What Are the Effects of Varenicline Compared With Nicotine Replacement Therapy on Long Term Smoking Cessation and Clinically Important Outcomes?

Depression Clinical Trial

Clinical Trial Details — Status: Active, not recruiting

Administrative data

• NCT number: NCT02681848
• Other study ID #: 14/49/94
• Status: Active, not recruiting
• Start date: September 1, 2006
• Est. completion date: September 30, 2019

Study information

• Verified date: March 2019
• Source: University of Bristol
• Contact: n/a
• Is FDA regulated: No
• Study type: Observational

Clinical Trial Summary

Introduction: Smoking is a major avoidable cause of ill-health and premature death. Treatments that help patients successfully quit smoking have an important effect on health and life expectancy. Varenicline is a medication that can help smokers successfully quit smoking. However, there are concerns that it may cause adverse effects, such as increase in the occurrence of depression, self-harm and suicide and cardiovascular disease. In this study the investigators aim to examine the effects of varenicline versus other smoking cessation pharmacotherapies on smoking cessation, health service use, all-cause and cause-specific mortality and physical and mental health conditions.

Methods: In this project the investigators will investigate the effects of varenicline compared to nicotine replacement therapies on: (1) long-term smoking cessation and whether these effects differ by area level deprivation; and (2) the following clinically-important outcomes: rate of general practice and hospital attendance; all-cause mortality and death due to diseases of the respiratory system and cardiovascular disease; and a primary care diagnosis of respiratory illness, myocardial infarction or depression and anxiety. The study is based on a cohort of patients prescribed these smoking cessation medications from the Clinical Practice Research Datalink (CPRD). The investigators will use three methods to overcome confounding: multivariable adjusted Cox regression, propensity score matched Cox regression, and instrumental variable regression. The total expected sample size for analysis will be at least 180 000. Follow-up will end with the earliest of either an 'event' or censoring due to the end of registration or death.

Ethics and dissemination: Ethics approval was not required for this study. This project has been approved by the CPRD's Independent Scientific Advisory Committee (ISAC). The investigators will disseminate the findings via publications in international peer-reviewed journals and presentations at international conferences.

Description:

1. Background

Smoking is the major avoidable cause of preventable morbidity and premature mortality in the UK and internationally. Smoking is also the principal cause of health inequalities and is responsible for most of the difference in healthy life-expectancy between the richest and poorest in our society and those with and without mental health problems. It has been estimated that smoking-related illnesses cost the NHS approximately £5bn per year. Varenicline has been shown to be the most clinically effective smoking cessation medicine for short-term abstinence in randomized controlled trials (RCTs). However, there is very little evidence for its long-term effectiveness and impact on clinical outcomes which are relevant to the NHS.

Although RCTs are considered the gold standard for the evaluation of the intended effects of medicines, they are less appropriate for determining unintended or unexpected beneficial or adverse effects for several reasons. Data from RCTs of varenicline typically have less than a year of follow-up, with relatively low power to detect effects on clinically-relevant outcomes, and are not informative about the longer-term consequences of varenicline treatment. Furthermore, these RCTs usually exclude smokers with chronic respiratory or heart disease and psychiatric illnesses, so their findings may not be generalisable to all patients. Observational studies using large primary care databases of electronic health records may be used to address these unanswered questions, as the investigators propose to do here.

The investigators recently investigated the short-term relationships between smoking cessation treatment and suicide, self-harm, depression and all-cause mortality using CPRD data. The investigators found no evidence of an increased risk of suicidal behaviour in patients prescribed varenicline. The investigators were concerned that the conventional observational estimates of effects of varenicline were biased by residual confounding by indication. If patients prescribed varenicline differed from patients prescribed other smoking cessation treatments prior to their first prescription, then the investigators could not be certain whether any differences in outcomes were due to the medication the patients received, or because of the pre-existing differences between the patients. In the investigators previous study the investigators tried to overcome confounding by indication using three statistical approaches: conventional multivariable-adjusted regression, propensity score matched analysis, and an instrumental variable analysis.

