[Congressional Record Volume 140, Number 85 (Wednesday, June 29, 1994)]
[Extensions of Remarks]
[Page E]
From the Congressional Record Online through the Government Printing Office [www.gpo.gov]


[Congressional Record: June 29, 1994]
From the Congressional Record Online via GPO Access [wais.access.gpo.gov]

 
                      HUMAN RADIATION EXPERIMENTS

                                 ______


                       HON. JOSEPH P. KENNEDY II

                            of massachusetts

                    in the house of representatives

                         Tuesday, June 28, 1994

  Mr. KENNEDY. Mr. Speaker, the Clinton administration and the 
Department of Energy Secretary Hazel O'Leary, in particular, are to be 
commended for launching their investigation of the Government's role in 
the human radiation experiments which started in the late 1940's. The 
Advisory Committee on Human Radiation Experiments has been established 
to investigate these concerns and consider the need for compensation to 
the victims.
  In my efforts to find an equitable settlement for those unjustly 
exposed to radiation during these experiments, I have been presented 
with a proposed compensation model, ``Calculating Compensation for 
Radiation Victims Based on a Retrospective Probability Analysis,'' by 
Robert Gary, one of the most knowledgeable and experienced litigators 
in the field. Mr. Gary has represented approximately one-third of all 
radiation related cases tried in the United States, including the 
class-action suit of 200,000 service personnel exposed to atomic 
radiation at the Nevada Test Site in the 1950's, and 600,000 people 
from the four counties surrounding Three Mile Island during the nuclear 
accident there.
  From his extensive study of the issue, Mr. Gary has devised the 
following model, which I recommend for consideration by my colleagues 
and members of the Advisory Committee on Human Radiation Experiments.

Calculating Compensation for Radiation Victims Based on a Retrospective 
                          Probability Analysis

                         (By Robert Gary, Esq.)

       Let us examine first the concept of a doubling dose, since 
     that is the way the biological effects of ionizing radiation 
     are currently measured and described. A doubling dose is that 
     amount of radiation which will double the natural incidence 
     of a certain kind of harm in a population. For example, if we 
     have 100 test subjects and the natural incidence of cancer is 
     17%, then if we expose that population to one doubling dose 
     of radiation, instead of getting 17 cancers in that 
     population, over time, we will get 34 cancers in that 
     population over time.
       If the doubling dose is 150 rems of radiation, then one 
     doubling dose will double the baseline figure for the 
     occurrence of the harm to which that doubling dose applies. A 
     second doubling dose would add another increment equal to the 
     baseline figure. For example, 300 rems, two doubling doses, 
     would produce an occurrence level of 34 + 17 = 51, and 450 
     rems, three doubling doses, would produce an occurrence level 
     of 34 + 17 + 17 = 68. If there were no doubling doses, i.e. 
     no special radiation exposure, there would be an occurrence 
     level of 17 which is the natural or background incidence.
       The biological effects of special radiation exposures have 
     traditionally been expressed in terms of the probability of 
     future harms occurring, or the number of future harms that 
     can be expected to occur, as results of a specific known 
     level of special radiation exposure. For example, if 100 mice 
     are exposed to 150 rems of radiation each, in a special 
     exposure, then 34 of them will develop harms consistent with 
     radiogenic origin, instead of the 17 that would have done so 
     without the special radiation exposure. We are looking 
     forward from the point of the known special radiation 
     exposure, and projecting anticipated effects in the future 
     caused by the known special radiation exposure. This is known 
     as a prospective probability analysis since it is looking 
     forward from the point of exposure.
       But the question that Congress has to address is: Is this 
     patient's condition the result of a special radiation 
     exposure he/she is known to have undergone in the past?'' 
     This requires a retrospective probability analysis projecting 
     backward from the known present harm to the known past 
     radiation exposure and asking, ``What is the probability of 
     causal connection?''
       For purposes of the argument presented here, we must know 
     that the patient received a special radiation exposure, and 
     we must know the number of rems received. Given these 
     variables, a retrospective probability analysis will help us 
     to fairly compensate radiation victims.
       We start with a thought experiment. The kind of harm that 
     will be considered is major birth defects. The baseline 
     incidence for these is approximately 10%. Scientific papers 
     including The Biological Effects of Ionizing Radiation 
     Report, (1972 Edition and all subsequent editions, suggests 
     that the doubling dose for major birth defects is 
     approximately 150 rems. If a human population is exposed to 
     150 rems, and each member becomes one of a pair of parents, 
     (assuming the other parent is not part of the exposed 
     population); and if each pair of parents has one child, we 
     would expect 20 children with major birth defects instead of 
     10. In other words the occurrence of the specific harm, major 
     birth defects, is doubled, by the administration of the 
     doubling dose to one parent in each parent pair.
       The child is the one that comes before the Judge or the 
     Congressperson, and the question is, ``What is the 
     probability that the child's birth defect was caused by the 
     special radiation exposure his/her father or mother 
     received?''
       The doubling dose was administered, in this thought 
     experiment, so the incidence went up from 10 to 20. One of 
     those 20 is in the office. The subcategories within the 20, 
     those who would have had a major birth defect without the 
     special exposure of their parent, and those that have the 
     defect because of the special exposure of their parent are 
     completely indistinguishable. They are indistinguishable in 
     principle, unalterably, and because of the laws of quantum 
     physics. But plainly the probability is P=.50 that our 
     patient's major birth defect was caused by the special 
     radiation exposure. This will always be true if exactly the 
     doubling dose is received. Half of the 20 were caused by the 
     special radiation insult, so, ceteris paribus, each of the 20 
     has a 50% chance of being a victim of that special radiation 
     insult.
       Prospectively, right after the exposure, each exposed 
     parent only had a P=.20 probability of having a child with a 
     major birth defect. How can the prospective probability be 
     P=.20 while the retrospective probability is P=.50? The 
     prospective probability is based on just doubling the 
     baseline incidence (10 goes to 20 so P=.20) but the 
     retrospective probability is based on the attributable 
     proportion out of the known injured population (10 out of 20 
     so P=.50).
       Let Pb represent the baseline probability.
       Let Pr represent the part of the total risk as 
     elevated by a special radiation exposure which is 
     attributable to that special radiation exposure.
       Let Pt represent the total risk as elevated by a 
     special radiation exposure.
       It follows that:

