These findings show a marked reduction in indoor particle levels (as PM2.5
) after the ban on smoking in a sample of pubs in Scotland. Although this work forms part of a wider project looking at the effect of changes in SHS exposure on the health of bar workers, analysis of the health data will be undertaken after a 1 year follow‐up to allow for seasonal factors.14
The SHS concentrations described here will help to inform our personal exposure estimates for our study participants to determine whether there is an exposure–response relationship for certain health parameters.
There was considerable debate in the media before 26 March 2006, suggesting that the ban on smoking in substantially enclosed public places would be difficult or impossible to enforce, particularly in pubs and bars.20
Our results indicate that there has been a high level of compliance with the ban, resulting in significantly reduced SHS levels in the cohort of pubs we have sampled.
Previous evaluation studies have used a range of methods to measure markers representative of SHS levels. Markers used include total inhalable dust,8
All these measurement methods have individual advantages and disadvantages in terms of ease of use, cost, specificity and ability to provide real‐time data as opposed to time‐weighted average levels. Increasingly, PM2.5
is considered a relatively specific marker of SHS concentrations and has the added advantage of being amenable to measurement using portable direct‐reading instrumentation to give real‐time data without the need for costly chemical analyses.
Our study used a real‐time aerosol monitor to measure PM2.5
levels in pubs. Our exposure metric is comparable to respirable suspended particulate (RSP) reported in several other studies. The pre‐ban levels we measured (mean PM2.5
) are broadly comparable with those reported by studies in pubs in north west England15
), New York10
(mean total inhalable dust 262 μg/m3
) and in restaurants in France (mean RSP 188 μg/m3
) and the UK (mean RSP 195 μg/m3
The US EPA and WHO air quality guidance levels for PM2.5
are designed to protect the general population from respiratory and cardiovascular ill‐health effects associated with exposure to fine particulate matter above these levels. It should be noted that the chemical composition of outdoor pollutants will be different from that of the indoor air measured in this study, and that air quality guidance typically refers to 24 h or annual average levels and not the 30 min concentrations that we have measured. Outdoor levels are influenced strongly by emissions from traffic and industrial processes, whereas indoor sources of particulate matter are more likely to be from cigarette smoke and other combustion sources such as cooking and open fires. The UK Committee on the Medical Effects of Air Pollutants suggests that it is not currently feasible to define a satisfactory guideline for indoor concentrations of particles owing to the lack of data linking indoor air pollution and health effects.23
In the absence of guidelines for indoor air quality particulate matter, comparisons with the health‐based US EPA and WHO outdoor guidelines have been used to provide perspectives on the indoor air quality measured in environments where smoking takes place.9,15
Our study measured levels exceeding the US EPA unhealthy level of 65 μg/m3
in over 80% of pre‐ban visits. This reduced to only 6% (n
3) of post‐ban visits. Smoking was observed on only one post‐ban visit. This occurred in an area of the pub reserved for a private function and generated the highest PM2.5
concentration (104 μg/m3
) of all 53 post‐ban visits. There was some indication in the other two examples where the post‐ban concentration exceeded the US EPA 65 μg/m3
unhealthy level that the source of PM2.5
may have been from patrons smoking at the entrance to the pub, with the smoke from their cigarettes drifting into the main area.
Other studies looking at changes in airborne concentrations of SHS markers in the hospitality sector after the introduction of smoke‐free ordinance have shown reductions in the order of 80–95%.9,10,11,17,18
Table 2 provides an overview of the results of several pre/post smoke‐free legislation evaluations carried out around the world. Our work has shown an average reduction in PM2.5
levels of 86% and so demonstrates that the Scottish legislation has achieved reductions in SHS within pubs comparable to those observed in other countries and cities that have implemented similar smoke‐free ordinances.
It is interesting to contrast the results presented in table 2 with those reported from an evaluation of the Finnish partial smoking restrictions requiring improved ventilation and restriction of the serving area available for smoking customers.24
A study of 20 Finnish bars and restaurants measured geometric mean airborne nicotine levels of 7.1 μg/m3
before the introduction of legislation and 7.3 μg/m3
after the measures were put into place. The authors state that the legislative measures introduced in Finland have had little effect in reducing SHS levels in pubs and restaurants.
