The partial smoking ban in licensed establishments and health inequalities in England: modelling studyBMJ 2005; 331 doi: https://doi.org/10.1136/bmj.38576.467292.EB (Published 01 September 2005) Cite this as: BMJ 2005;331:488
- Alan A Woodall, specialist registrar in public health ()1,
- Emma J Sandbach, health intelligence analyst1,
- Catherine M Woodward, director of public health1,
- Paul Aveyard, senior lecturer2,
- Graham Merrington, environmental health manager3
- 1 Telford and Wrekin Primary Care Trust, Telford TF1 5RY
- 2 Department of Primary Care and General Practice, University of Birmingham, Edgbaston, Birmingham B15 2TT
- 3 Borough of Telford and Wrekin Council, Social Care, Telford TF3 4LE
- Correspondence to: A A Woodall
- Accepted 22 July 2005
The UK government's white paper Choosing Health proposes prohibiting smoking in public places in England, but exempts public houses (pubs) not serving catered food and licensed establishments that require membership.1 However, passive inhalation of smoke at work may cause 600 deaths per year in the United Kingdom and increases morbidity and mortality among bar workers.2 Furthermore, people attempting to quit smoking find that socialising with other smokers makes quitting difficult, and lapses in quitting are more likely in premises where smoking is permitted.3
Concerns exist that exempt establishments are located primarily in deprived areas with the highest smoking prevalence and that a partial ban worsens health inequalities.4 We examined if exempt establishments were located predominantly in deprived areas in the borough of Telford and Wrekin.
Methods and results
We determined the catering status of pubs from regularly updated records of local authority licensing and environmental health. These allowed us to identify premises preparing catered food, those serving manufactured snacks, and those not serving food. Choosing Health proposes prohibiting smoking only in pubs serving catered food; all others can be exempt. All licensed members' clubs may choose exemption, which we assumed for this study.
Postcode mapping of establishments to each super-output area allowed extraction of scores on the index of multiple deprivation scores by using the software MapInfo, version 8.0 (MapInfo Limited, Windsor, UK). The super-output area is a contiguous aggregate of census output areas containing some 1500 people.
We used logistic regression to examine the relation between the proportion of exempt premises and deprivation score for pubs and licensed members' establishments in Telford and Wrekin borough. With the regression coefficients generated, we used median deprivation scores for each English fifth to estimate the probability of a typical English establishment being exempt.
Out of 174 pubs in the borough, 99 (57%) served catered food. Hence, 75 (43%) would be exempt from smoking prohibition. Including all licensed members' establishments in the model showed that 127 (56%) would be exempt. The model predicted that two thirds of English pubs in deprived areas would be exempt, whereas only a quarter would be exempt in affluent areas. Including members' clubs in the model showed that two fifths of establishments in affluent areas and four fifths of establishments in deprived areas would be exempt (table). A significant linear trend between deprivation and probability of exemption existed in all analyses.
What is already known on this topic
Passive smoking is a serious risk to health
Smoking in public places is banned completely in many countries, but the current proposals for England will allow for some establishments licensed to sell alcohol to be exempt from a smoking ban
What this study adds
Most licensed establishments in the poorest areas would be exempt from the workplace ban on smoking, while most in the more affluent areas would be subject to the ban. This is likely to worsen socioeconomic inequalities in health and smoking prevalence
Prohibiting smoking only in pubs that serve catered food and allowing exemptions for other licensed drinking establishments may worsen health inequalities. Choosing Health estimates that only 10-30% of pubs could be smoking (p100), but our data indicate that the proportion of exempt pubs is higher (43%). This is a small study in one borough, so care must be taken extrapolating the findings. However, Telford and Wrekin is similar to England in terms of demographics and socioeconomic profile. Higher exemption rates were also observed in a survey of 29 local authorities, but no data on deprivation were obtained.5 Our results show that people in deprived areas are more likely to live near licensed establishments exempt from legislation to protect them against smoking. It is possible that people from deprived neighbourhoods may visit establishments in affluent areas, whereas those living in affluent neighbourhoods make the reverse journey. It is more likely that the poorest people with the worst health and highest smoking prevalence would be those most likely to be harmed by passive smoking either working in pubs or as customers, and would be those most likely to have their attempt to stop smoking undermined. We urge the UK government to ban smoking in all enclosed public places, similar to the ban proposed in Scotland and enacted in Ireland, to prevent worsening health inequalities.
A statistical appendix in on bmj.com
This article was first posted on bmj.com on 19 August 2005: http://bmj.com/cgi/doi/10.1136/bmj.38576.467292.EB
We thank Andrea Roalfe for statistical advice. AAW dedicates this work to the memory of Thomas Edward Kennedy.
Contributors AAW initiated the study. AAW, PA, and CW contributed to the design of the study. GM provided data from the local authority and assisted with study design. AAW, PA, and ES carried out the statistical analyses. AAW wrote the paper. All authors contributed to interpretation of findings and contributed critical comments to the paper. AAW is the principal guarantor of the paper.
Competing interests None declared.
Ethical approval Ethical approval was not sought as the data is in the public domain and no issues regarding breach of confidentiality were identified.