|
About the Department
At a glance
About NDPBs
NDPBs
Advisory Bodies
COMEAP |
|
Home
Public Health
Air Pollution
COMEAP Meetings |
|
COMMITTEE ON THE MEDICAL EFFECTS OF AIR POLLUTANTS DRAFT Minutes of the meeting held on Friday 9 February 2001 in Room 149 Richmond House, Whitehall, London
ITEM 1. APOLOGIES FOR ABSENCE AND ANNOUNCEMENTS 1. Apologies for absence had been received from Professor Peter Blain, Professor Jon Ayres, Professor Roy Richards, Professor William MacNee, Dr Anna Gavin, Mr Fintan Hurley and Dr John Pritchard. 2. The Chairman noted that Dr Brian Miller (IOM) was attending the meeting for Item 4, and that Dr Richard Atkinson (St George's Hospital) would be present for Items 5 and 7. He also welcomed Ms Claire Townsend (DH) who had recently joined PH5 in order to work on some issues concerning risk assessment, as well as dealing with some aspects of air pollution work.
3. The minutes were reviewed page by page and the following amendment was agreed: (i) On the Action Sheet it was noted that OSC should be replaced by OST (Office of Science and Technology).
3.1 Consultation Paper on Air Pollution Information Service 4. Members were informed that this paper had been circulated by DETR. COMEAP had contributed to the paper with its discussion of the numerical index and the statements on the health effects expected at the EC alert thresholds. Members were encouraged to send any further comments directly to DETR, the deadline for comments being 10 April 2001. [ACTION: Members] 3.2 Investigating the Health Impacts of Emissions to Air from Local Industry 5. It was noted that in December 2000, DH published guidelines entitled Investigating the Health Impacts of Emissions to Air from Local Industry. These guidelines were issued to district and regional directors of public health, consultants in communicable disease control, consultants in public health medicine, and chief environmental officers. 6. It was noted that despite being placed on the Department's website, there had been much demand for the guidelines, and that it was likely that a reprint of the publication would be ordered in the coming months. 3.3 Aldehydes 7. The COMEAP statement on the health effects of aldehydes in ambient air had been agreed by post since the last meeting, and the final statement was now available on the COMEAP website. 3.4 Prepublication of Material on the Internet 8. The Secretariat commented that replies had been received from several editors of peer reviewed, scientific journals, setting out their views on whether prepublication of material on the Internet was likely to prejudice future publication in their journals. These, including a recent reply from the editor of Thorax, had been circulated to the Committee by post. It was noted that the New England Journal of Medicine, which had previously had the stance that it would not publish material that had been made available on the Internet, was now softening its approach to this issue. Members were also made aware of an editorial in Nature, which addressed the issue in relation to the publishing information about sequencing of genes. Nature was of the view that this did not amount to prior publication in that particular case.
Professor Roy Harrison declared an interest in this item. 9. The Secretariat reminded Members that at the last meeting the Committee had agreed that it remained prudent to regard an association between long-term exposure to particles and mortality as causal and that quantification of the effects should be attempted provided the uncertainties were acknowledged. A range of possible estimates (now expressed per µg/m3 of PM2.5) had, therefore, been provided in COMEAP/2001/1 for the Committee's comments. 10. The paper was presented as a draft statement with a supporting draft report. For such a complex issue, based on specially commissioned work by the Institute of Occupational Medicine (IOM), the Secretariat felt that there was a need for a more detailed report than could be provided in a short statement, and the Committee's view on this was invited. 11. It was commented that the covering paper raised some key questions for discussion which would have a bearing on the preferred range of estimates. These key issues were: (i) a few of the factors examined in the HEI analysis had resulted in adjusted relative risks with lower confidence intervals below 1. The Committee was asked to what extent did this weaken the credibility of the entire effect? (ii) the HEI analysis had suggested that cardiovascular mortality, but not respiratory mortality was affected. This was an unexpected result. (iii) risks were similar in people both aged under 50 and over, with and without pre-existing lung and heart disease or reduced FEV1 and in smokers and nonsmokers. Whilst this did not support a long term exacerbation of disease mechanism, it was recognised that other mechanisms were possible. 12. The Secretariat commented that the paper included an estimate of the life years gains from reducing the acute effects of exposure to particles, assuming that the loss of life expectancy was between 2 and 6 months. It also assumed that the effects did not differ with age. The Committee's view of this figure for a comparison with the long term effects was requested. An additional option of a 0.1% reduction in hazard rate was added to represent the relative risks in the HEI analysis after adjustment for those confounders that markedly reduced the central estimate below the previous lower confidence interval. The COMEAP subgroup had decided not to select a specific relative risk from the HEI reanalysis, but this same reduction in hazard rate could be derived by using the relative risk of 1.03 adjustment for sulphur dioxide. However, it was pointed out that this was not statistically significant. 13. The paper also included some rough comparisons with other risks to put the estimates into context. It was suggested that the Committee might wish to comment on the validity of these comparisons. It was noted that they suggested that the size of the effect was less than that for active smoking and similar to the effect of passive smoking on heart disease. The covering paper also looked at what sort of reduction in life expectancy might be expected from the reductions in black smoke observed over the past 30 years. This exceeded that actually seen, but the estimate was based on several questionable assumptions. 14. The Secretariat concluded that whatever estimates the Committee considered as best, it was important that it was made clear how the uncertainties compared with that for other air pollution health effects, and the Committee was asked to consider whether the figures should only be used for sensitivity analyses. 