- Prof. John R. Goldsmith, M.D., M.P.H.
Epidemiology and Health Services Evaluation Unit,
Faculty of Health Sciences, Ben Gurion University of the Negev, P.O.B. 653, 84105 Beer-Sheva, Israel
Professional interaction over fifteen years between myself, an epidemiologist, and a lawyer started in 1974, when we were both in Washington, evaluating environmental health problems. The lawyer, recently disappointed with the outcome of a case which hinged on the testimony of an epidemiologist, began a dialogue about the criteria for use of probabilities in the scientific and judicial system. We agreed on the importance of making clear these differences, and he documented them in an article.
These differences can be misused in both legal and scientific procedures, under circumstances in which the failure to demonstrate conventional statistical significance (scientifically) is erroneously interpreted as meaning that preventing exposure would not be a reasonable public health measure.
When the lawyer started his private practice he sought expert epidemiological advice in the case of foreign service workers with cancer who had been exposed to microwave radiation in the US Embassy in Moscow.
The trail then led to a major investigation of health risks of Embassy staff by a leading U.S. epidemiologist. The report of this study was said to be negative but actually had some disturbing findings. The trail took a sharp turn when the lawyer provided me copies of documents, obtained under the Freedom of Information Act, which indicated persistent cover-up and deliberate distortions of views of highly regarded scientists with respect to risks from these exposures. A published report on personnel risks from radar exposure in the U.S. Navy diluted the experience of increased leukemia in an exposed group with the low rates in a less exposed group, bringing down likelihood of a significant result and concluding that no effect occurred.
The ethical issues concern whether a scientist who inadvertently finds this evidence should disclose it, in light of security considerations among other matters. The trail, in this presentation, ends with an application of the legal use of probability in interpreting epidemiological evidence on the central scientific issue, the possible health risks from microwave radiation.
Because I used to be a hiker and earlier was a Forest Service guard and lookout, I use a trail as a simile for scientific processes, based on the experience that one can not usually predict exactly where either a scientific process or a trail in the forest may lead, nor what hazards may have to be faced on the way.
As far as I know, the cases referred to here, or related ones, may still be pending, and in order to avoid the possibility that these comments may be misunderstood as an attempt to influence the hearing process, the lawyer is given a pseudonym, Eric.
Eric said as we left the meeting, "You epidemiologists use evidence differently than we lawyers, but we both use the same probability scales. No wonder we keep disappointing one another." The meeting had been on sharing data and interpretations on the possible health effects of organic compounds in drinking water. Eric had been representing the Federal Council for Environmental Quality, and I was on assignment in 1974 to the U.S. National Cancer Institute in the Office of the Associate Director for Field Studies and Statistics.
We discussed it over lunch, and I learned that his disappointment had been based on his involvement in the Silver Bay case in which taconite waste had been flushed into Lake Superior. Before lunch was over, we had reached agreement that there were legal uses of probabilities and uses in the biological sciences which differed quite widely. We agreed that in future efforts to present scientific evidence on environmental regulation, both groups would be well served if these criteria were better known. For this reason we decided that presentation of a comparative table in a suitable publication would be a good idea. Such a scheme was eventually published, and I felt that I had learned an important lesson (Hills 1976).
Several years passed and I had returned to the California Department of Health and Eric had joined a private law firm. One day he called me and said he needed some help from an epidemiologist, "Could I provide it, he asked ?". "Tell me more, I replied."
The following story came out: He had a client who had worked overseas for the US State Department, and the client's wife had developed breast cancer. He learned, after his rotation back to Washington, that the Moscow embassy building, in which his family and many of the staff members also lived in those years, had been subject to microwave (radar) irradiation by the Russians. His client didn't know much about the radiation, but he did know that of their acquaintances, cancer seemed to be unusually frequent.
I wasn't very hopeful, because of the lack of information as to the nature and intensity of the radiation exposure. I suggested that if his client could provide a reasonably valid estimation of the numbers of persons living in the Embassy apartments and of how many of them had developed cancer over what periods of time, it might be possible to compare the data to what could be expected. If the ratio were high, this would suggest that the residents were at increased risk and perhaps this would motivate the State Department to provide protection and to pay more attention to estimation of the risk.
The client did this, and turned over the data to me and I compared it to the expected numbers based on the U.S. National Cancer Institutes. DHEW (1975) Although I had some reservations about the completeness of the data, it seemed that indeed the incidence of cancer was several times greater (6 to 8 times depending on assumptions) than expected. I told Eric I felt that this would, in ordinary circumstances, justify at least a case-referent epidemiological study, and had begun to discuss how much it might cost and who could possible do it.