The investigators used physicians' prescribing preferences as instrumental variables, as described in Section 1, for the physicians' prescriptions of varenicline or nicotine replacement products. In the conventional observational analysis, after adjustment for a wide range of observable patient characteristics, patients prescribed varenicline had a 51% lower risk of all-cause mortality than those prescribed other nicotine replacement products in the nine months after their first prescription, (hazard ratio=0.49, 95% confidence interval (95%CI): 0.40 to 0.61). However, the investigators found little evidence of a reduction in all-cause mortality using instrumental variable analysis (risk difference per 1,000 prescriptions=0.7, 95% CI: -3.3 to 4.7). This suggests that the conventional multivariable-adjusted regression may have suffered from bias due to unmeasured confounding, as the effect sizes are implausible.

Smoking is strongly patterned by socio-economic status in the UK; in the 2011 Government Health Survey, prevalence of smoking was 31% amongst individuals in routine occupations compared to 10% amongst individuals in higher managerial and professional occupations (ASH). A recent systematic review reported that the NHS stop-smoking services may be helping to reduce inequalities in smoking prevalence by preferentially targeting smokers of lower socio-economic status (SES). These findings are supported by data from primary care records (The Health Improvement Network), which show that between 2008 and 2010, smokers in more deprived groups were more likely to receive smoking cessation interventions. A recent telephone survey in England found no evidence for differences in rates of self-reported smoking cessation medication use or type of medication used by category of SES. However, this survey demonstrated that there were clear differences in success of quit attempts by SES, with rates being lowest in the lowest SES group. More research is urgently required in this area using large clinical databases to further understand the differences in smoking cessation medication effectiveness by SES.

Evidence explaining why this research is needed now:

The US Food and Drug Administration (FDA) has mandated that varenicline carry a black box warning (the agency's strongest safety warning) on its labelling highlighting the increased risk of suicidal ideation and depression in patients prescribed varenicline based on spontaneous reports to the FDA Adverse Events Reporting (FDA AERS) database. These warnings are meant to indicate causal effects of pharmaceuticals. However, there is an increasing number of experimental and observational studies that suggest there is little difference in risk of adverse neuropsychiatric effects of varenicline compared to nicotine replacement products. On October 16th 2014, the FDA will hold a joint meeting of the Psychopharmacologic Drugs Advisory Committee and the Drug Safety and Risk Management Advisory Committee to:

(i) Discuss safety data from observational studies and a meta-analysis of RCTs since the emergence of the original signal of serious neuropsychiatric adverse events with varenicline which led to the black box warning in 2009 and; (ii) Determine whether any changes should be made to the description of this risk in varenicline's product labelling.

This project is a significant extension of previous research undertaken by this group as part of a Medicines and Healthcare products Regulatory Agency (MHRA) grant (SDS 33437) on "Medicines that may promote suicide: pharmacoepidemiological analysis of risk and risk assessment". The focus of this earlier grant was to investigate whether varenicline caused suicide and self-harm. The investigators produced a written report of the findings for the MHRA and presented these findings orally to the Pharmacovigilance Expert Advisory Group (PEAG) of the Commission on Human Medicines. This committee deliberates on potential new signals of risks associated with pharmacotherapies. Two major questions were identified as future research priorities by the PEAG: how should regulators interpret the results of instrumental variable studies when they conflict with conventional multivariable-adjusted observational analysis? Is it possible to give definitive guidance about when instrumental variable analysis is likely to have lower bias? Currently, the epidemiological literature is not sufficiently developed to conclusively do this. The investigators will address this gap in the literature in the methodological components of this proposal. This will significantly further understanding of when instrumental variable analysis should be used to address problems of confounding by indication. Additionally, this proposal will provide evidence regarding the effectiveness of varenicline with respect to long term smoking cessation and clinically relevant outcomes. This will aid regulators determine whether the benefits of varenicline treatment outweigh possible risks. The investigators will report the results of this research to regulators.

2. Aims

1. To investigate the long-term smoking abstinence of patients prescribed either varenicline or nicotine replacement products in the CPRD.

2. What are the causal effects of smoking cessation on:

1. frequency of GP attendance,

2. frequency of all-cause and cause-specific hospitalisation,

3. all-cause and cause-specific mortality,

4. incident diagnosis of respiratory illness,

5. incident diagnosis of myocardial infarction and

6. incident diagnosis of depression or anxiety.

3. To investigate differences in the effectiveness of smoking cessation medications for long-term smoking abstinence by area level deprivation.