15. It was noted that sulphur dioxide could be looked upon as a confounding factor. This pollutant therefore might have as much importance as particles, and it was remarked that the statement and report should make clear that it was concentrating on the effects of particles on health. However, it was acknowledged that if an analysis of "pollutant cocktails" rather than individual pollutants were being considered then this would not be so important. Some Members felt that the HEI reanalysis did not change the overall conclusions, but did increase the levels of uncertainty surrounding the conclusions. The comment was made that it seemed that effects of particles on health varied according to educational status and area in which individuals resided. The latter was thought to be important - and the question of whether the UK was more similar to the Eastern or Western areas of the US was raised. 16. Members thought that the precision of risk estimates was slightly weakened by additional confounding factors. However, it was agreed that the HEI reanalysis had substantiated the original associations for the effects of particles on health. 17. Members went on to discuss the effects of particles on cardivoascular mortality. It was noted that the HEI reanalysis had concluded that cardiovascular mortality, but not respiratory mortality was affected by exposure to particles. The remark was made that an a priori hypothesis would have expected an effect on respiratory mortality. The fact that cardiovascular and respiratory diseases coexisted was raised. It was commented that it was often difficult to distinguish between these diseases, and Members were of the opinion that effects on total mortality was a more helpful and reliable endpoint. Some Members advised that caution should be used with regard to placing reliance on cause-specific mortality even in a study involving large numbers of people. However, others considered that misclassification would widen the confidence intervals but would not affect the relative risk value. There was some further discussion about diagnostic misclassification of cause of death on death certificates, and a recent paper published in "Circulation" was mentioned. Professor Poole-Wilson undertook to circulate this study to the Committee. [ACTION: Professor Poole-Wilson] 18. A further point was raised in relation to what was put down as the underlying cause of death on a death certificate. This has been looked at in an early study of death certificates versus clinical hospital records in Scotland, when it had been concluded that there was a reasonable correlation between the two. It was suggested that the main issue was not whether there was misclassification of individual cases but whether respiratory effects were systematically misclassified in more polluted cities. 19. Members went on to discuss a point about relative risks in different groups. It was noted that if the relative risks for smokers and nonsmokers were the same, then smokers could still be more affected by exposure to air pollution than nonsmokers since the baseline rate of mortality, and thus the absolute risk was higher in smokers. However, it was observed that deaths in smokers and nonsmokers were usually due to very different causes, and that the same size of disparity might not be true for differences in lung function. Some Members were surprised at the effects seen in people actively smoking, but other noted that the particles in cigarette smoke were not the same as ambient particles, ambient particles being a very different type of chemical species. A point was also raised about the rate of clearance of particles from the respiratory tract; it was thought that air pollutants might accumulate more readily in nonsmokers. 20. There followed some discussion of the table S1 on page 6 of the draft report. Some comments had been submitted by members prior to the meeting. A range of estimates of effects had been provided in this table. Of those who had commented, most felt that the total life years gained (millions) was at least as large as the value of 0.007-0.02. Several Members felt that the reduction in mortality rate was likely to be between 0.3-0.6%, but others felt that the figure of 0.1% was a more useful intermediate estimate. It was suggested that the Committee's report should not be pinned to a specific estimate, and that rather than saying that the estimates had been "informed by" the HEI reanalysis, a better form of words would be "taken as among the lowest estimates from the HEI reanalysis". It was remarked that the results were consistent with that from the approach of Bart Ostro, and it was agreed that the report should make this point. 21. Some Members did not agree with the statement in paragraph 41 of the draft report which referred to the losses of life expectancy exceeding a few days and up to at least a month. It was felt that the most recent evidence did suggest that losses of life expectancy could be more than one month. After some further discussion of these points it was agreed that the estimates in Table S1 should be recirculated including all the above comments, for Members final agreement. [ACTION: Secretariat] 22. Members thought that paragraphs 51 and 52 of the draft report which put the risks into context with those of smoking and passive smoking were helpful. There was also some discussion as to whether there was any need to say anything about exposure to black smoke in the report. It was suggested that a more useful comparison could be made using 1990-2000 data rather than data from the 1960s. 23. The Committee proceeded to discuss the draft statement. There was some concern that pressure groups might use the statement in relation to pollutant emissions from industrial sources, and it was emphasised that the statement should make clear that the data currently available did not support this, and that the Committee should say that it did not know whether it was applicable to point sources. It was also pointed out that emissions from industrial sources might be very different chemical species from pollutants in ambient air in cities. 24. Members felt that the statements made in paragraphs 8 and 9 of the draft seemed overcautious and needed to be more positive. More was now known about the long term effects of air pollutants and this should be made clear. It was also suggested that the first sentence of paragraph 8 should be rephrased since there was now enough consensus to say that the associations between health effects and exposure to air pollutants were more likely than not. It was remarked that the report did not use all the UK data on morbidity that were available, and specific mention was made of an old study by Scarlett et al and the effects of air pollution on chronic disease. Although there was no overall relationship, seen in this study it was felt that it should be mentioned in the report for completeness. Professor Ross Anderson undertook to send a note of the references to the Secretariat. [ACTION: Professor Anderson, Secretariat] 25. There was some discussion as to the timing of publication of the statement and report. It was felt that this should be mid-March, and after further discussion, it was agreed that the Secretariat would redraft the statement and report as soon as possible in order to finalise by post.