Our study was restricted by the short time period between inception and the implementation of the legislation and so it was not possible to measure the levels in all pubs at both quiet and busy time periods. However, we believe that the selection of pubs was representative and the fall in levels so large that these data can be extrapolated to all premises adhering to the ban.
It is difficult to determine how representative our sample is of the whole Scottish pub cohort. The pubs visited were selected at random from a database of all central Edinburgh and Aberdeen pubs, together with pubs from small communities in Aberdeenshire and the Borders areas. We did exclude a small number of pubs for logistical reasons and to ensure the safety of the researchers. Most of these excluded pubs were from more out‐lying areas or in postcodes with high deprivation category scores. Edwards et al15
have examined SHS levels in affluent compared with deprived pubs in north‐west England and found that pubs in deprived areas had mean PM2.5
levels that were almost double those in affluent areas (384 vs 187 μg/m3
). Although our selection bias may have led us to measure slightly lower mean SHS levels in the pre‐ban visits, it seems likely that our post‐ban measurements are reasonably representative. Figures from the enforcing authorities suggest that compliance with the legislation is high throughout the community, with over 99.4% of premises found to be free of smoking activity during more than 3900 inspections across Scotland between March and May 2006.25
Hence, we think that the 86% reduction in PM2.5
levels found in our cohort may be a conservative estimate of the reductions achieved across the whole population of Scottish licensed premises.
Our pre‐ban visits were conducted in February and March and our post‐ban visits were completed in May and June. It is possible that there was some seasonal influence on the reductions of PM2.5 we observed. Windows and doors were more likely to be open during the post‐ban visits and so the number of air changes per hour in pubs was likely to have been higher and this would tend to reduce PM2.5 or SHS concentrations.
We noted substantial numbers of customers smoking outside the entrance to pubs during the post‐ban visits. It will be interesting to note whether this practice continues when the weather conditions become less favourable in the winter months. This potential seasonal effect will be considered when we complete the 1‐year follow‐up.
The average PM2.5
concentration outdoors on the days we performed the visits was 7 μg/m3
(range 3–9 μg/m3
) in Aberdeen and 12 μg/m3
(range 5–23 μg/m3
) in Edinburgh. There was little difference in outdoor air pollution levels between the pre and post‐ban visits so the contribution to indoor levels from outdoor particles will have been consistent. Previous work suggests that SHS contributes between 90% and 95% of RSP air pollution in pubs when smoking is permitted and hence we are confident that SHS is likely to have been the dominant PM2.5
source within pubs during pre‐ban visits.9
Although we report a slight reduction in the average number of customers per bar between the pre‐ and post‐ban visits, we note that this reduction was mainly driven by the results from two pubs where the pre‐ban visit coincided with a drinks promotion. We also consider that this apparent reduction may be a result of a methodological problem in that we did not count customers who sat at outdoor tables owing to difficulties in observing these areas in most premises. By the post‐ban visits in May and June, the weather conditions were more favourable and there was an increase in patrons sitting outside many pubs.
Our measurements were carried out discreetly to try to ensure that the behaviour of bar workers or customers was not influenced by our presence. We postulated that knowledge of our activity could have induced bar managers or workers to switch extraction systems on (or off) depending on their attitude to the legislation, and that customers could also change their smoking behaviour depending on how they perceived our presence. We are reasonably confident that our measurement activity went undetected and that our entry did not influence smoking or other activities that could have changed the PM2.5 levels in these pubs.
What this paper adds
- This is the first evaluation of the changes in secondhand smoke (SHS) levels in pubs as a result of the introduction of the Smoking, Health and Social Care (Scotland) Act 2005 restricting smoking in enclosed public spaces.
- This is the largest data set of pre‐ and post‐ban measurements made in pubs where smoke‐free legislation has come into force. The study characterises the SHS concentrations that bar workers and customers were exposed to before the introduction of the smoke‐free legislation and describes the improvements in indoor air quality that have been achieved.
- This work shows that the Scottish measures have led to reductions in fine particulate matter of about 86%, similar to the SHS marker reductions observed in other countries and cities that have implemented similar ordinances.