and checked by a second reviewer, with any discrepancies resolved by a third reviewer: aims and design, numbers of participants, outcomes, data collection, analysis methods, and findings. Where studies reported interim analyses, further details were requested from authors. Quality Assessment Quality assessment was performed using Newcastle–Ottawa Scales18 for cohort or cross-sectional studies and the Cochrane Risk of Bias Assessment for RCTs.19 Initial assessment was made by the lead reviewer and checked by a second reviewer, and discrepancies resolved by a third reviewer. Where a trial had already been quality-assessed for a Cochrane review, we used this published assessment. Data Synthesis A priori, we anticipated that review studies might be so diverse that meaningful data synthesis could be challenging. Hence, we planned making final decisions on whether or not meta-analyses were possible once data extraction was finished, and outcomes of this deliberation are reported alongside review findings. If performed, we anticipated that meta-analyses would be conducted in Stata version 1420 using a random effects DerSimonian and Laird model to generate pooled means and 95% confidence intervals with heterogeneity quantified by the I2 statistic.21 Rather than not pool studies in the presence of a high I2 value, we planned to present this statistic alongside meta-analysis findings to inform the reader of the extent to which pooled estimates should be treated cautiously. Results We identified 772 studies (1053 including duplicates); no ongoing trials were identified from registries and no completed, unpublished studies were identified from pharmaceutical company databases. We identified 30 articles for retrieval in full, with 18 being included in the review (Figure 1); study details and outcomes are shown in Supplementary Material Table 1. Figure 1. View largeDownload slide PRISMA flowchart of included studies. Figure 1. View largeDownload slide PRISMA flowchart of included studies. Study Design, Outcome Measures, and Quality Assessment Included studies comprised two randomized controlled trials (RCTs),22,23 eleven cohort studies,24–34 two case–control studies,35,36 and three case reports.37–39 Maternal or fetal adverse outcomes were reported in fourteen bupropion22–31,36–38 and four varenicline studies.32–34,39 Although bupropion can be used as an anti-depressant or for smoking cessation, only two RCTs specified that bupropion had been prescribed for smoking cessation.22,23 The three observational studies evaluating varenicline exposure did not explicitly state that varenicline was used for smoking cessation, but all discussed its sole indication as a cessation pharmacotherapy.32–34 Congenital malformations were reported in eight studies (six following bupropion27–30,35,36 and two following varenicline33,34); reported malformation classification systems are described in Supplementary Material Table 1. Birthweight and gestational age at delivery were reported in five bupropion studies.22–24,26,27 Other outcomes included: fetal loss or stillbirth25,27,30,32,34; fetal length or head circumference22,23; preterm birth22–24; maternal medication adverse effects23,25,39; and pre-eclampsia.31 An overview of reported outcomes is shown in Table 1. Table 1. Overview of Reported Outcomes by Study   Fetal outcomes  Maternal outcomes  Paper  Drug of interest  Congenital malformations (incl CV)  Birth weight  GA at delivery  Pregnancy outcome  Length/Head circumference  Preterm birth  Other*  Pregnancy/Maternal adverse effects  Cotinine levels  Pre-eclampsia  Alwan  2010  Bupropion  ✓                    Berard  2016  Bupropion    ✓  ✓      ✓  ✓        Boshier  2003  Bupropion        ✓        ✓      Chan  2005  Bupropion    ✓  ✓                Chun-Fai-Chan  2005  Bupropion  ✓  ✓  ✓  ✓              Cole  2007  Bupropion  ✓                    Einarson  2009  Bupropion  ✓                    Gisslen  2011  Bupropion              ✓        GSK  2008  Bupropion  ✓      ✓              Leventhal  2010  Bupropion              ✓        Louik  2014  Bupropion  ✓                    Nanovskaya  2017  Bupropion    ✓  ✓    ✓  ✓  ✓    ✓    Palmsten  2013  Bupropion                    ✓  Stotts  2015  Bupropion    ✓  ✓    ✓    ✓  ✓  ✓    Harrison- Woolrych  2013  Varenicline        ✓      ✓        Kaplan  2014  Varenicline                ✓      Olsen  2015  Varenicline  ✓                    Richardson  2017  Varenicline  ✓      ✓                Fetal outcomes  Maternal outcomes  Paper  Drug of interest  Congenital malformations (incl CV)  Birth weight  GA at delivery  Pregnancy outcome  Length/Head circumference  Preterm birth  Other*  Pregnancy/Maternal adverse effects  Cotinine levels  Pre-eclampsia  Alwan  2010  Bupropion  ✓                    Berard  2016  Bupropion    ✓  ✓      ✓  ✓        Boshier  2003  Bupropion        ✓        ✓      Chan  2005  Bupropion    ✓  ✓                Chun-Fai-Chan  2005  Bupropion  ✓  ✓  ✓  ✓              Cole  2007  Bupropion  ✓                    Einarson  2009  Bupropion  ✓                    Gisslen  2011  Bupropion              ✓        GSK  2008  Bupropion  ✓      ✓              Leventhal  2010  Bupropion              ✓        Louik  2014  Bupropion  ✓                    Nanovskaya  2017  Bupropion    ✓  ✓    ✓  ✓  ✓    ✓    Palmsten  2013  