4. Design and theoretical/conceptual framework

The investigators will use the CPRD to conduct a cohort study of all patients prescribed varenicline or nicotine replacement products. The study type is "Hypothesis testing". Exposure will be defined as the first prescription of either varenicline or nicotine replacement therapy. The investigators will investigate differences in the following outcomes described above.

For the statistical analysis the investigators will use Cox-regression models adjusted for a range of baseline confounders, propensity score matched Cox-regression, and instrumental variable analyses using physicians' prescribing preferences as instruments for the prescriptions issued.

Power calculations The following power calculations are based on effect sizes and confidence intervals observed in the investigators previous analyses, which had data on 110,000 individuals prescribed either varenicline or nicotine replacement therapy. Based on the rate of 18,000 new prescriptions per year observed in the CPRD from 2006 to 2011, with a further 4 years of follow-up the number of patients prescribed either varenicline or nicotine replacement therapy will have increased by 72,000. Therefore the total expected sample size for analysis will be around 180,000.

In the investigators previous analysis using CPRD data the age- and sex-adjusted hazard ratio for self-harm/suicide for varenicline vs. nicotine replacement therapy at nine months was 0.73 (95% CI: 0.54 to 0.99); after adjusting for possible confounders this became: 0.90 (95% CI: 0.66 to 1.22). A 70% increase in sample size would lead to a reduction of the standard error by a factor of 1.3, reducing the breadth of the above-adjusted confidence interval from 0.56 to 0.43.

Rare outcomes, self-harm and suicide, were used in previous analyses; the investigators will have greater power to explore more common outcome measures within this project. For example, in the previous analysis the nine month age- and sex-adjusted hazard ratio for all-cause mortality nine months after first prescription for varenicline vs. nicotine replacement therapy was 0.43 (95% CI: 0.35 to 0.53); after controlling for possible confounders this became: 0.49 (95% CI: 0.40 to 0.61). A 70% increase in sample size would lead to a reduction of the standard error by a factor of 1.3, reducing the breadth of the above-adjusted confidence interval from 0.21 to 0.16.

For the effects of varenicline versus nicotine replacement therapy on all-cause mortality, instrumental variable analysis found a risk difference of 0.7 (95% CI: -3.3 to 4.7) per 1,000 patients treated after nine months. A 70% increase in sample size would narrow the confidence intervals from 8.0 to 6.2.

Data collection and analysis The investigators will use the latest available release of the CPRD. This is because General Practices enrolled with the CPRD send regular tranches of data which are released to researchers throughout the year. This will guarantee that there is the largest possible sample of patients for the analysis.

Data analysis Defining exposures, outcomes and covariates

Exposures First time users of the smoking cessation therapies (varenicline or nicotine replacement therapy) will be defined as people who received at least one prescription of the product after the 1st of September 2006 but with no use of a related product during the 12 months before the index date (the first date on which a prescription was issued). Langley et al. (2010) found the smoking cessation prescription data in the THIN database, which is closely related to the CPRD, to be highly comparable to national dispensing data. The analysis will be limited to the first treatment episode. This will mimic an intention to treat analysis in a RCT. The prescriptions will be defined by the therapy file in the CPRD, which contains a list of all prescriptions issued to patients at the practices. Each therapy record records the date a prescription was issued, the quantity of drug prescribed and the dosage.

To mimic an intention-to-treat analysis in an RCT in the primary analysis patients who are initially prescribed nicotine replacement therapy, but later switch to varenicline will be allocated to nicotine replacement therapy and vice-versa.

Outcomes

Smoking abstinence In the CPRD smoking status is indicated by whether the patient is a current, former or never smoker. As GPs are paid to record smoking status smoking behaviour is robustly recorded in the CPRD. Marston et al. (2014) found that 84% of patients had smoking status recorded within a year of registering at a practice, and that smoking prevalence rates by age were similar in CPRD and the Health Survey of England. Booth et al. (2013) found that the difference in prevalence of smoking estimate between the CPRD and the Health Survey for England was less than 1%, and the mean difference was 0.1% (95% CI: -1.5% to 1.7%). Using unpublished data from CPRD sampled as part of the research reported in Thomas et al. (2013) the investigators found that 74% of patients prescribed smoking cessation medication had a subsequent record indicating smoking status. Of these 66% were indicated as current smokers and 33% as ex-smokers. The investigators will initially define a patient as relapsed if they have any record indicating that the patient is a current smoker after their first prescription of a smoking cessation therapy. The investigators will not be able to determine the smoking status of patients who do not return to the GP. Therefore, the investigators will perform sensitivity analyses to examine whether the assumptions made about the smoking status of individuals who are not observed affect the results. For example, the investigators will conduct a sensitivity analysis to see if the results are altered by assuming that patients with missing data have relapsed, or by assuming that patients with missing outcomes have achieved abstinence.

