House of Commons - Health - Written Evidence

Select Committee on Health Written Evidence

Supplementary memorandum by Dr Richard Edwards Royal College of Physicians (SP 51)

COMMENTARY ON STUDIES CARRIED OUT BY DR ANDREW GEENS OF THE UNIVERSITY OF GLAMORGAN SCHOOL OF TECHNOLOGY

INTRODUCTION

During my recent appearance to give evidence to the House of Commons Health Select Committee investigating Smoking in Public Places. I was asked about the work of Dr Andrew Geens. I replied that I was aware of his work and gave a brief critique.

Given that I had only looked briefly at his work and that was some time ago, I thought it worth carrying out a more thorough investigation and appraisal of his work for the committee. These are the results of my investigations.

GENERAL COMMENTS

I have come across reports from two studies during previous searches of the internet for publications by Dr Geens in this field[4];[5]. These reports were published from Dr Geen's institution. I can no longer find these two reports on the internet. I have also come across an opinion piece in the Building Services Journal co-authored with Max Graham[6], which contains some data. I am uncertain if this is a peer reviewed publication.

I have been unable to locate any other research published on this topic by Dr Geens in the peer-reviewed literature, the usual standard for communicating findings within the scientific community. I have checked with colleagues who have worked in this field for many years and they have also been unable to locate any peer-reviewed publications from this research by Dr Geens. By contrast, almost all the research studies on this subject which were quoted in the Royal College of Physicians report "Going smoke free"[7], are from the peer reviewed literature.

I said in my evidence that I thought that some of this work was funded by the Tobacco Industry. Having checked the published reports, I do not have direct evidence for this, though the source of funding is somewhat obscure. The first report from the Airport Hotel, Manchester was commissioned by "Corporate Responsibility Consulting Ltd" (CRC Ltd). The website for AIR ("Atmosphere Improves Results") the organisation which promotes the Public Places Charter, and campaigns against smoke-free legislation and promotes ventilation solution states: "AIR is managed by Corporate Responsibility Consulting Ltd and receives funding from the Tobacco Manufacturers' Association"[8].

This suggests there is a link between CRC Ltd and the tobacco industry, though the CRC Ltd website does not disclose any details of funders or stakeholders. The source of funding for the Phoenix and Doublet pub study is not declared.

BACKGROUND ABOUT PARTICULATE AIR POLLUTION AND MARKERS OF SECOND-HAND SMOKE POLLUTION

Several markers of second hand smoke (SHS) are available. A commonly used specific marker is vapour phase nicotine (VPN). Other studies have also used a range of particulate markers including tobacco specific markers such as solanesol and less specific markers such as PM2.5[9]. The term "PM2.5"refers to the diameter of the particles—2.5 microns or less. PM2.5 is particularly useful and pertinent as a marker for SHS, as although it is not specific to SHS (PM2.5 is released from the combustion of a variety of fuels for example), it has the following characteristics:

— PM2.5 is released in large amounts in mainstream and sidestream smoke, and hence levels are raised in the presence of cigarette smoking.

— PM2.5 is known as a "respirable particle" and on theoretical grounds is likely to affect health. This is because due to its size and mass it is easily inhaled deep into the respiratory system where it can be locally adsorbed or absorbed.

— PM2.5 can be measured in real time using relatively cheap and portable monitors such as the TSI SifePak.

Study 1: A ventilation strategy study at Airport Hotel, Manchester

The study consisted of monitoring carbon monoxide (CO), carbon dioxide (CO2) and particulate levels on four consecutive days (Monday-Thursday) in December 2002 behind the bar of a hotel close to Manchester Airport. The bar had a dilution ventilation system fitted which was turned off on days 1 and 3, and switched on during days 2 and 4. On days 3 and 4, a policy of no-smoking at the bar was also introduced. Smoking levels were recorded as the numbers of cigarettes smoked per hour. Monitoring was also conducted during a busier period on 23 December and Christmas Eve.

The main findings were that the CO2 and CO levels were lower (CO maximum levels 2-4 ppm) and rose less through the day on the days when the ventilation was switched on compared with the days without ventilation (CO max levels 10-14 ppm).

This study has numerous limitations, some of which are acknowledged by the author:

1. Little detail on the methods used is given. For example, no details of the monitoring equipment are supplied. The method of monitoring the amount of smoking in the bar is not described. Particulate matter results are presented, but the type of particle being monitored is not revealed. The volume of the bar is not given, so the figures for the number of cigarettes smoked per hour are difficult to interpret in terms of smoking density. The specification of the ventilation system is not described (eg air changes per hour).

2. Carbon monoxide is a relatively insensitive indicator of SHS levels. CO2 levels are not related to SHS, but is influenced mainly by occupancy levels in indoor environments.