Bupropion                    ✓  Stotts  2015  Bupropion    ✓  ✓    ✓    ✓  ✓  ✓    Harrison- Woolrych  2013  Varenicline        ✓      ✓        Kaplan  2014  Varenicline                ✓      Olsen  2015  Varenicline  ✓                    Richardson  2017  Varenicline  ✓      ✓              View Large Table 1. Overview of Reported Outcomes by Study   Fetal outcomes  Maternal outcomes  Paper  Drug of interest  Congenital malformations (incl CV)  Birth weight  GA at delivery  Pregnancy outcome  Length/Head circumference  Preterm birth  Other*  Pregnancy/Maternal adverse effects  Cotinine levels  Pre-eclampsia  Alwan  2010  Bupropion  ✓                    Berard  2016  Bupropion    ✓  ✓      ✓  ✓        Boshier  2003  Bupropion        ✓        ✓      Chan  2005  Bupropion    ✓  ✓                Chun-Fai-Chan  2005  Bupropion  ✓  ✓  ✓  ✓              Cole  2007  Bupropion  ✓                    Einarson  2009  Bupropion  ✓                    Gisslen  2011  Bupropion              ✓        GSK  2008  Bupropion  ✓      ✓              Leventhal  2010  Bupropion              ✓        Louik  2014  Bupropion  ✓                    Nanovskaya  2017  Bupropion    ✓  ✓    ✓  ✓  ✓    ✓    Palmsten  2013  Bupropion                    ✓  Stotts  2015  Bupropion    ✓  ✓    ✓    ✓  ✓  ✓    Harrison- Woolrych  2013  Varenicline        ✓      ✓        Kaplan  2014  Varenicline                ✓      Olsen  2015  Varenicline  ✓                    Richardson  2017  Varenicline  ✓      ✓                Fetal outcomes  Maternal outcomes  Paper  Drug of interest  Congenital malformations (incl CV)  Birth weight  GA at delivery  Pregnancy outcome  Length/Head circumference  Preterm birth  Other*  Pregnancy/Maternal adverse effects  Cotinine levels  Pre-eclampsia  Alwan  2010  Bupropion  ✓                    Berard  2016  Bupropion    ✓  ✓      ✓  ✓        Boshier  2003  Bupropion        ✓        ✓      Chan  2005  Bupropion    ✓  ✓                Chun-Fai-Chan  2005  Bupropion  ✓  ✓  ✓  ✓              Cole  2007  Bupropion  ✓                    Einarson  2009  Bupropion  ✓                    Gisslen  2011  Bupropion              ✓        GSK  2008  Bupropion  ✓      ✓              Leventhal  2010  Bupropion              ✓        Louik  2014  Bupropion  ✓                    Nanovskaya  2017  Bupropion    ✓  ✓    ✓  ✓  ✓    ✓    Palmsten  2013  Bupropion                    ✓  Stotts  2015  Bupropion    ✓  ✓    ✓    ✓  ✓  ✓    Harrison- Woolrych  2013  Varenicline        ✓      ✓        Kaplan  2014  Varenicline                ✓      Olsen  2015  Varenicline  ✓                    Richardson  2017  Varenicline  ✓      ✓              View Large Study quality was variable (Table 2) with seven of the eleven observational studies assessed as of low methodological quality (score of <7);25,26,29,30,32–34 the three case reports were considered to provide only low-quality evidence.37–39 Table 2. Quality Assessment of Included Studies Cohort studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Outcome (max 3*)  Total (max 9*)  Berard, 2016  ☆☆☆  ☆☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Palmsten, 2013  ☆☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Chun-Fai-Chan, 2005  ☆☆☆  ☆☆  ☆☆  ☆☆☆☆☆☆☆  Cole, 2007  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Einarson, 2009  ☆☆☆  ☆  ☆☆  ☆☆☆☆☆☆  Boshier, 2004  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Chan, 2005  ☆☆  ☆  ☆☆  ☆☆☆☆☆  Harrison-Woolrych, 2013  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Olsen, 2015  ☆☆☆  0  ☆☆  ☆☆☆☆☆  GlaxoSmithKleine, 2008  ☆  0  ☆☆  ☆☆☆  Richardson, 2017  ☆☆  0  ☆  ☆☆☆  Case–control studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Exposure (max 3*)  Total (max 9*)  Alwan, 2010  ☆☆☆☆  ☆☆  ☆  ☆☆☆☆☆☆☆  Louik, 2014  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Randomized controlled trials—Cochrane Risk of Bias    Random sequence generation  Allocation concealment  Blinding of participants and personnel  Blinding of outcome assessment  Incomplete outcome data  Selective outcome reporting  Nanovskaya, 2017  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Stotts, 2015*  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Case reports  Gisslen, 2011  High risk by design, unblinded assessments  Kaplan, 2014  High risk by design, unblinded assessments  Leventhal, 2010  High risk by design, unblinded assessments  Cohort studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Outcome (max 3*)  Total (max 9*)  Berard, 2016  ☆☆☆  ☆☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Palmsten, 2013  ☆☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Chun-Fai-Chan, 2005  ☆☆☆  ☆☆  ☆☆  ☆☆☆☆☆☆☆  Cole, 2007  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Einarson, 2009  ☆☆☆  ☆  ☆☆  ☆☆☆☆☆☆  Boshier, 2004  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Chan, 2005  ☆☆  ☆  ☆☆  ☆☆☆☆☆  Harrison-Woolrych, 2013  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Olsen, 2015  ☆☆☆  0  ☆☆  ☆☆☆☆☆  GlaxoSmithKleine, 2008  ☆  0  ☆☆  ☆☆☆  Richardson, 2017  ☆☆  0  ☆  ☆☆☆  Case–control studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Exposure (max 3*)  Total (max 9*)  Alwan, 2010  ☆☆☆☆  ☆☆  ☆  ☆☆☆☆☆☆☆  Louik, 2014  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Randomized controlled trials—Cochrane Risk of Bias    Random sequence generation  Allocation concealment  