Service use

The investigators will define service use as the number of visits to GP and hospitals in the 3, 6, 9, 12, 24 and 48 months after first prescription. The investigators will define the GP appointments using the clinical data file of the CPRD. This includes all the diagnoses and symptoms that GPs record about all of their patients. As with the other outcomes, the vast majority of diagnoses and symptoms include the date on which the data were added to the database. The investigators will use these dates to define visits to primary care. The investigators will define the hospital visits outcome using the linked Hospital Episodes Statistics data. This is available for approximately half of the sample. Again these data contain the date on which the event occurred, which the investigators will use to define attendance to secondary care within 3, 6, 9, 12, 24 and 48 months after first prescription.

All-cause and cause-specific mortality

The investigators will define all-cause and cause-specific mortality using the linked Office of National Statistics mortality dataset. These include the date of death and cause of death using ICD-9 codes. The investigators will investigate three specific causes of mortality, 1) diseases of respiratory system (ICD-9=460-519), 2) cardiovascular disease (ICD-9=390-459) and 3) mental disorders (ICD-9=290-319).

Adverse events

The investigators will define the adverse event outcomes using the diagnosis records from the Clinical and Referral files in the CPRD. These files record all the diagnoses that the GPs input into their computer system. Each record in the table is given a diagnosis code based on the Read code categorisation. The investigators will use validated Read code lists, for the three adverse event outcomes, respiratory illnesses, myocardial infarction or depression and anxiety. For eligible patients the investigators will extract all records from the Clinical and Referral Tables that indicate the patient either received a specific diagnosis or were referred for a specific diagnosis. As with the therapy records for prescriptions described above, each Clinical and Referral Record indicates the date the information was inputted into the system. The investigators will use this date to define the date that the diagnosis was made.

Covariates

The investigators will include gender, age in years at time of first prescription, previous psychiatric illness/consultation, previous use of psychotropic medications such as hypnotics, antipsychotics and antidepressants, previous self-harm, measures of alcohol consumption where appropriate mean/median number of GP visits per year, body mass index, socioeconomic position (deprivation score for area or residence) and major chronic illness (including diabetes, cancer, arthritis) using the Charlson index. Relevant Read codes will be identified either by validated code lists or by searching for each of these events in the Read code dictionaries to identify any missing Read codes. Collider bias could occur if the investigators conditioned on events which happened as a result of the prescription the patient was issued. To prevent this bias from affecting the results, the investigators will define each covariate using data inputted prior to the first prescription.

5. Statistical Analysis For investigating the effects of varenicline use on each outcome (long-term smoking cessation, frequency of GP and hospital attendance, all-cause and cause-specific mortality, primary care diagnosis of respiratory illness, myocardial infarction, depression or anxiety), the investigators will report a conventional multivariable-adjusted regression, propensity score regression and instrumental variable analysis.

A. Conventional Cox-regression In the first analysis, a conventional observational analysis, the investigators will estimate hazard ratios of the outcomes using Cox-proportional hazards models and the actual prescriptions issued to the patients. Each patient's date of entry into the cohort will be the date they were first prescribed a smoking cessation therapy. The date of exit for each outcome will be the date on which they first have an event, or are censored due to end of follow-up or death or leaving the practice. The investigators will report these associations adjusted for basic confounders (age and gender), and results adjusted for all measured covariates described above.

B. Propensity score regression In the second analysis the investigators will construct a sample of patients balanced on covariates and risk factors using a propensity score. The investigators will construct propensity scores using a logistic regression of the actual treatment received on the covariates described above. Therefore, each participant's propensity score will be their conditional probability (odds) of receiving varenicline versus nicotine replacement therapy. The investigators will match each patient receiving varenicline to another patient receiving nicotine replacement therapy with the closest propensity score on a ratio of 1:1 using a nearest neighbour algorithm with no replacement, and matching will be restricted to the common support region. Patients outside the common support region are those prescribed varenicline with propensity scores higher than any patient prescribed nicotine replacement therapy and vice versa. The investigators will estimate hazard ratios of the outcomes using the propensity score matched sample using Cox-regressions using the same entry and exit information as the conventional Cox-regression analysis described above.