The State Department suddenly announced that because of concern over exposures at the Moscow embassy, they were contracting with Professor Abraham Lilienfeld, head of the Epidemiology Department at the Johns Hopkins School of Public Health to do a full scale study of the matter and that he was to proceed at once. I advised Eric that Lilienfeld's reputation was very high, he being one of the best epidemiologists in the world, and that it would be reasonable to await the results of his study, so a few more years went by.
In this interval I had in 1978 moved to Israel, and Eric's client, the State Department employee, developed brain cancer and died, but his wife continued to be Eric's client.
Finally, I heard that the Lilienfeld report was out and was essentially negative, (Lilienfeld et al , 1978). I asked Eric to send me a copy, and it arrived in due time. I spent some hours going through all the tables and text. My first impression was that by presenting such a huge volume of data (106 Tables !) it was intended to show that every possibility had been explored. The study, as had been announced, would compare the health data for those working at the Moscow embassy, whether or not State Department employees, with comparable populations working at other Eastern European embassies. This was presumably because in those years it was true that Moscow duty was bleak and isolated and the climate dreadful, while these attributes occurred to some extent in other Eastern European Embassies as well. This comparison assumes that no irradiation was occurring at the other embassies. (see below) However, I excerpted data on cancer incidence and mortality among the Moscow employees and found that, for some categories, it was elevated compared to expectation, as well as compared to data from other Eastern European embassies. For other cancers, the rates for the other embassies was high as well as that for Moscow employees (though not always by enough to be considered statistically significant , that is with a probability of chance occurrence of less than p = .05) I also suggested that Eric get other medical and epidemiological views, but I encouraged him to use the data excerpted from the Lilienfeld report as a basis for claiming that the cancer of his client had been related to working conditions at the Moscow embassy.
When I next heard from Eric, (May 24, 1979) he sent me a packet of material "we developed" from the State Department files under the Freedom of Information Act. I read it with astonishment. It contained information that:
A. A study was done and reported Sept, 1967 of a group of 43 workers, (37 exposed and 7 not exposed) tested for abnormalities in chromosomes on stimulated division. 20 out of the 37 were above the normal range among the exposed, compared to 2/7 among the non-exposed. In a final report, the scientists urged repeat and follow-up which was clinically indicated for 18 persons, but was not undertaken by the end of the contract period, June 30, 1969.
B. A study of blood counts among exposed persons in Moscow, compared to comparable persons in Washington reported to the State Department on October 7, 1976, showed the statistical comparison significantly different for Moscow subjects in almost every comparison.
C. The ambassador had complained to the Russian authorities of the irradiation in 1967, and apparently knew about the studies mentioned in A., above. Concerning the inquiry of a doctor, Ambassador Thompson wrote.."I do not believe that you should communicate to him any particulars of the results of your investigations to date...."
D. The President of the AFSA (American Foreign Service Association, presumably, an employee group) in his report 6 April, 1976, to the Board on "Radiation, UHF and Electro-magnetic Dangers " alleged that exposures were occurring in other Embassies, that not all of the exposures were due to external causes, that the Department of State was possibly involved in a cover-up, while it had evidence of health damage.
E. That data on exposure and occurrence of some cases of cancer were withheld from Prof. Lilienfeld until the report was complete, and it was too late to include the results,
F. That the views of Prof. Lilienfeld were altered or deleted at the request of the contract officer, according to comparisons between the original and final versions of the report. Notes taken at the meeting document that these changes were made at the request of the contract officer. Lilienfeld himself was hospitalized at the time of the meeting, but was said to have agreed to the changes.
G. That Lilienfeld had urged that follow-up studies be done, since the latency period for some possible types of cancer had not yet been sufficient at the time of his survey.
H. That reviews of the work done by the contract investigators were interpreted by consultants as inconclusive because the State Department had failed to complete the follow-up work recommended by its contractors. (although this reason was not made clear).
To this day, I don't know from which files the data were extracted, who did the selection, for what purpose, and whether additional data would make the apparent impact of the data less damming as evidence of a cover-up. As an epidemiologist I usually try to find out whether evidence is likely to be biassed or whether it is likely to be representative. As to these excerpts, I can never know unless I try to find and examine the original files.
Of course Eric resubmitted the claim July 19, 1979, and added in June 16 1981 that for another client whose wife had had leukemia following work in the Moscow Embassy.
I took no other action feeling that the claims submitted, including information obtained under the Freedom of Information Act, had priority.
I heard no more from Eric, but from time to time I reexamined the file, especially when a proposed Voice of America broadcast facility was planned for the Arava (the portion of the valley of the Jordan and Dead Sea rift valley) on Israel's Eastern border. (I testified against the project, and it has now been cancelled).