Blinding of participants and personnel  Blinding of outcome assessment  Incomplete outcome data  Selective outcome reporting  Nanovskaya, 2017  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Stotts, 2015*  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Case reports  Gisslen, 2011  High risk by design, unblinded assessments  Kaplan, 2014  High risk by design, unblinded assessments  Leventhal, 2010  High risk by design, unblinded assessments  *Quality assessment for Stotts 2015 as assessed in the Cochrane Review “Pharmacological interventions for promoting smoking cessation during pregnancy” (2015)10 View Large Table 2. Quality Assessment of Included Studies Cohort studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Outcome (max 3*)  Total (max 9*)  Berard, 2016  ☆☆☆  ☆☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Palmsten, 2013  ☆☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Chun-Fai-Chan, 2005  ☆☆☆  ☆☆  ☆☆  ☆☆☆☆☆☆☆  Cole, 2007  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Einarson, 2009  ☆☆☆  ☆  ☆☆  ☆☆☆☆☆☆  Boshier, 2004  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Chan, 2005  ☆☆  ☆  ☆☆  ☆☆☆☆☆  Harrison-Woolrych, 2013  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Olsen, 2015  ☆☆☆  0  ☆☆  ☆☆☆☆☆  GlaxoSmithKleine, 2008  ☆  0  ☆☆  ☆☆☆  Richardson, 2017  ☆☆  0  ☆  ☆☆☆  Case–control studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Exposure (max 3*)  Total (max 9*)  Alwan, 2010  ☆☆☆☆  ☆☆  ☆  ☆☆☆☆☆☆☆  Louik, 2014  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Randomized controlled trials—Cochrane Risk of Bias    Random sequence generation  Allocation concealment  Blinding of participants and personnel  Blinding of outcome assessment  Incomplete outcome data  Selective outcome reporting  Nanovskaya, 2017  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Stotts, 2015*  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Case reports  Gisslen, 2011  High risk by design, unblinded assessments  Kaplan, 2014  High risk by design, unblinded assessments  Leventhal, 2010  High risk by design, unblinded assessments  Cohort studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Outcome (max 3*)  Total (max 9*)  Berard, 2016  ☆☆☆  ☆☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Palmsten, 2013  ☆☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆☆  Chun-Fai-Chan, 2005  ☆☆☆  ☆☆  ☆☆  ☆☆☆☆☆☆☆  Cole, 2007  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Einarson, 2009  ☆☆☆  ☆  ☆☆  ☆☆☆☆☆☆  Boshier, 2004  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Chan, 2005  ☆☆  ☆  ☆☆  ☆☆☆☆☆  Harrison-Woolrych, 2013  ☆☆☆  0  ☆☆  ☆☆☆☆☆  Olsen, 2015  ☆☆☆  0  ☆☆  ☆☆☆☆☆  GlaxoSmithKleine, 2008  ☆  0  ☆☆  ☆☆☆  Richardson, 2017  ☆☆  0  ☆  ☆☆☆  Case–control studies—Newcastle-Ottawa scale    Selection (max 4*)  Comparability (max 2*)  Exposure (max 3*)  Total (max 9*)  Alwan, 2010  ☆☆☆☆  ☆☆  ☆  ☆☆☆☆☆☆☆  Louik, 2014  ☆☆☆  ☆  ☆☆☆  ☆☆☆☆☆☆☆  Randomized controlled trials—Cochrane Risk of Bias    Random sequence generation  Allocation concealment  Blinding of participants and personnel  Blinding of outcome assessment  Incomplete outcome data  Selective outcome reporting  Nanovskaya, 2017  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Stotts, 2015*  Low risk  Unclear  Low risk  Low risk  High risk  High risk  Case reports  Gisslen, 2011  High risk by design, unblinded assessments  Kaplan, 2014  High risk by design, unblinded assessments  Leventhal, 2010  High risk by design, unblinded assessments  *Quality assessment for Stotts 2015 as assessed in the Cochrane Review “Pharmacological interventions for promoting smoking cessation during pregnancy” (2015)10 View Large Potential for Meta-Analyses From the distribution of outcomes across studies and study designs (Table 1), it was clear that the few studies investigating varenicline were so different in design and outcome measurement that meaningful meta-analyses were not possible. For bupropion studies, meta-analyses investigating effects on congenital malformation, mean birthweight, and mean gestational age at birth were considered feasible and were undertaken by combining studies or study arms with sufficiently similar designs. For the meta-analysis investigating effects of bupropion exposure on congenital malformations, we included only cohort studies.27–30 For birthweight and gestational age at birth meta-analyses, we pooled data from bupropion-exposed arms in cohort studies24,26,27 and RCTs22,23 to determine the mean value associated with each outcome. Bupropion Congenital Malformations Six studies, four cohort27–30 and two case control studies,35,36 reported congenital malformations. Cohort studies included 3376 pregnancies (Figure 2) and from these studies the pooled estimate for the percentage of congenital malformations amongst live-born infants exposed to bupropion at any point during gestation was 1.0% (95% CI = 0.0%–3.0%, I2 = 80.9%) (Figure 3a). As individual studies classified congenital malformations in different ways (Supplementary Material Table 1), we accepted the presence or absence of malformations was as defined within each study and no attempt was made to derive a single classification system applied to all studies. Pregnancies which ended in stillbirth, miscarriage, intra-uterine fetal death or termination were excluded from the