A-Z site index | Search
| Links
| Contact DH | Help
|
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||
|
Executive summary 1.1 The Department of Health (DH) asked the Committee on the Medical Effects of Air Pollutants (COMEAP) to advise on the extent of effects of air pollutants on health in the United Kingdom (UK), including an estimate of the numbers of people affected. The Committee formed a sub-group which reviewed the available literature and drafted this report. The report has been endorsed by the Committee. 1.2 The terms of reference of the sub-group were to advise on: "the number of people in the UK whose health was affected by exposure to air pollution". 1.3 The sub-group has reviewed in detail the available information, has drawn heavily on the reports published by COMEAP and by the earlier Advisory Group on the Medical Aspects of Air Pollution Episodes (MAAPE), and has reviewed work published since the publication of these earlier reports.
Approach 1.4 A framework for estimating the impact of air pollution on health was defined. This involved identification of the most appropriate risk estimates linking concentrations of pollutants and effects on health, considering the extent of exposure of people in the UK to air pollutants and estimating the effects of this exposure on background rates of the relevant health effects. 1.5 It was decided that estimates of effects would only be provided when there were available:exposure-response relationships (coefficients) which, in the view of the sub-group, could be applied in the UK with reasonable confidence; and, adequate data on the distribution of concentrations of air pollutants across the country which could be combined with data on population to provide estimates of population exposure. These provisos led to the analysis being focused on the population of Great Britain (GB) (excluding Northern Ireland) and, in terms of health effects, on numbers of deaths and numbers of hospital admissions. Studies of the effects of sulphur dioxide, particles, nitrogen dioxide and carbon monoxide have tended to focus on urban areas. Satisfactory data on the concentrations of these pollutants were available for urban areas of GB and thus estimates of effects of these pollutants were limited to these areas. For ozone, data from rural areas were available and thus for this pollutant the analysis was widened to "all GB". 1.6 The estimates of the exposure-response relationships are based on the results of time-series studies. These studies examine the relationship between daily levels of pollution and the risk of adverse health effects, on the same day or subsequent days, adjusting for climate and other factors. Risks have been expressed as percentage change in health effect per unit change in daily pollutant concentration. In subsequently estimating impacts the country was divided into grid squares. For each grid square the effects of air pollutants on individual health outcomes were calculated by multiplying the exposure-response coefficient (derived from time-series studies) by the ambient concentration (using the appropriate averaging time), the background rate for the health outcome considered (eg, deaths per 100,000 population per year) and the population in the grid square. 1.7 It was agreed that sufficient data were available to allow estimates to be made of the effects of ozone, particulate matter and sulphur dioxide on both all causes of deaths and on admissions to hospital for respiratory disorders. For nitrogen dioxide we are much less sure about the reliability of estimates of effects. For this reason the results of the calculations relating to nitrogen dioxide are not included in this Executive Summary but are discussed in Chapter 8. For carbon monoxide there were insufficient data to allow estimates of effects in the UK to be made with acceptable accuracy. 1.8 For reasons explained in the Introduction (Chapter 2) the sub- group has chosen to present data derived from these calculations in terms of the number of deaths or hospital admissions affected by pollutants in the course of the year. Deaths are affected by bringing forward the date of death; unfortunately it is not possible to estimate by how long. It is believed that for hospital admissions, which are not once-only events as are deaths, the available data can be extrapolated to say: air pollution contributes to the causes for the admission to hospital of n people per year (this includes readmissions). Some hospital admissions may be brought forward whilst others may be truly additional. The split between these groups, if any, is unknown. 1.9 With respect to deaths a number of other workers have stated the results of similar calculations in terms of extra events occurring in a given year. We think this form of presentation is misleading because it implies that the events would not have taken place during the given year had it not been for exposure to the air pollutants. There is no certainty that this is true. Both deaths and hospital admissions of the same individuals may well have occurred during the given year without the added effects of exposure to air pollution. It should be stressed that both the deaths and hospital admissions affected are likely to occur in patients with severe pre-existing disease. 1.10 An assumption of causality has been made in making the calculations reported here. This reflects the previous work of COMEAP in which most of the members of the sub-group had been involved. The sub-group's work included a new review of the methodological issues involved in assessing causality and how these might apply in the context of air pollution and health by a professional statistician (Professor M J Campbell). The purpose was to highlight to members the methodological issues in case anything important had previously been overlooked; it was not intended to fully re-review the question of the causality of air pollution. This methodological review, included here as Appendix 1, did not lead the sub-group to revise its judgement: that the associations are causal is accepted as likely.
1.11 The overall results are given in the following tables (see Tables 1.1 and 1.2). The estimates provided in these tables should be read in conjunction with the relevant chapters of the report. The effects of particles and sulphur dioxide are estimated assuming no threshold for the health effects of these pollutants. The main impacts are on the urban population and rural areas are not included in the calculations. For ozone, both urban and rural areas are considered but for the summer months only. The ozone estimates are strongly sensitive to assumptions regarding a threshold. In assuming no threshold of effect it has been accepted that extrapolation beyond the range of the reported data is allowable. This point is arguable and where it has been done our estimates should be interpreted as worst case estimates. A detailed discussion of the subject of thresholds is presented in Chapter 2. 1.12 It is stressed that the effects on mortality have not been fully quantified. Many of the deaths associated with days of higher air pollution are in the elderly and the sick. Episodes of cold weather and epidemics of the common cold hasten the deaths of such people and it seems likely that air pollutants could act in a similar manner, hastening death by a few days or weeks. If this is the major effect, the impact of air pollution episodes on mortality will be relatively small, but we have been unable to establish the extent by which the time of death has been altered. Table 1.1 Numbers of deaths and hospital admissions for respiratory diseases affected per year by PM10* and sulphur dioxide in urban areas of Great Britain
* PM10: particulate matter generally less than 10 µm
in diameter
(a) co-variation of pollutants means that in some instances we do not know which individual pollutant or mixtures of pollutants has caused the recorded effects or whether some additive or synergistic effects have taken place; (b) it follows that a reduction in the concentration of a single pollutant may produce different benefits than predicted by exposure-response relationships based on observational studies.
1.14 In the view of the sub-group and COMEAP, in addition to the effects recorded here, it is likely that long-term exposure to air pollutants also damages health. At present there are insufficient UK data to allow acceptably accurate quantification of these effects and the sub-group was not confident in applying to the UK estimates of exposure-response coefficients from long-term studies undertaken elsewhere. However, if estimates made elsewhere, especially in the USA, do apply in the UK, they suggest that the overall impacts may be substantially greater than those that we have as yet been able to quantify.
1.15 In the view of the sub-group and COMEAP, the results presented in this report provide a compelling case for more research. It is recognised that, to some extent, this recommendation is already being met by the current DH/DETR/MRC research initiative on air pollution and health. However, we feel that research on the following is needed to allow an improvement of the estimates provided in this report: the years of life lost as a result of day to day variations in levels of air pollutants; the impact on health of long-term exposure to current levels of air pollutants; research on groups at special risk, the elderly and especially the chronic sick; and studies of the effects of air pollutants on outcomes other than death and hospital admissions.
|
|
HMSO, 1998 Available from The Stationery Office |
||||||||||||||||||||||||||||||
You
are here: 
Executive summary