In 1990 with the break-up of the Former Soviet Union, I began to think that the publication of the material should be useful in health evaluation of radar or microwave exposures due to military, broadcast or industrial sources. I wrote Eric's firm in June, 1990 to ask what was the outcome of his cases, ostensibly because I had never been paid for my work on them. His widow replied that he had died of cancer February 22, 1990. She knew nothing about his case material.
I still did nothing
In the Spring of 1994, I was asked to contribute a paper to a meeting on "The Biological Effects of Radio-frequency (RF) Radiation" at Skrunda Latvia, the site of a huge microwave broadcast and receiving station for tracking missiles. The paper was to be on epidemiological evidence to be considered in such an evaluation. I again reviewed the file and concluded that evidence was suggestive for four health effects, (a) chromosomal changes, (b) hematological changes, (c) reproductive effects, and (d) increased cancer incidence from the microwave irradiation in Moscow, and all four of these effects had been found independently in other studies as well. I prepared the paper presenting evidence that these four effects might be found for the exposed groups at Skrunda. There were few physicians and epidemiologists in the audience, so although it was agreed to publish proceedings, I undertook to publish , independently, these findings and I did so Goldsmith, (1995)
I provided for the Skrunda symposium a manuscript, "Epidemiological Studies of Radio-frequency Radiation: Current Status and Areas of Concern", which is intended to be published in the Proceedings in "Science of the Total Environment" .
At a meeting on Ethics and Environmental Epidemiology 16 Sept. 1994 at the Annual meeting of the International Society for Environmental Epidemiology I presented " Balancing the Interests of Patients, Science, an Employers: A case Study of RF Exposure."
I was approached by a man whose son served in the military for two years in a naval patrol boat with extensive radar exposure. The year after his discharge, the son developed testicular cancer, for which he received surgical and chemotherapeutical treatment. The father asked me to review the evidence on cancer among military personnel exposed to radar, in order to see if there may be any reason to feel that his exposure during service in the Navy might have been partly responsible for his sons cancer.
The following facts were discerned:
A.The major study following-up Naval personnel exposed to radar during the Korean War compared two groups of men, both of which were apparently exposed, the more exposed one was exposed in testing and repair, and the presumably less exposed one was in day-to-day use of radar equipment. A significant increase in leukemia in the most exposed group was diluted with a group with no increase with leukemia and the combined group had a small, but not significant increase. (Robinette et al., 1980) The abstract reports"No adverse effects....could be attributed to potential microwave exposure..."
B. Dogs used during the Vietnam war had significant increases in testicular cancer.
C. A subsequent study of active duty Naval personnel showed that men with exposure to exhaust fumes had increased rates. The young man , whose father sought my advice had such an exposure as well as to radar. Active duty personnel would be unlikely to show testicular cancer increase, if it required many years of latency.
D. A study of testicular cancer among persons seen at mostly military hospitals in the Washington, D.C. area, showed an excess among Naval personnel but not among men from other service branches.
E. A study of persons exposed to Agent Orange, a defoliant used heavily in Vietnam, showed an excess of testicular cancer among Naval personnel, but not among persons exposed to Agent Orange.
F. Testicular cancer rates were increased among traffic control officers using radar guns which they usually held in the lap and thereby exposed testicular tissue.
Using Eric's diagram, I had little difficulty reaching the conclusion that if a man had Naval service with heavy exposure to exhaust fumes and radar and then developed testicular cancer, the chances were substantially better than 50% that the exposure was causally related to the disease.
The evidence so far indicates that hematological changes, increased spontaneous abortion, mutational changes in the circulating lymphocytes and onset of cancer, including lymphatic and testicular cancers may be occurring among persons with increased military, industrial or occupational exposure to radar or microwaves.
The military importance of such communication-relevant radiation has prevented or delayed a full and objective disclosure of the hazards and therefore adequate protection from them. Compared to the evidence concerning power line exposures and health, that for radar appears to be stronger.
The Ethical Issues:
A. Given that the legal profession and scientists may use probabilities in different ways in coming to conclusions, we must be wary of agencies involved in dissemination of pollution hazards using lack of statistical significance as a substitute for lack of effect, especially when public health protection can often be based on whether an effect is more likely than not.
B. What are the ethical obligations of a scientist who, without intending to, comes into possession of evidence of cover-up of hazards, distortion of evidence about risk from those affected, and misuse of the reputation of scientific peers?
I DO NOT FEEL THAT THE EVIDENCE SO FAR HAS SHOWN A SERIOUS RISK FOR ORDINARY HOUSEHOLD USES OF RADAR, AS IN COOKING.
THERE IS ALREADY AN INDUSTRY-SUPPORTED GROUP STUDYING POSSIBLE CONSEQUENCES OF RADAR OR MICROWAVE EXPOSURES ASSOCIATED WITH THE USE OF CELLULAR TELEPHONES.