Members will recall that this issue was discussed at the last meeting when a draft statement was considered. This statement has now been amended in light of Members' comments, and the updated version is attached for Members' further consideration. Members may also find it helpful to bring the previous paper (COMEAP/2001/2) to the meeting to aid with discussion.

COMEAP Secretariat June 2001

Draft Statement

We have been asked to advise on the use of statistical associations between daily concentrations of particles and admissions to hospital for treatment of cardiovascular diseases in a cost benefit analysis of the UK Air Quality Objective for particles. It is recognised that the evidence for such an association has developed since we last advised on this question in 1998. Not only has the number of relevant epidemiological studies, some from the UK, increased, but a number of mechanistic hypotheses to explain the associations have been advanced. Some support for these hypotheses has been provided by both epidemiological and experimental studies. We take the view that it would be prudent to regard the reported associations as causal.

It is our view that the evidence has developed sufficiently for us to advise that the reported association should be used for sensitivity analyses in consideration of the benefits of reducing concentrations of particles in the UK. In advising that the data might be used for sensitivity analysis in cost-benefit calculations we have deliberately adopted a cautious approach and this advice should be regarded as tentative. We are aware of large European and US studies, the results of which will soon be published, that will shed more light on this issue. Once such evidence has been assessed it may be possible to provide firmer advice. We advise that that the association should be summarised as:

A 10 µg/m3 increase in 24 hour average PM10 is likely to be associated with a 0.6%* increase in all age, all cause cardiovascular admissions. If a confidence interval around this estimate is required we suggest it to be 0.4-0.8%, but uncertainty regarding this confidence interval is greater than that relating to the central estimate. We stress that this is our first attempt to quantify the effect of particles on cardiovascular admissions and that the summary coefficient we have advised should be taken as a first approximation that may need amendment as further studies appear.

The need for more studies in the UK is clear, as is the need for further work to clarify the possible mechanistic basis of the reported effects.

*See Annex 2 (Association between ambient particles and daily admissions for cardiovascular diseases. Systematic review - Report to the Department of Health) in COMEAP/2001/2