Impact of Smoke-free Legislation on Early-life Mortality and Low Birth Weight in England

Impact of Smoke-free Legislation on Early-life Mortality and Low Birth Weight in England: a Quasi-experimental Study

The purpose of this study is to investigate whether there has been a change in low birth weight and perinatal and infant mortality following the July 2007 introduction of a ban on smoking in public places and workplaces in England.

Study Overview

• Status: Completed
• Conditions: Sudden Infant Death; Infant Mortality; Infant, Very Low Birth Weight; Stillbirth; Infant, Low Birth Weight
• Intervention / Treatment: Other: Smoke-free legislation

Detailed Description

Primary research question Has there been a change in the numbers of babies being born with low birth weight or dying in the perinatal or infant period following the 1 July 2007 introduction of a ban on smoking in public places in England?

Study design Retrospective cohort study (using prospective routinely collected health care data)

Study population All singleton births in England between 1 January 1995 and 31 December 2011.

Intervention The intervention under study is the ban on smoking in enclosed public places and the workplace implemented in England overnight on 1 July 2007.

Inclusion and exclusion criteria We will include all registered singleton births in England occurring between 1 January 1995 and 31 December 2011. This is the maximum time period surrounding the ban's introduction for which the required birth data are available through the data source. Data were originally extracted for 1 January 1993 to 31 December 2011. However, postcode was not recorded in 1993-1994, leading to missing values for Index of Multiple Deprivation (IMD) quintile, region, and urbanisation level in this period. As these variables were considered key potential confounders in the primary analyses, a decision was made to restrict the modelling to the time period 1 January 1995-31 December 2011.

International Classification of Disease (ICD) coding changed from version 9 to 10 as of January 2001, leading to an important drop in recorded SIDS cases. Therefore, analyses of SIDS are restricted to the time period 2001-2011.

Babies with chromosomal anomalies will be excluded.

Outcome

The primary outcomes are:

• Low birth weight (birth weight < 2500 grams)
• Stillbirth (intrauterine death from 24+0 weeks gestation)
• Neonatal mortality (death within the first 28 days of life)
• Sudden infant death syndrome (SIDS; death within first year of life with mentioning on the death certificate of ICD-10-U code R95, or R99 with no other specification)

To assess whether smoke-free legislation had a selective impact on certain subgroups of outcomes we furthermore identified a number of secondary outcomes:

• Very low birth weight (birth weight < 1500 grams)
• Early neonatal mortality (death within first week of life)
• Late neonatal mortality (death between 7 and 28 days of life)
• Post-neonatal mortality (death between 28 days and 1 year of life)
• Infant mortality (death within the first year of life)

Data sources Data are obtained via the Office for National Statistics (ONS). All registered stillbirths and livebirths occurring in England between 1 January 1995 and 31 December 2011 are included. These are linked to death certificates for all deaths occurring before the first birthday.

Data extraction and handling Individual perinatal and mortality data are linked by ONS in a single database including the following individual-level covariates: month of birth, year of birth, month of death, year of death, age at death, sex, birth weight, maternal age, maternal marital status, parity, IMD, region, urbanisation level.

The following covariates are categorised for information governance reasons:

• Age at death: early neonatal, late neonatal, post-neonatal
• Birth weight: <1000 grams, 1000-1499 grams, 1500-2499 grams, 2500-3999 grams, ≥4000 grams
• Maternal age: <20 years, 20-24 years, 25-29 years, 30-34 years, 35-39 years, >40 years
• Parity: 0, 1, 2, ≥3
• IMD: quintiles
• Region: 10 regions
• Urbanisation level: urban, rural

Sample size Sample size calculation for time-oriented analyses is complicated given the complexity of the models. We will use national data for the current study, which will - to the best of our knowledge - be the largest evaluation of the impact of smoke-free legislation on perinatal health, both regarding population size and time span. As we use the maximum time span and population available, sample size calculation can in a way be considered redundant.

We are aware of only one published study on smoke-free legislation and early-life mortality (reference 1). Due to design issues it is not possible to involve data from this study for comparison to the current study.

A number of studies have previously assessed the impact of smoke-free legislation on low birth weight. Our proposed approach is best comparable to that performed earlier in Scotland (reference 2). Using Scottish data on 757,795 deliveries occurring between 1996 and 2009, they showed an immediate -9.9% (95%CI -14.2; -5.2) drop in low-birth-weight-babies. Given the longer study period (1993-2011) and the much larger population size (n>10 million) our study can be expected to have sufficient power to detect a similar reduction in low-birth-weight-babies in England, if present.