3. The study was carried over a restricted time period with monitoring on week days from 10.10 to 19.45 each day. This is likely to have been a very quiet period. It is not explained why monitoring did not continue during the busier evening period. The author himself notes:

"The level of smoking during the test period was very light and so the ventilation system has not been monitored under very testing conditions".

4. It is apparent from graph 12 that the rate of smoking was lower on one of the "ventilation on" days. The figures are not given, but can be estimated from the graphs as about 22.5 cigs per hour smoked on average on Monday (day 1, vent off), but only 15.5 per hour on Tuesday (day 2, ventilation on). Figures for days 3 and 4 are both about 21 cigarettes smoked per hour. Therefore, SHS levels would be expected to be lower at least on day 2.

5. On the busier days (23/24 December) where monitoring continued after 8 pm, no counts were recorded and only a very crude estimate of occupancy and smoking rates was available based on ratio of daily takings. During the evening on these days, CO levels rose to around 10ppm, similar to the levels seen on non-ventilated days. No data on particulate levels is given for these busier days. The reason for this is not explained.

In conclusion, due to the lack of information about the methods, inadequacies in the approach, and the particulate data being limited to time periods where smoking and SHS levels were mostly very low, this study provides very little evidence about the effectiveness or otherwise of ventilation in controlling SHS levels in busy pubs and bars where smoking is allowed.

Study 2: A comparative ventilation effectiveness study at the Doublet and Phoenix Public Houses

In this study monitoring with ventilation switched on and off was conducted over parallel 5/6 day time periods in a smoke free pub (The Phoenix) and a non smoke-free pub (The Doublet) in Glasgow. Both pubs had dilution ventilation systems. Real time CO2, CO and PM2.5 levels were measured in the bar serving areas of both pubs, and CO2 and CO in the customer area in the Doublet. Numerous graphs are used to present the monitoring data. Since for the reasons noted above, particulate levels are the most appropriate SHS markers of those measured, I will focus the commentary on this.

The author quotes in this study and in the Building Services Journal article[13] an Health and Safety Executive EH40 occupational standard for respirable particles (8 hour time weighted average) of 4mg/m3 (or 4,000(g/m3, as 1 mg = 1,000g). This is an enormously high level at which the atmosphere would be thick with visible particles. However, this occupational standard is a reference to the following statement in EH40: "The COSHH definition of a substance hazardous to health includes dust of any kind when present at a concentration in air equal to . . . 4 mg/m3 8-hr TWA [time-weighted average] of respirable dust." [14]

The authors do not make clear that this is not a standard for SHS nor for PM2.5, the particle which is measured in their study. Rather, it is a general level for determining appropriate control of exposure of respirable particles for which no health effects are known to exist, except those associated with the effects on the lung due to presence of a large amount of inert particles. Where dusts have their own limit then exposures will need to comply with the appropriate limit and where dust contain components that have their own assigned workplace exposure limits, all relevant limits should be complied with.

The authors fail to note that the HSE has declined to set an occupational limit for SHS exposure, because the level at which health effects are negligible is not known. Similarly, the Chartered Building Service Engineers (CIBSE) guide A3 states that ". . . regardless of the ventilation used, the health risks of ventilation cannot be eliminated". As a complex mixture of 4,000+ substances, with over 50 known carcinogens, and established adverse health effects, SHS can in no way be described as inert particle with limited or no effect on human health. To relate the achievement of the ventilation systems in Geens' studies to the EH40 4mg/m3 occupational limit for respirable particles is therefore impossible to justify.

As noted above, the UK does not yet use PM2.5 indicators for air quality standards, relying on the slightly different PM10 particle level instead. In the US, the Environmental Protection Agency (EPA) health-related standards for ambient air are: 5 g/m3 (ie 0.015 mg/m3) for the annual mean, and 65 g/m3 (ie 0.065 mg/m3) for a 24 hour period. The US EPA Air Quality index incorporates PM2.5 levels as indicators of air quality. Less than 15 g/m3 is described as "good" air quality, 16-40 g/m3 "moderate", 41-65 g/m3 "unhealthy" for sensitive groups, 66-150 g/m3 "unhealthy", 151-250 g/m3 "very unhealthy", and ~ 251 g/m3 "hazardous". Note that the EPA hazardous level (~ 0.251 mg/m3) is way below the 4mg/m3 "standard" quoted by Dr Geens. The author does acknowledge in passing that DEFRA may suggest a PM2.5 annual exposure limits of 40-50 g/m3 (0.04-0.05 mg/m3) for the UK.