analysis, so we defined the proportion of pregnancies with congenital malformations as follows:  Proportion with any congenital malformation=number of live born infants with a malformationtotal number of live born infants exposed in utero Figure 2. View largeDownload slide Overview of pregnancies included in meta-analysis of congenital malformations following bupropion exposure. Figure 2. View largeDownload slide Overview of pregnancies included in meta-analysis of congenital malformations following bupropion exposure. Figure 3. View largeDownload slide (A) Proportion of congenital malformations following bupropion exposure. (B) Mean birthweight following bupropion exposure. (C) Mean gestational age at delivery following bupropion exposure. Figure 3. View largeDownload slide (A) Proportion of congenital malformations following bupropion exposure. (B) Mean birthweight following bupropion exposure. (C) Mean gestational age at delivery following bupropion exposure. The two case–control studies had conflicting results.35,36 Both used National Birth Defects Prevention Study (NBDPS) criteria40 to classify congenital cardiac defects; overall, Alwan found no evidence that maternal bupropion exposure in pregnancy increased infants’ risks of developing congenital cardiac defects (adjusted odds ratio 1.4, 95% CI = 0.8–2.5).35 Alwan did however, report an increased risk of left outflow tract cardiac defects (adjusted odds ratio 2.6, 95% CI = 1.2–5.7) which was not found by Louik36 (adjusted odds ratio 0.4, 95% CI = 0–2.4). Louik investigated the risk of developing eight different cardiac defects following bupropion exposure but did not attempt to estimate the overall risk of any cardiac defect and reported an increased risk of ventricular septal defects (adjusted odds ratio 2.9, 95% CI = 1.5–5.5) which was not found by Alwan (adjusted odds ratio 1.2, 95% CI = 0.5–3.4). Louik found no increased risks for other sub-categories of cardiac defects following bupropion exposure. Birthweight Two RCTs (combined n = 35) found no significant differences in birthweight between bupropion or placebo groups (Supplementary Material Table 1) but both will likely have been under-powered to detect clinically significant differences.22,23 One controlled cohort study conducted in smokers found significantly higher mean birthweights amongst infants born after exposure to bupropion (3315.9 g, SD = 553.3, n = 72) compared to those who smoked and used no treatment (2943.5 g, SD = 733.5, n = 900, p < .05).24 However, this finding was not replicated in two other bupropion cohort studies.26,27 Meta-analysis of the 262 pregnancies in bupropion-exposed arms from cohorts and RCTs gives a pooled estimate for mean birthweight amongst infants exposed to bupropion of 3305.9 g (95% CI = 3173.2–3438.7 g, I2 = 77.6%, n = 262) (Figure 3b). Gestational Age at Delivery Of five studies reporting gestational age at birth, two were RCTs22,23and three cohort studies.24,26,27 No significant differences were found between trial groups within the likely under-powered RCTs or between exposure groups in two of the cohort studies.26,27 However, one cohort study found that infants born after bupropion exposure had a significantly higher mean gestational age at birth (39.1 weeks, SD = 1.3, n = 72) compared to non-exposed infants born to smokers (37.5 weeks, SD = 3.3, n = 900, p < .05).24 The pooled estimate for mean gestational age at delivery in the five studies which included 260 pregnancies was 39.2 weeks (95% CI = 38.8–39.6, I2 = 69.9%) (Figure 3c). Foetal Loss Three cohort studies reported fetal loss following bupropion exposure,25,27,30 with only one study including control group data.27 The GlaxoSmithKline “Bupropion Pregnancy Registry” cohort reported data from 994 prospectively registered pregnant women (featuring 1005 monitored fetuses) following gestational bupropion exposure. Following first trimester bupropion exposure, there were 669 live births, 3 fetal deaths occurring at or later than 20 weeks gestation, 38 induced abortions, and 96 spontaneous pregnancy losses occurring before 20 weeks. Following second trimester exposure, there were 145 live births, 1 induced abortion and 1 spontaneous pregnancy loss and after bupropion exposure in the third trimester there were 51 live births and 1 fetal death. In prospectively registered pregnancies, 603 were either lost to follow-up or pending delivery when the register closed, resulting in a loss of outcome data.30 One study found significantly higher rates of spontaneous abortions (p = .009) and therapeutic abortions (p = .015) in a bupropion cohort compared with those exposed to “non-teratogenic” agents.27 A small uncontrolled, cohort study of 12 pregnancies, reported 5 live births, 2 therapeutic terminations (no explanation of reasons for termination given), 1 intra-uterine death, and 4 cases that were lost to follow-up following gestational bupropion exposure.25 Varenicline Four varenicline studies reported relevant adverse outcomes; three cohort studies32–34 and one case report study.39 No study explicitly stated that such exposures were unintentional, though this was probably the case as there were no smoking cessation studies and varenicline has no therapeutic indications in pregnancy. Richardson reported outcomes and congenital malformations following exposure