Statistical analysis

Relevant population characteristics will be described. Logistic regression analysis will be performed to investigate the association between introduction of smoke-free legislation and sudden ('step') and/or gradual ('slope') changes (as appropriate) in the odds of developing each outcome. Analyses will be adjusted for birth weight, sex, maternal age, maternal marital status, parity (secondary analyses only, see below), IMD quintile, region, and urbanisation level. Seasonal patterning and non-linearity of the underlying time trend will be accounted for as appropriate. Final model selection will be based on Akaike's and Bayesian information criteria (AIC and BIC). The denominator for the analyses will differ according to the various outcomes:

• stillbirths: all births in the dataset
• low birth weight, very low birth weight, neonatal mortality, early neonatal mortality, infant mortality: all livebirths in the dataset
• late neonatal mortality: all livebirths in the dataset surviving the early neonatal period
• post-neonatal mortality: all livebirths in the dataset surviving the neonatal period

The primary analyses will be performed on cases with complete data on all covariates. Parity is the only variable that has >10% missing data (approximately 40-50%), as it is only recorded in married women. As parity is not expected to be a key confounder, we will perform the primary analyses without involving parity in the models in order to maximise population size.

Sensitivity analyses To assess possible confounding by parity, sensitivity analyses will be performed that include parity in the model. In a second set of sensitivity analyses, imputation will be performed to investigate the robustness of the findings to missing data. In order to minimise issues regarding multiple testing, sensitivity analyses will be performed for the primary outcomes only.

All analyses will be performed using Stata 12.0.

• Study Type: Observational
• Enrollment (Actual): 10291118

Contacts and Locations

Study Locations

• United Kingdom
  • Midlothian
    • Edinburgh, Midlothian, United Kingdom, EH8 9AG
      • University of Edinburgh

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: No older than 1 year (Child)
• Accepts Healthy Volunteers: Yes
• Genders Eligible for Study: All

• Sampling Method: Probability Sample

Study Population

All singleton births in England. See 'statistical analysis' for specification of base population for each outcome.

Description

Inclusion Criteria:

• singleton birth occurring in England between 1 January 1995 and 31 December 2011

Exclusion Criteria:

• chromosomal anomalies
• stillbirth (applies to all outcomes other than stillbirth)

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
full cohort; All singleton births in England. See 'statistical analysis' for denominator specification for individual outcomes.	Other: Smoke-free legislation; The intervention under study is the smoke-free legislation in England introduced overnight on 1 July 2007. As of this date virtually all enclosed public places and workplaces were by law required to be smoke-free. More detail can be found via the link provided at the end of this protocol.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
low birth weight	birth weight <2,500 grams	1 Jan 1995 - 31 Dec 2011
stillbirth	intrauterine death from 24+0 weeks gestation	1 Jan 1995 - 31 Dec 2011
neonatal mortality	death within the first 28 days of life	1 Jan 1995 - 31 Dec 2011
sudden infant death syndrome (SIDS)	death within first year of life with mentioning on the death certificate of ICD-10-U code R95, or R99 with no other specification	1 Jan 2001 - 31 Dec 2011

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
very low birth weight	birth weight < 1,500 grams	1 Jan 1995 - 31 Dec 2011
early neonatal mortality	death within the first week of life	1 Jan 1995 - 31 Dec 2011
late neonatal mortality	death between 7 and 28 days of life	1 Jan 1995 - 31 Dec 2011
post-neonatal mortality	death between 28 days and 1 year of life	1 Jan 1995 - 31 Dec 2011

Collaborators and Investigators

• Sponsor: University of Edinburgh
• Collaborators: Imperial College London; Maastricht University; University of Glasgow
• Investigators: Principal Investigator: Jasper V Been, MD MPH PhD, University of Edinburgh

Publications and helpful links

General Publications

• Mackay DF, Nelson SM, Haw SJ, Pell JP. Impact of Scotland's smoke-free legislation on pregnancy complications: retrospective cohort study. PLoS Med. 2012;9(3):e1001175. doi: 10.1371/journal.pmed.1001175. Epub 2012 Mar 6.
• Markowitz S. The effectiveness of cigarette regulations in reducing cases of Sudden Infant Death Syndrome. J Health Econ. 2008 Jan;27(1):106-33. doi: 10.1016/j.jhealeco.2007.03.006. Epub 2007 Apr 8.
• Oberg M, Jaakkola MS, Woodward A, Peruga A, Pruss-Ustun A. Worldwide burden of disease from exposure to second-hand smoke: a retrospective analysis of data from 192 countries. Lancet. 2011 Jan 8;377(9760):139-46. doi: 10.1016/S0140-6736(10)61388-8.

Helpful Links

• description of intervention

Study record dates

Study Major Dates

• Study Start: January 1, 1995
• Primary Completion (Actual): December 1, 2011
• Study Completion (Actual): January 1, 2014

Study Registration Dates

• First Submitted: January 16, 2014
• First Submitted That Met QC Criteria: January 16, 2014
• First Posted (Estimate): January 17, 2014

Study Record Updates

• Last Update Posted (Estimate): January 17, 2014
• Last Update Submitted That Met QC Criteria: January 16, 2014
• Last Verified: January 1, 2014

More Information

Terms related to this study

• Keywords: infant; live birth; infant, newborn; second-hand smoke; tobacco smoke pollution; smoke-free; smoke-free policy
• Additional Relevant MeSH Terms: Pathologic Processes; Death, Sudden; Pregnancy Complications; Death; Fetal Death; Infant Death; Sudden Infant Death; Body Weight; Birth Weight; Stillbirth
• Other Study ID Numbers: TRF NR-0166-1.2