C. Instrumental variable analysis In the third analysis, the investigators will estimate the effects of smoking cessation therapies on the outcomes using physicians' prescribing preferences as instruments for the prescriptions the GPs issue to their patients. The investigators cannot directly measure the physicians' preferences, therefore the investigators will use the prescriptions they issued to their previous patients as a proxy for their preferences. For example, if the instrument was based on just one previous prescription, physicians who previously prescribed varenicline would categorised as a varenicline prescriber. See Section 2 for further details. As with the investigators' previous studies the investigators will use seven prior prescriptions to improve the strength of the instruments. Using multiple prior prescriptions will maximise power; as with previous studies, the investigators will include multiple prior prescriptions to improve the efficiency of the estimators. The investigators will report risk differences in the outcomes using additive structural mean models estimated via the generalised method of moments.

The investigators will categorise each of the adverse event outcomes as occurring within 3, 6, 9, 12, 24 and 48 months of first prescription. The investigators will do this because methods for conducting survival analysis using instrumental variables are not well developed. The investigators will use Stata 13.1 SE to generate all results. The instrumental variable analysis will be conducted using the ivreg2 command and psmatch2 will be used to construct the propensity score. All standard errors will be estimated using cluster robust standard errors which accounts for clustering of patients within practices.

D. Socio-economic variation in effectiveness of smoking cessation treatments

Area level deprivation will be assigned to each patient using their home address postcode and to each GP practice using the practice postcode. Deprivation levels will be based on the Indices of Multiple Deprivation (IMD) which are available from the Office of National Statistics. IMD statistics are updated every two years. The investigators will use the most recent IMD statistics preceding the date of entry into the study for each patient. Although area level deprivation statistics will only be a proxy for individual level deprivation, these demonstrate the expected associations with smoking prevalence. By using both individual and GP level IMD codes, the investigators will investigate whether the effects of smoking cessation therapies differ by IMD at both the level of GP practice and at the individual level. For each patient, the investigators will investigate treatment compliance by reporting the total number of prescriptions issued after the initial prescription.

The investigators will estimate the effects of smoking cessation therapies within sub-groups defined by IMD level both at the individual and practice level using the three methods described above, multivariable-adjusted Cox regression, propensity score regression and instrumental variable analysis. The cohort of patients will be defined as described above. The investigators will report these associations adjusted for basic confounders (age and gender), and results adjusted for all measured covariates described above. Analyses will account for clustering of patients by GP practice.

E. Sensitivity analyses The investigators will perform sensitivity analyses to examine whether the assumptions made about the smoking status of individuals who are not observed affect the results. For example, the investigators will conduct a sensitivity analysis to see if the results are altered by assuming that patients with missing data have relapsed, or by assuming that patients with missing outcome data have achieved abstinence.

Recruitment information / eligibility

• Status: Active, not recruiting
• Enrollment: 180000
• Est. completion date: September 30, 2019
• Est. primary completion date: March 31, 2014
• Accepts healthy volunteers: No
• Gender: All
• Age group: 18 Years and older

Eligibility

Inclusion Criteria:

Patients:

• With CPRD records aged 18 and over.

• Who were prescribed medicines in BNF category 4.10.2 from 1st September 2006, when varenicline was introduced to the UK, to the present.

• With records that were classified as 'acceptable' by the CPRD from all up to standard practices at least 18 months prior to date of entry of each cohort (1st March 2005).

• Who have data defined as "acceptable" by the CPRD if they meet minimum quality control standards, for example their registration period with their GP is valid. "Up to standard" practices are GP practices defined by the CPRD to be providing data of sufficient quality for research purposes.

Exclusion Criteria:

• Patients who registered at a practice less than 365 days before the first recorded prescription to allow for high quality assessment of baseline data and possible confounders.

• Patients prescribed bupropion in the year before their index prescription will be excluded from the analysis. It is relatively rare for patients to be prescribed both NRT and varenicline on the same day. In the investigators' previous study this only occurred for 0.248% of all prescription events. In the primary analysis for this study the investigators will exclude patients initially prescribed both NRT and varenicline.