26. Members were reminded that a preliminary discussion of this issue had been held at the previous meeting (November 3, 2000), when it was noted that epidemiological and toxicological/mechanistic evidence linking exposure to air pollutants and effects on the cardiovascular system had been accumulating for some time. The Secretariat directed Members to COMEAP/2001/2, which provided a summary of some of this evidence. The Secretariat went on to suggest that the evidence was not yet sufficiently well developed to allow Members to come to as firm a view on the possible effects of exposure to particles on cardiovascular admissions as they had reached with regard to respiratory admissions and a possible range of uncertainty as outlined in paragraph 17 of COMEAP/20001/2 (see website: www.doh.gov.uk/comeap/index.htm (items currently under consideration). Members were requested to consider these options and also to consider a draft statement, attached at Annex 3 of COMEAP/2001/2, the wording of which would necessarily change depending on the views of the Committee on the four options set out in paragraph 17. 27. Attention was drawn to a paper by researchers at St George's Hospital Medical School, which was attached at Annex 2. This report summarised the results of time series studies of cardiovascular admissions and particle measures included in a systematic review of the literature undertaken at St George's Hospital Medical School. The authors explained that papers for inclusion were identified using routine searches of online databases such as Medline, and that potential time series studies were first reviewed using their abstracts before a more detailed assessment using the full paper. Studies were finally classified as suitable for inclusion after examination against a number of criteria, which included the provision of regression coefficient, appropriate control for seasonality and duration of study period. Studies excluded from the database included those rejected on statistical grounds, reviews of the time series literature and studies focusing on methodology. 28. The authors noted that studies varied in their use of ICD codes to categorise cardiovascular disease outcomes and also varied in the pollutant lag(s) analysed and reported. Attention was drawn to tables 1-3 which listed the results of the analyses. It was commented that the most common cardiovascular disease groups studied were all-cardiovascular diseases and cardiac disease. Two studies of London data had been included and finally, it was noted that the effects of PM10 on cardiovascular admissions had shown a trend towards higher effect estimates in North America and Canada than in Europe. 29. Members proceeded to discuss the issue of whether it was necessary to separate specific cardiovascular endpoints from general cardiorespiratory disease categories. It was pointed out that classification of cause-specific hospital admissions was very variable and that diagnosis of actual cause of death was also difficult. It was felt that an opinion from clinicians on this issue might be useful. Professor Poole-Wilson undertook to circulate a list of appropriate clinicians to the Secretariat with a view to discussing the matter further. [ACTION: Professor Poole-Wilson] 30. It was commented that the NMMAPS study had not been included in the meta-analysis paper, and the authors commented that this had been a matter of the timing of publication of the NMMAPS report (too late for inclusion) rather than exclusion of this large study. Members asked whether the NMMAPS analysis had shown heterogeneity of results. Whilst it was not clear whether a formal test for heterogeneity had been undertaken, there had been variation in individual estimates, although not all of these estimates had been statistically significant. Further comments were made about the difficulties of diagnosis and classification of diseases, eg, some studies included stroke and others did not. Members also commented that the general public tended to be more worried about the risk of experiencing a heart attack as a result of exposure to elevated levels of air pollutants, and felt that it was important for the Committee to say whether this question (in addition to whether there was a relationship with heart disease in general) could be answered given the present evidence. The Secretariat acknowledged that if a group of experts could be convened to discuss the issues here, it would greatly help in reconciling some of the problems being encountered, but commented that attempts to gather together such a group had, so far, met with little success. 