to varenicline in pregnancy (n = 89) in a study which compared pregnant women exposed to non-teratogenic agents (n = 267) with those exposed to either NRT or bupropion (combined group, n = 267). As determined by the EUROCAT classification system for congenital malformations,41 seven infants (7.87%) in the varenicline group were reported to have a congenital malformation; two of which were “major” and five “minor.” No significant between-group differences were found in malformation rates.34 Another cohort study reported malformation rates in infants both exposed (4.3%, n = 254) and not exposed to varenicline in utero and also in those exposed to maternal smoking during gestation (4.2%, n = 5296), and those exposed to neither varenicline nor smoking in utero (4.2%, n = 656139). Rates appeared similar but no statistical comparison of groups was undertaken.33 One uncontrolled cohort of 23 varenicline-exposed pregnancies reported 14 live full-term births (61%), two live preterm (<37 weeks gestation) births (9%), four terminations of pregnancy (17%), two spontaneous or missed abortions (9%), and one ectopic pregnancy (4%). Within this cohort five fetal adverse events were reported, as shown in Supplementary Material Table 1.32 One case report described a normal pregnancy, delivery, and infant health until six months following gestational varenicline exposure for 4 weeks from the last menstrual period.39 Discussion We believe that this is the first systematic review investigating the safety of bupropion and varenicline in pregnancy. We found no evidence that either of these treatments might be harmful in pregnancy but available evidence is of poor quality and there is also no strong evidence to suggest safety. Most studies investigated outcomes following bupropion exposure and pooled estimates for birthweight, gestation at birth, and congenital abnormality rates do not suggest that any of these outcomes are adversely affected. However, estimates’ confidence intervals were relatively wide and more data would be required to improve precision. Far fewer studies investigated outcomes following varenicline exposures and overall there is probably insufficient evidence to make firm conclusions about the safety on any of either therapy in pregnancy. This study has some limitations. Relatively few studies were eligible for inclusion reflecting a paucity of relevant data and the majority of those in the review were small and observational, with only two RCTs.22,23 This restricted the assessment of potential causal relationships. Included studies generally had low methodological quality; some are case reports37–39 or cohort studies which lack control groups.25,30,32 Relatively few studies reported similar outcomes, restricting the potential for meta-analysis and, where these were conducted, heterogeneity was high, presumably due to differences between study designs and comparator groups; this heterogeneity means that this study’s calculated pooled estimated need to be treated with caution. As there were so few studies investigating bupropion specifically for smoking cessation purposes,22,23 we combined those which stated explicitly that bupropion had been used for smoking cessation and also those where the purpose of bupropion use was unspecified and this could have been prescribed for either smoking cessation or depression. Consequently, some of the data in our meta-analyses will have come from nonsmokers who would be expected to have better birth outcomes than smokers. As the evidence regarding the safety of varenicline during pregnancy is sparser than that for bupropion, with safety data identified in only four studies,32–34,39 this review predominantly focuses on bupropion. Because of the low-quality designs (eg, uncontrolled32) and complex comparator groups (eg, pregnant women exposed to non-teratogenic agents or those using either NRT or bupropion as a single group), no conclusion as to the safety of varenicline can be made.34 A strength of this review is its novelty and systematic approach. By including studies with any design, it is likely we have identified the majority of available safety evidence. Additionally, the rigorous quality assessment indicates that further investigation of the safety of pharmacotherapy during pregnancy is required. We have aimed to maximize use of available data and believe we have made the best use of this whilst also being sensitive to the limitations inherent in empirical studies’ designs. Our pooled estimate for the proportion of congenital malformations in live-born infants following gestational bupropion exposure (1%) is similar to those reported in comparable populations. From 2011–2015 EUROCAT, a European network of population-based registries, reported a congenital abnormalities rate of 2.5% amongst live births, fetal deaths, stillbirths, and terminations for fetal abnormalities.42 In addition, the MACDP (Metropolitan Atlanta Congenital Defects Program) determined the rate in five central counties of Atlanta between 1968 and 2003, to be 2.67%.43 Included papers reported only abnormalities within live-born infants; however, population-based registries generally include pregnancies in which the fetus dies before term. As abnormalities are less likely to be present in live-born infants, the 1% review-derived rate may underestimate