Follow-up

• Follow-up will end with the earliest of either an "event" or censoring due to the end of registration or death.

Related Conditions & MeSH terms

• Anxiety
• Cardiovascular Disease
• Cardiovascular Diseases
• Depression
• Infarction
• Myocardial Infarction
• Respiration Disorders
• Respiratory Disease
• Respiratory Tract Diseases
• Smoking Cessation

Intervention

Drug:
• Varenicline
Smoking cessation medication
• Nicotine replacement therapy

Locations

Country	Name	City	State
n/a

Sponsors (1)

Lead Sponsor	Collaborator
University of Bristol

References & Publications (42)

Action on Smoking and Health, Smoking statistics who smokes and how much (2014), (available at http://www.ash.org.uk/files/documents/ASH_106.pdf).

Allender S, Balakrishnan R, Scarborough P, Webster P, Rayner M. The burden of smoking-related ill health in the UK. Tob Control. 2009 Aug;18(4):262-7. doi: 10.1136/tc.2008.026294. Epub 2009 Jun 9. Review. — View Citation

B. R. Kirkwood, J. A. . Sterne, Essential medical statistics (Wiley-Blackwell, 2003).

Booth HP, Prevost AT, Gulliford MC. Validity of smoking prevalence estimates from primary care electronic health records compared with national population survey data for England, 2007 to 2011. Pharmacoepidemiol Drug Saf. 2013 Dec;22(12):1357-61. — View Citation

Brookhart MA, Wang PS, Solomon DH, Schneeweiss S. Evaluating short-term drug effects using a physician-specific prescribing preference as an instrumental variable. Epidemiology. 2006 May;17(3):268-75. — View Citation

Brown T, Platt S, Amos A. Equity impact of European individual-level smoking cessation interventions to reduce smoking in adults: a systematic review. Eur J Public Health. 2014 Aug;24(4):551-6. doi: 10.1093/eurpub/cku065. Epub 2014 Jun 1. Review. — View Citation

C. Baum, M. Schaffer, S. Stillman, IVREG2: Stata module for extended instrumental variables/2SLS and GMM estimation. Stat. Softw. Compon. (2002).

Cahill K, Stevens S, Perera R, Lancaster T. Pharmacological interventions for smoking cessation: an overview and network meta-analysis. Cochrane Database Syst Rev. 2013 May 31;(5):CD009329. doi: 10.1002/14651858.CD009329.pub2. Review. — View Citation

Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373-83. — View Citation

Clarke PS, Windmeijer F. Identification of causal effects on binary outcomes using structural mean models. Biostatistics. 2010 Oct;11(4):756-70. doi: 10.1093/biostatistics/kxq024. Epub 2010 Jun 3. — View Citation

Cole SR, Platt RW, Schisterman EF, Chu H, Westreich D, Richardson D, Poole C. Illustrating bias due to conditioning on a collider. Int J Epidemiol. 2010 Apr;39(2):417-20. doi: 10.1093/ije/dyp334. Epub 2009 Nov 19. — View Citation

D. of Health, Smoking kills: a White Paper on tobacco (1998), (available at http://webarchive.nationalarchives.gov.uk/+/www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_4006684).

Davies NM, Gunnell D, Thomas KH, Metcalfe C, Windmeijer F, Martin RM. Physicians' prescribing preferences were a potential instrument for patients' actual prescriptions of antidepressants. J Clin Epidemiol. 2013 Dec;66(12):1386-96. doi: 10.1016/j.jclinepi.2013.06.008. Epub 2013 Sep 24. — View Citation

Davies NM, Smith GD, Windmeijer F, Martin RM. COX-2 selective nonsteroidal anti-inflammatory drugs and risk of gastrointestinal tract complications and myocardial infarction: an instrumental variable analysis. Epidemiology. 2013 May;24(3):352-62. doi: 10.1097/EDE.0b013e318289e024. — View Citation

Douglas L, Szatkowski L. Socioeconomic variations in access to smoking cessation interventions in UK primary care: insights using the Mosaic classification in a large dataset of primary care records. BMC Public Health. 2013 Jun 5;13:546. doi: 10.1186/1471-2458-13-546. — View Citation

Drug Safety and Availability > FDA Drug Safety Communication: Safety review update of Chantix (varenicline) and risk of neuropsychiatric adverse events, (available at http://www.fda.gov/Drugs/DrugSafety/ucm276737.htm).