31. The Committee went on to discuss whether it might be appropriate to support the use of a coefficient for the effect of air pollutants on cardiovascular disease, and more particularly whether that meta-analysis derived summary coefficient proposed in the St George's Hospital Medical School paper (0.6% increase in all age, all cardiovascular admissions) could be supported. Some Members felt that careful explanation of how this value had been arrived at was needed since this value was not a simple average but a weighted average. Other Members commented that the studies in London and Los Angeles had determined this value, but others were of the opinion that choosing a quantitative estimate from these time series studies was difficult, and that there were many uncertainties around the value of 0.6% which should be reflected in an appropriate statement from the Committee. It was noted that it might be helpful to look at cardiovascular admissions as a proportion of total admissions since other factors (such as availability of beds) could influence admission rates. The importance of the effects of pre-existing heart conditions was emphasised and some Members suggested that the importance of this should be weighed against other factors. Attention was drawn to a paper by Professor Raymond Agius which had looked at relative risk of hospital admissions for cardiovascular disorders in people with a previous admission. Another study by Goldberg et al, which had looked at hospital admissions in Montreal using data from the health care system was also mentioned, and it was agreed that these papers should be considered further. [ACTION: Secretariat] 32. Members proceeded to discuss
the four options outlined in paragraph 17 of COMEAP/2001/2. Overall, it
was felt that there was most support for the conclusion that with regard
to the effects of particles on cardiovascular admissions "it would
be prudent to regard the associations as causal". Some members felt
that the draft statement did not fully reflect the concerns of the Committee.
There was some discussion of whether all of the effect could be
regarded as causal. In addition, it was considered that if only studies
using two pollutant models were used, there would be insufficient studies
to come to a conclusion. The use of single pollutant models would need
to be considered as one of the components of uncertainty. However, it
was agreed that for the present time a coefficient of 0.6% could be used
for sensitivity analysis. It was also noted that more results from APHEA
were expected in addition to the NMAPPS results. After some further discussion,
it was agreed that the draft statement should be amended in light of Members'
comments and discussed further at the next meeting of the Committee in
June. ITEM 6. SEMINAR ON THE EFFECTS OF AIR POLLUTANTS ON CARDIOVASCULAR DISORDERS 33. The Secretariat informed Members that it was intended that a group should be convened in the near future and a seminar be held to discuss the effects of air pollutants on cardiovascular disorders. The Secretariat said that further details of the seminar would be circulated in the next few weeks. [ACTION: Secretariat]
34. The Secretariat commented that maintaining an up-to-date and detailed grasp of the effects of air pollutants so that expert advice could be provided to DH was a key function of COMEAP. It was noted that Members had discussed this on several occasions with the result that an informal subgroup had met to consider the issues involved. This had led to a tangible result: the setting up of an unique database at St George's Hospital Medical School of time-series studies of the effects of air pollutants on health. This database could now be seen to be providing information to the Committee: an example was in COMEAP/2001/2 on admissions to hospital for treatment of cardiovascular diseases. 35. The Secretariat went on to say that in 1999, a series of draft papers on the mechanisms of action of the common air pollutants and on volunteer studies of their effects had been prepared. However, it was recognised that these were becoming dated; in addition, changing priorities had made continued updating impossible. It was also noted that, for example, updating of the COMEAP 1995 report on particles would not only involve a considerable amount of further work, but that much of that work would be a duplication of the work undertaken by the US EPA, and thus probably a poor use of the limited resources available at DH.
[ACTION: Secretariat, Members, Dr Atkinson]
37. The Secretariat said that the numbers of people applying for the various new positions on the Committee had slightly increased since the paper COMEAP/2001/4 had been circulated; it was further noted that a letter had been sent out to Members to explain that it was open to them to apply for the post of Chairman irrespective of their years of service as Members, and all Members were encouraged to do so.
38. The following papers were circulated for information: 9.1 Use of COMEAP Website 39. At the last meeting, Members requested some information on how often the COMEAP website was visited during the year. The Secretariat obtained some data from colleagues in Information Services Division which listed the number of times the website was visited on a weekly basis. This is circulated as COMEAP/2001/7 for information. The data were also broken down in terms of the number of times each particular webpage was visited. The Secretariat has asked for more up-to-date information in light of the fact that several new sections have been added to the website since these data were collated. This can be made available to Members at a later date if they so wish.
40. This was circulated primarily for Members' information. Members were asked to inform the Secretariat if there were any inaccuracies in the data presented. [ACTION: Members]
10.1 Register of Members' Interests 41. It was noted that the Secretariat had an obligation to udpate the register of Members' interests on an annual basis. The Secretariat, therefore, requested that Members check and the relevant section and send any amendments to Miss Cumberlidge as soon as possible. [ACTION: Members] ACTION SHEET
Published by the
Department of Health
|