prevalence in all pregnancies and the interpretation of our data is not completely straightforward. Despite this, it is reassuring that the upper 95% CI for the estimate (3%) is close to population estimates; further studies would increase the precision of this estimate and possibly provide further reassurance that congenital abnormality rates are not higher after bupropion exposure. Although 95% confidence intervals are consistent with wide range of values, the meta-analysis derived point estimate for mean birthweight following bupropion exposure 3305.9 g (95% CI = 3173.2–3438.7 g) was similar to the population average of the countries in which the studies reporting this outcome were conducted. Studies included in the birthweight meta-analysis were predominantly North American and only one was UK-based.27 Population-based data show that the average birthweight for those born between 37 and 41 weeks of gestation in the United States in 2005 was 3389 g (SD = 466),44 and in 2009 the mean birthweight of Canadian babies was 3364 g.45 Calculating the effects of bupropion on birthweight is also complicated by the known reduction in birthweight associated with maternal smoking; for example, one large study of 3338 mothers reported an adjusted birthweight deficit within babies born to active smokers averaging 226 g.46 Four of the studies contributing to the pooled estimate for birthweight following bupropion exposure included only pregnant smokers22–24,26 and the remaining study controlled for the effects of smoking by matching study groups by smoking status.27 None of the review studies reported a mean birthweight within the bupropion-exposed groups that was significantly less than their control groups22–24,26,27; in four studies, birthweights were higher in the bupropion cohorts,22,23,26,27 and in one study this finding was statistically significant.24 The latter study reported increasingly heavier birthweights between pregnant smokers who used no cessation pharmacotherapy, who used a nicotine patch, and who used bupropion, with rates of smoking cessation during pregnancy of 0%, 79%, and 81%, respectively. The high rates of smoking cessation in the bupropion-exposed cohort in this study may have been the driving factor behind the higher birthweight within the group, rather than being associated with bupropion pharmacotherapy itself, but this nonetheless is a beneficial outcome. We calculated the pooled mean gestational age at delivery following bupropion exposure as 39.2 weeks (95% CI = 38.8–39.6), as shown in Figure 3. This is comparable to the normal 40-week gestation and clinically insignificant. When assessing the studies that compared bupropion-exposed infants to pregnant smokers not using bupropion, there was also no evidence of a significant negative effect. One study found that the mean gestational age at birth for infants born to pregnant smokers using bupropion was significantly later than that of pregnant smokers using nicotine patch or no cessation pharmacotherapy, which may be in some part associated with higher smoking cessation rates within the bupropion-exposed group.24 The remainder of the studies either found no significant differences in mean gestational age at delivery22,26 or reported similar findings between exposed and non-exposed groups with no determination of significance levels.23,27 Whilst this review demonstrates the paucity of safety evidence, the authors are aware of several ongoing studies which will provide further insight. These include the Australian “Smoking MUMS Study,” a population-based investigation to further assess the safety of these agents in pregnancy,47 two investigating bupropion,48,49 and one of varenicline.50 Conclusion This review finds no conclusive evidence for the safety of gestational use of bupropion or varenicline. Pooling the limited available evidence suggests that bupropion has no major positive or negative impacts on the rates of congenital abnormalities, birthweight, or premature birth. Supplementary Material Supplementary data is available at Nicotine & Tobacco Research online. Funding No funding was sought to undertake this review. Declaration of Interests The authors declare no conflicts of interest. Acknowledgments The authors would like to thank April McCambridge for her assistance with this review. Professor Coleman is a National Institute for Health Research (NIHR) Senior Investigator. The views expressed in this article are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care. ET, MJ, LC, and TC were involved in the development of the research question. ET performed the electronic searches and initial screening by title and abstract. Articles were reviewed independently by ET and one of MJ, LC, or TC, and agreement was sought on whether or not these met inclusion criteria. If required, consensus was achieved by consulting a third author. Data extraction was completed by ET and checked by MJ, LC, or TC, with discrepancies resolved by consensus or by involving a third researcher, where necessary. ET was responsible for conducting the qualitative review. ET, MJ, LZ, and TC all contributed to the drafting of the final manuscript. References 1. Clarke S, Woodcock A, Bewley B. Smoking and the young - summary of a report of a working party of the Royal College of Physicians. J R Coll Physicians of Lond . 