E. Leuven, B. Sianesi, PSMATCH2: Stata module to perform full Mahalanobis and propensity score matching, common support graphing, and covariate imbalance testing. Stat. Softw. Compon. (2012) (available at http://ideas.repec.org/c/boc/bocode/s432001.html).

Glynn RJ, Schneeweiss S, Stürmer T. Indications for propensity scores and review of their use in pharmacoepidemiology. Basic Clin Pharmacol Toxicol. 2006 Mar;98(3):253-9. Review. — View Citation

Hernán MA, Alonso A, Logan R, Grodstein F, Michels KB, Willett WC, Manson JE, Robins JM. Observational studies analyzed like randomized experiments: an application to postmenopausal hormone therapy and coronary heart disease. Epidemiology. 2008 Nov;19(6):766-79. doi: 10.1097/EDE.0b013e3181875e61. — View Citation

Hernán MA, Robins JM. Instruments for causal inference: an epidemiologist's dream? Epidemiology. 2006 Jul;17(4):360-72. Erratum in: Epidemiology. 2014 Jan;25(1):164. — View Citation

Herrett E, Thomas SL, Schoonen WM, Smeeth L, Hall AJ. Validation and validity of diagnoses in the General Practice Research Database: a systematic review. Br J Clin Pharmacol. 2010 Jan;69(1):4-14. doi: 10.1111/j.1365-2125.2009.03537.x. Review. — View Citation

Hiscock R, Bauld L, Amos A, Fidler JA, Munafò M. Socioeconomic status and smoking: a review. Ann N Y Acad Sci. 2012 Feb;1248:107-23. doi: 10.1111/j.1749-6632.2011.06202.x. Epub 2011 Nov 17. Review. — View Citation

J. D. Angrist, G. W. Imbens, D. B. Rubin, Identification of causal effects using instrumental variables. J. Am. Stat. Assoc.. 91, 444-455 (1996).

Jha P, Peto R. Global effects of smoking, of quitting, and of taxing tobacco. N Engl J Med. 2014 Jan 2;370(1):60-8. doi: 10.1056/NEJMra1308383. — View Citation

Joffe MM. Confounding by indication: the case of calcium channel blockers. Pharmacoepidemiol Drug Saf. 2000 Jan;9(1):37-41. doi: 10.1002/(SICI)1099-1557(200001/02)9:1<37::AID-PDS471>3.0.CO;2-U. — View Citation

Khan NF, Perera R, Harper S, Rose PW. Adaptation and validation of the Charlson Index for Read/OXMIS coded databases. BMC Fam Pract. 2010 Jan 5;11:1. doi: 10.1186/1471-2296-11-1. — View Citation

Kotz D, West R. Explaining the social gradient in smoking cessation: it's not in the trying, but in the succeeding. Tob Control. 2009 Feb;18(1):43-6. doi: 10.1136/tc.2008.025981. Epub 2008 Oct 20. — View Citation

L. Hansen, K. Singleton, Generalized instrumental variables estimation of nonlinear rational expectations models. Econometrica. 50, 1269-1286 (1982).

Langley TE, Szatkowski L, Gibson J, Huang Y, McNeill A, Coleman T, Lewis S. Validation of The Health Improvement Network (THIN) primary care database for monitoring prescriptions for smoking cessation medications. Pharmacoepidemiol Drug Saf. 2010 Jun;19(6):586-90. doi: 10.1002/pds.1960. — View Citation

Marston L, Carpenter JR, Walters KR, Morris RW, Nazareth I, White IR, Petersen I. Smoker, ex-smoker or non-smoker? The validity of routinely recorded smoking status in UK primary care: a cross-sectional study. BMJ Open. 2014 Apr 23;4(4):e004958. doi: 10.1136/bmjopen-2014-004958. — View Citation

McMahon AD. Approaches to combat with confounding by indication in observational studies of intended drug effects. Pharmacoepidemiol Drug Saf. 2003 Oct-Nov;12(7):551-8. — View Citation

Office of National Statistics, Do smoking rates vary between more and less advantaged areas? (2014), (available at http://www.ons.gov.uk/ons/rel/disability-and-health-measurement/do-smoking-rates-vary-between-more-and-less-advantaged-areas-/2012/sty-smoking-rates.html).

