Psychosocial Function of Accident and Emergency Departments

any cause of death. Indeed, the consultants had not even seen the dosage-response data for any causes other than leukaemia and related diseases. These dosage-response data were subsequently obtained by the threat of a Freedom of Information Act lawsuit. No competent biostatistician or epidemiologist who was analysing this PNS data would have 'countless thousands of permutations available to them'. Lung cancer and leukaemia are among the few radiogenic causes of death that are named in the 1981 Lancet paper. It is well known that there is a IS-year latent period for lung cancer before increased risks are detectable. Finally, lung cancer is one of the few radiogenic causes of death in the PNS data where (after a IS-year latent period) the series size is sufficiently large to have much chance of detecting a radiogenic effect. So it is hardly surprising that the first table that Bross & Driscoll (1981) analysed was the dose-latency table for lung cancer. The point, of course, is that if NIOSH had taken the trouble to do valid statistical dosageresponse analyses (taking account of latent period) for the PNS causes of death, they could hardly have missed the relationship for lung cancer, and they could not have published the misleading report in The Lancet. Indeed, when they later carried out the analyses, they found some other interesting cause-of-death relationships (that have never been reported). Moreover, NIOSH has just completed (December 1982) a draft report on the casecontrol study which was carried out to confirm or deny our lung cancer findings. Not only does this new study confirm our findings (Bross & Driscoll 1981), as well as the original report of excess cancer at PNS that Najarian & Colton published in 1978, but it also contains a major new scientific finding. What we find is a strong synergism with chemicals (asbestos, etc.) at levels of radiation that are considered to be very low. Thus the synergism occurs with lifetime badge dose exposures that are generally less than the S rem which is permitted annually by the Nuclear Regulatory Commission. Unfortunately, once again through inadequate statistical analysis, NIOSH has in their latest draft report again reached a negative conclusion on strongly positive data. British readers should realize that NIOSH reports and those of other US agencies in the interagency task force on low-level radiation hazards should be read as policy statements and not as scientific papers. Sincerely