1992; 26( 4): 352– 356. 2. Curtin SC, Matthews TJ. Smoking prevalence and cessation before and during pregnancy: Data from the birth certificate, 2014. Natl Vital Stat Rep . 2016; 65( 1): 1– 14. Google Scholar PubMed  3. Cui Y, Shooshtari S, Forget EL, Clara I, Cheung KF. Smoking during pregnancy: Findings from the 2009-2010 Canadian Community Health Survey. PLoS One . 2014; 9( 1): e84640. Google Scholar CrossRef Search ADS PubMed  4. McAndrew F, Thompson J, Fellows L, et al.   Infant Feeding Survey 2010 . The NHS HSE Info Centre; 2012. https://digital.nhs.uk/catalogue/PUB08694. Accessed July 06, 2017. 5. Leonardi-Bee J, Jere ML, Britton J. Exposure to parental and sibling smoking and the risk of smoking uptake in childhood and adolescence: a systematic review and meta-analysis. Thorax . 2011; 66( 10): 847– 855. Google Scholar CrossRef Search ADS PubMed  6. Roberts KH, Munafò MR, Rodriguez D, et al.   Longitudinal analysis of the effect of prenatal nicotine exposure on subsequent smoking behavior of offspring. Nicotine Tob Res . 2005; 7( 5): 801– 808. Google Scholar CrossRef Search ADS PubMed  7. Godfrey C. Estimating the Costs to the NHS of Smoking in Pregnancy for Pregnant Women and Infants . PHRC, University of York; 2010. http://phrc.lshtm.ac.uk/papers/PHRC_A3-06_Final_Report.pdf. Accessed July 06, 2017. 8. Ruger JP, Weinstein MC, Hammond SK, Kearney MH, Emmons KM. Cost-effectiveness of motivational interviewing for smoking cessation and relapse prevention among low-income pregnant women: A randomized controlled trial. Value Health . 2008; 11( 2): 191– 198. Google Scholar CrossRef Search ADS PubMed  9. Dhalwani NN, Szatkowski L, Coleman T, Fiaschi L, Tata LJ. Prescribing of nicotine replacement therapy in and around pregnancy: A population-based study using primary care data. Br J Gen Pract . 2014; 64( 626): e554– e560. Google Scholar CrossRef Search ADS PubMed  10. Coleman T, Chamberlain C, Davey MA, et al.   Pharmacological interventions for promoting smoking cessation during pregnancy. Cochrane Database Syst Rev . 2015( 12): CD010078. doi:10.1002/14651858.CD010078.pub2 11. Vaz LR, Aveyard P, Cooper S, Leonardi-Bee J, Coleman T; SNAP Trial Team. The association between treatment adherence to nicotine patches and smoking cessation in pregnancy: A secondary analysis of a randomized controlled trial. Nicotine Tob Res . 2016; 18( 10): 1952– 1959. Google Scholar CrossRef Search ADS PubMed  12. Kotz D, Viechtbauer W, Simpson C, van Schayck OC, West R, Sheikh A. Cardiovascular and neuropsychiatric risks of varenicline: A retrospective cohort study. Lancet Respir Med . 2015; 3( 10): 761– 768. Google Scholar CrossRef Search ADS PubMed  13. Cahill K, Stead LF, Lancaster T. Nicotine receptor partial agonists for smoking cessation. Cochrane Database Syst Rev . 2012; 18( 4). CD006103. doi:10.1002/14651858.CD006103.pub6. 14. Agency EM. Champix: EPAR - Scientific Discussion . Pub European Medicines Agency; 2006. http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Scientific_Discussion/human/000699/WC500025254.pdf. Accessed July 06, 2017. 15. Wu P, Wilson K, Dimoulas P, Mills EJ. Effectiveness of smoking cessation therapies: A systematic review and meta-analysis. BMC Public Health . 2006; 6: 300. Google Scholar CrossRef Search ADS PubMed  16. Turner E, Jones M, Vaz L, et al.   A systematic review to assess the safety of drug treatments which are used rarely for smoking cessation in pregnancy: Dual nicotine replacement therapy, varenicline and bupropion. PROSPERO  2017: CRD42017067064. http://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42017067064. Accessed May 30, 2017. 17. Liberati A, Altman DG, Tetzlaff J, et al.   The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: Explanation and elaboration. PLoS Med . 2009; 6( 7): e1000100. Google Scholar CrossRef Search ADS PubMed  18. Wells G, Shea B, O’Connell D, et al.   The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-Analyses. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed May 7, 2017. 19. Higgins JPT, Altman DG, Gøtzsche PC, et al.   The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ . 2011; 343: d5928. Google Scholar CrossRef Search ADS PubMed  20. StataCorp. 2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP. https://www.stata.com/support/faqs/resources/citing-software-documentation-faqs/ 21. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials . 1986; 7( 3): 177– 188. Google Scholar CrossRef Search ADS PubMed  22. Nanovskaya TN, Oncken C, Fokina VM, et al.   Bupropion sustained release for pregnant smokers: A randomized, placebo-controlled trial. Am J Obstet Gynecol . 2017; 216( 4): 420.e1– 420.e9. Google Scholar CrossRef Search ADS   23. Stotts AL, Northrup TF, Cinciripini PM, et al.   Randomized, controlled pilot trial of bupropion for pregnant smokers: Challenges and future directions. Am J Perinatol . 2015; 32( 4): 351– 356. Google Scholar PubMed  24. Bérard A, Zhao JP, Sheehy O. Success of smoking cessation interventions during pregnancy. Am J Obstet Gynecol . 2016; 215( 5): 611.e1– 611.e8. Google Scholar CrossRef Search ADS   25. Boshier A, Wilton LV, Shakir SA. Evaluation of the safety of bupropion (Zyban) for smoking cessation from experience gained in general practice use in England in 2000. Eur J Clin Pharmacol . 