P. R. Rosenbaum, D. B. Rubin, Constructing a Control Group Using Multivariate Matched Sampling Methods That Incorporate the Propensity Score. Am. Stat.. 39, 33 (1985).

P. R. Rosenbaum, D. B. Rubin, The Central Role of the Propensity Score in Observational Studies for Causal Effects. Biometrika. 70, 41 (1983).

P. S. Clarke, F. Windmeijer, Instrumental Variable Estimators for Binary Outcomes. J. Am. Stat. Assoc.. 107, 1638-1652 (2012).

Peto R, Darby S, Deo H, Silcocks P, Whitley E, Doll R. Smoking, smoking cessation, and lung cancer in the UK since 1950: combination of national statistics with two case-control studies. BMJ. 2000 Aug 5;321(7257):323-9. — View Citation

Quint JK, Müllerova H, DiSantostefano RL, Forbes H, Eaton S, Hurst JR, Davis K, Smeeth L. Validation of chronic obstructive pulmonary disease recording in the Clinical Practice Research Datalink (CPRD-GOLD). BMJ Open. 2014 Jul 23;4(7):e005540. doi: 10.1136/bmjopen-2014-005540. — View Citation

Ratschen E, Britton J, McNeill A. The smoking culture in psychiatry: time for change. Br J Psychiatry. 2011 Jan;198(1):6-7. doi: 10.1192/bjp.bp.110.081372. — View Citation

Stata Statistical Software: Release 13. (College Station, TX: StataCorp LP, 2013).

The Pharmacovigilance Expert Advisory Group, Summary report based on the minutes of the Pharmacovigilance Expert Advisory Group meeting held on Wednesday 3 July 2013 (2013), (available at http://www.mhra.gov.uk/home/groups/l-cs-el/documents/committeedocument/con322356.pdf).

Thomas KH, Martin RM, Davies NM, Metcalfe C, Windmeijer F, Gunnell D. Smoking cessation treatment and risk of depression, suicide, and self harm in the Clinical Practice Research Datalink: prospective cohort study. BMJ. 2013 Oct 11;347:f5704. doi: 10.1136/bmj.f5704. — View Citation

Van Staa TP, Abenhaim L, Cooper C, Zhang B, Leufkens HG. The use of a large pharmacoepidemiological database to study exposure to oral corticosteroids and risk of fractures: validation of study population and results. Pharmacoepidemiol Drug Saf. 2000 Sep;9(5):359-66. doi: 10.1002/1099-1557(200009/10)9:5<359::AID-PDS507>3.0.CO;2-E. — View Citation

* Note: There are 42 references in all — Click here to view all references

Outcome

Type	Measure	Description	Time frame	Safety issue
Other	Primary care attendance	Number of visits to primary care after first prescription.	3, 6, 9, 12, 24 and 48 months after first prescription
Other	All-cause and cause specific mortality	All-cause patient mortality after treatment identified via linked Office of National Statistics data.; Cause specific mortality by respiratory disease (ICD-10=J00-J99) or cardiovascular disease (ICD-10=I00-I52)	3, 6, 9, 12, 24 and 48 months after first prescription
Other	Respiratory illness	Incident primary care diagnosis of respiratory illness identified via Read codes.	3, 6, 9, 12, 24 and 48 months after first prescription
Other	Myocardial infarction	Incident primary care diagnosis of myocardial infarction identified via Read codes.	3, 6, 9, 12, 24 and 48 months after first prescription
Other	Depression or anxiety.	Incident primary care diagnosis of depression or anxiety identified via Read codes.	3, 6, 9, 12, 24 and 48 months after first prescription
Other	Smoking abstinence	Number of patients who successfully abstain from smoking after treatment.	3, 6, 9, 12, and 48 months after first prescription
Other	Secondary care attendance	Number of visits to secondary care after first prescription. The investigators will define a patient as relapsed on the day they have their first record indicating that the patient is a current smoker after their first prescription of a smoking cessation therapy.	3, 6, 9, 12, 24 and 48 months after first prescription
Primary	Smoking abstinence	Number of patients who successfully abstain from smoking after treatment.; The investigators will define a patient as relapsed on the day they have their first record indicating that the patient is a current smoker after their first prescription of a smoking cessation therapy.	24 months after first prescription