Re: Researchers should talk to workers Am. J. Ind. Med. 2000. 37:668

Dr. McDonald errs in his reply to my earlier letter "Researchers Should Talk to Workers," stating that the Amandus et al. [1987] study did not rely on McDonald's exposure estimates. Amandus [1987] states "the eight hour TWA fiber‐exposure for jobs was estimated from the short term samples employing a scheme derived by McDonald et al. [1986a,b]." As this sentence indicates, Amandus et al. used Dr. McDonald's formula for the individual dose estimates. While it may be, as Dr. McDonald claims, that the Amandus team decided separately to rely on this exposure estimate, this does not obviate the fact that Amandus et al. did indeed rely on the McDonald et al. exposure scheme. Dr. McDonald's publications make no mention of his "numerous discussions with past and present employees." The only "employee" thanked for assistance in the 1986 study that was the focus of my original letter was Earl D. Lovick, the W.R. Grace Co. plant manager, an accountant by training [McDonald, 1986]. Moreover, in his sworn testimony Lovick said that he reviewed the exposure estimates with top executives at corporate headquarters, none of whom ever worked in the mine or mill. If Dr. McDonald's discussions with employees contributed to his estimates, then he should have cited (and perhaps thanked) them in his original paper. Dr. McDonald's interviews should have revealed a variety of problems with the Company's estimates. One worker, for example, testified that dust samples that were "too heavy to count" were thrown out [Egilman, 2000]. Dr. McDonald's assertion that fiber estimates which included the observations of workers would mean that tremolite was less hazardous than previously thought reveals other problems with his data. For example, Dr. McDonald underreported the occurrence of lung cancer at Libby. He reported 21 cases, but as of 1984, Grace knew of 32 asbestos lung cancers (R.C. Walsh to H.A. Eschenbach, October 18, 1983). The protocol explains why six of those lung cancers were omitted, but that still leaves 26 lung cancers that should have been included in the published papers. Since data on the higher number of lung cancers was available at the time of the study, I am hard‐pressed to understand why this was omitted from the published paper. Neither Dr. McDonald nor the Grace Company has made the raw data from the tremolite studies public. Given the number and seriousness of the flaws in the McDonald et al. study, no conclusions about the toxicity of tremolite can be drawn from the Libby studies.