Data from the Doublet during opening hours (assuming this to be 12 noon to 12 midnight or just after) with ventilation switched on vary from about 100-1,200 g/m3 on Monday, 100-2,000 g/m3 on Tuesday, 100-1,300 g/m3 on Wednesday, 150-1,700 g/m3 on Thursday, 100-2,300 g/m3 on Friday, and 100-1,500 g/m3 on Saturday. When the pub is closed overnight, levels reduce to close to zero (although due to the scale of the graph, this may be up to 50 g/m3). When the ventilation was switched off on the Friday, levels increased to a peak of 5,500 g/m3.

Levels in the Phoenix varied between 10-25 g/m3 on the lowest day to 80-160 g/m3 on the highest day. Levels differed little between the pub open hours of 12 noon to 12 midnight, and during closure at night.

These results confirm the very high levels of exposure in the Doublet, the pub where smoking is allowed, despite the best efforts of the ventilation system (though levels are even worse when smoking is allowed without ventilation). Every day bar staff and customers were exposed to peak levels of PM2.5 of at least 1,200 g/m3. These peaks are enormously in excess of the US EPA annual and 24 hour air quality standards, and five times or more greater than the level used to define "hazardous" air quality in the US. The rapid fall in levels of PM2.5 overnight, presumably to close those observed in the external ambient air emphasise the inability of the ventilation system to maintain air quality whilst the pub was open and smoking allowed. The results contrast sharply with the far lower levels seen in the smoke-free pub.

Somewhat incredibly, these results are interpreted by the author as demonstrating: "the ability of the ventilation system in the Doublet to limit and control the concentrations of the parameters under consideration"; and "the performance of the ventilation system in dealing with Environmental Tobacco Smoke"; and that the results "confirm that significant improvements in indoor air quality are achievable with simple inexpensive ventilation systems". No comment is made about the much better air quality that is achieved in the smoke-free pub, nor about the very high levels of PM2.5 observed in the pub with smoking and ventilation relative to the US EPA air quality standards and criteria.

RESULTS FROM OTHER STUDIES

The RCP report describes several studies in which particulate levels in pubs have been monitored. Two from the US (both published in the peer reviewed literature) demonstrate high levels in a range of ventilated pubs and other venues where smoking is allowed which were much reduced following the introduction of smoke-free policies.

In the first from New York, the mean PM2.5 level in 14 bars and restaurants where smoking was allowed was 412 g/m3. After introduction of the smoke-free legislation, this reduced to 27 g/m3 two months later, a 90% reduction. Similarly in Delaware, average post-smoke free levels were 9% of pre-legislation levels in eight hospitality establishments.

The sort of levels seen in pubs with smoking in the UK is shown in the two graphs below. These are taken from research underway in the northwest to explore levels of particulates in pubs within deprived and non-deprived areas.

The first graph (figure 1) shows PM2.5 data from a pilot study in which we visited a smoke-free pub followed by two pubs where smoking was allowed in Manchester City Centre with a real time portable monitor. This demonstrates firstly, how particulate levels even in a large city centre next to busy roads are relatively low (about 10-20 g/m3, 0.01 to 0.02 mg/m3). Secondly, levels in the smoke free pub were about the same as in the ambient air (if anything lower). Thirdly, levels increased to 100-350 g/m3, 0.10 to 0.20 mg/m3) in the pubs where smoking was allowed. Levels were still quite modest as this was early evening and the pubs were relatively quiet with few smokers.

Figure 1

PILOT STUDY DATA FOR PUBS AIR QUALITY STUDY IN THE NORTH WEST

The second graph (figure 2) shows data from one night's monitoring in four pubs in a north west town. This shows firstly how particulate levels increase immediately on entering a pub where smoking is allowed. Secondly, it shows again how SHS levels can reach levels of 1,200g/m3—more than five times above the level of PM2.5 described as hazardous in the US Air Quality Standards. Finally, it demonstrates the variability between pubs, with in this example PM2.5 levels far higher in a pub which serves a deprived community, and which would be exempt under the current Public Health White Paper proposals from being smoke free as it does not serve food. We will be collecting more data in the next few weeks, but preliminary data suggests that the highest levels of SHS are in exactly these pubs, and it will be the most heavily exposed staff working in non-food serving pubs in more deprived areas who will not be protected if the current White Paper proposals are introduced.

Figure 2

DATA FROM ONE NIGHT'S MONITORING IN PUBS AIR QUALITY STUDY IN NORTH WEST

CONCLUSION

This appraisal has shown that the studies carried out by Dr Geens have many weaknesses in design and execution, compounded by highly selective presentation and interpretation of the results. These studies do not provide evidence that ventilation can reduce SHS in pubs where smoking is allowed to levels that will protect the health of staff or customers from the adverse health effects of SHS. Detailed review of the studies reveals the opposite. By contrast, studies of the impact of smoke free legislation show that air quality is rapidly improved and that the health of bar staff is improved.

November 2005