2003; 59( 10): 767– 773. Google Scholar CrossRef Search ADS PubMed  26. Chan B, Einarson A, Koren G. Effectiveness of bupropion for smoking cessation during pregnancy. J Addict Dis . 2005; 24( 2): 19– 23. Google Scholar CrossRef Search ADS PubMed  27. Chun-Fai-Chan B, Koren G, Fayez I, et al.   Pregnancy outcome of women exposed to bupropion during pregnancy: A prospective comparative study. Am J Obstet Gynecol . 2005; 192( 3): 932– 936. Google Scholar CrossRef Search ADS PubMed  28. Cole JA, Modell JG, Haight BR, Cosmatos IS, Stoler JM, Walker AM. Bupropion in pregnancy and the prevalence of congenital malformations. Pharmacoepidemiol Drug Saf . 2007; 16( 5): 474– 484. Google Scholar CrossRef Search ADS PubMed  29. Einarson A, Choi J, Einarson TR, Koren G. Incidence of major malformations in infants following antidepressant exposure in pregnancy: Results of a large prospective cohort study. Can J Psychiatry . 2009; 54( 4): 242– 246. Google Scholar CrossRef Search ADS PubMed  30. GlaxoSmithKline. The Bupropion Pregnancy Registry: Final Report . 2008. http://pregnancyregistry.gsk.com/documents/bup_report_final_2008.pdf. Accessed May 27, 2017. 31. Palmsten K, Huybrechts KF, Michels KB, et al.   Antidepressant use and risk for preeclampsia. Epidemiology . 2013; 24( 5): 682– 691. Google Scholar CrossRef Search ADS PubMed  32. Harrison-Woolrych M, Paterson H, Tan M. Exposure to the smoking cessation medicine varenicline during pregnancy: A prospective nationwide cohort study. Pharmacoepidemiol Drug Saf . 2013; 22( 10): 1086– 1092. Google Scholar PubMed  33. Olsen M, Petronis KR, Froslev T, et al.   Maternal use of varenicline and risk of congenital malformations. Pharmacoepidemiol Drug Saf . 2015; 24: 244. 34. Richardson JL, Stephens S, Yates LM, et al.   Pregnancy outcomes after maternal varenicline use; analysis of surveillance data collected by the European Network of Teratology Information Services. Reprod Toxicol . 2017; 67: 26– 34. Google Scholar CrossRef Search ADS PubMed  35. Alwan S, Reefhuis J, Botto LD, Rasmussen SA, Correa A, Friedman JM; National Birth Defects Prevention Study. Maternal use of bupropion and risk for congenital heart defects. Am J Obstet Gynecol . 2010; 203( 1): 52.e1– 52.e6. Google Scholar CrossRef Search ADS   36. Louik C, Kerr S, Mitchell AA. First-trimester exposure to bupropion and risk of cardiac malformations. Pharmacoepidemiol Drug Saf . 2014; 23( 10): 1066– 1075. Google Scholar CrossRef Search ADS PubMed  37. Gisslen T, Nathan B, Thompson T, Rao R. Hyperinsulinism associated with gestational exposure to bupropion in a newborn infant. J Pediatr Endocrinol Metab . 2011; 24( 9-10): 819– 822. Google Scholar CrossRef Search ADS PubMed  38. Leventhal K, Byatt N, Lundquist R. Fetal cardiac arrhythmia during bupropion use. Acta Obstet Gynecol Scand . 2010; 89( 7): 980– 981. Google Scholar CrossRef Search ADS PubMed  39. Kaplan YC, Olgac Dündar N, Kasap B, Karadas B. Pregnancy outcome after varenicline exposure in the first trimester. Case Rep Obstet Gynecol . 2014; 2014: 263981. Google Scholar PubMed  40. Reefhuis J, Gilboa SM, Anderka M, et al.  ; National Birth Defects Prevention Study. The National Birth Defects Prevention Study: A review of the methods. Birth Defects Res A Clin Mol Teratol . 2015; 103( 8): 656– 669. Google Scholar CrossRef Search ADS PubMed  41. EESoC. Guidelines for Registration. http://www.eurocat-network.eu/aboutus/datacollection/guidelinesforregistration/guide1_4. Accessed July 06, 2017. 42. EESoC. Prevalence Tables. http://www.eurocat-network.eu/accessprevalencedata/prevalencetables. Accessed July 06, 2017. 43. Program MACD. Metropolitan atlanta congenital defects program: Executive summary. Birth Defects Res A Clin Mol Teratol . 2007; 79( 2): 66– 93. CrossRef Search ADS   44. Donahue SM, Kleinman KP, Gillman MW, Oken E. Trends in birth weight and gestational length among singleton term births in the United States: 1990-2005. Obstet Gynecol . 2010; 115( 2 Pt 1): 357– 364. Google Scholar CrossRef Search ADS PubMed  45. Statistics Canada CVS. Birth Database (CANSIM table 102–4510). http://www5.statcan.gc.ca/cansim/a26?lang=eng&id=1024510. Accessed July 06, 2017. 46. Ellard GA, Johnstone FD, Prescott RJ, Ji-Xian W, Jian-Hua M. Smoking during pregnancy: The dose dependence of birthweight deficits. Br J Obstet Gynaecol . 1996; 103( 8): 806– 813. Google Scholar CrossRef Search ADS PubMed  47. Havard A, Jorm LR, Preen D, et al.   The Smoking MUMS (Maternal Use of Medications and Safety) Study: protocol for a population-based cohort study using linked administrative data. BMJ Open . 2013; 3( 9): e003692. Google Scholar CrossRef Search ADS PubMed  48. Kranzler H. Placebo-controlled Trial of Bupropion for Smoking Cessation in Pregnant Women (BIBS). NLM identifier: NCT02188459. https://clinicaltrials.gov/ct2/show/NCT02188459. Accessed July 06, 2017. 49. Miller H. Bupropion for Smoking Cessation in Pregnancy. NLM identifier: NCT01875172. https://clinicaltrials.gov/ct2/show/NCT01875172. Accessed July 06, 2017. 50. Pfizer. Varenicline Pregnancy Cohort Study. NLM identifier: NCT0 1875172. https://clinicaltrials.gov/ct2/show/NCT01290445. Accessed July 06, 2017. © The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Nicotine and Tobacco ResearchOxford University Press

Published: Mar 22, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off