     Pt=Pb + Pr

     and

     Pr=Pt-Pb
       In general Pt may be calculated using the following 
     formula: Pt=Pb [(100/D)(x)+100]100.
       Where x = the number of rems in the special radiation 
     exposure, and
       Where D = the doubling does for the kind of harm being 
     considered
       Say the client is a cancer case and was exposed to 75 rems.
       Pb for cancer is about .17 (according to the 
     literature), and doubling dose will be taken as 150 rems for 
     purposes of this calculation. We want to find Pc which 
     we'll call the probability of causal connection. 
     Pc=Pr/Pt.
       So we start out finding Pt: Pt=.17[(100/
     150)(75)]100=.2550
       Now we find Pr:

     Pr=Pt-Pb
     Pr=.2550-.1700=.0850
       Now, Pc or the probability of causal connection, is 
     the ratio between the risk attributable to the special 
     radiation exposure and the total risk after elevation by that 
     special radiation insult.

     Pc=Pr/Pt
     Pc=.0850/.2550=.3333

       So the answer is that this patient has a probability of 
     .3333 that his/her cancer was caused by the special radiation 
     exposure specified for this case.
       A quick table of Pc values might be helpful. All 
     values not included on the table can be calculated using the 
     formula and the method outlined.

------------------------------------------------------------------------
                                x                                   Pc  
------------------------------------------------------------------------
10..............................................................   .0625
50..............................................................   .2500
100.............................................................   .4000
150.............................................................   .5000
300.............................................................   .6667
400.............................................................   .7273
450.............................................................   .7500
------------------------------------------------------------------------

       Intuitively it checks out that the doubling does yield a 
     Pc of .5000, and we can see that two doubling does (300 
     rems) would create an attributable proportion of 2 out of 3 
     parts or .6667, similarly 3 doubling doses would result in an 
     attributable proportion of 3 out of 4 parts of .7500. We 
     don't need the formula for these obvious cases, but the 
     formula is useful for the less obvious cases like 10 rems or 
     400 rems.
       The literature is not always consistent about what the 
     doubling dose is for a particular kind of harm. If we assume 
     150 rems is the doubling dose the expression in the Pt 
     equation is (100/150)x, but if we took 50 rems as our 
     doubling does the Pt equation would have (100/50)x, and 
     if 5 rems were the doubling dose it would be (100/5)x. 
     Similarly if 500 rems were the doubling dose it would be 
     (100/500)x.
       Apart from radiation, there are other toxic agents that 
     have a linear or doubling dose relationship to the harms they 
     cause. These equations will work for all of them and provide 
     compensation guidelines for clients injured by any toxic 
     agent within this broad category which may include Agent 
     Orange, Sarin, and Isomethylcyante.
       In the end we want to convert our Pc value into 
     dollars because that's what the compensation client is asking 
     for. This is done by taking the Fair Jury Value (FJV) of the 
     injury, assuming no question about causation, and simply 
     multiplying it by the probability of causal connection 
     Pc.
       We put proximate cause on a sliding scale. The question is 
     not, ``Is causation more probable than not?'' but rather, 
     ``How probable is causation?'' The more probable causation is 
     the more compensation the alleged victim gets. The darkness 
     that surrounds the causation issue in radiation cases, and 
     which must do so because of the rules of quantum physics, is 
     left unobscured. We will never know, nor can we ever know, 
     who is really a victim of a special radiation insult (unless 
     its an immediately lethal dose). What we are looking for is 
     fairness. We want to provide compensation, but not clean out 
     the Federal Treasury. We want to pay victims, or possible 
     victims, but not provide a windfall to everyone who has been 
     exposed to any amount of any toxic agent. Most important, we 
     want to avoid sending real victims away emptyhanded because 
     they haven't been able to meet the ``more probable than not'' 
     standard of the Restatement of Torts, 2d. The Agency or 
     organization releasing radiation or other toxins should take 
     responsibility for the uncertainties that are inevitably 
     connected with those materials or physical processes. It's 
     not fair to expose people to radiation and then say that the 
     uncertainties which cannot in principle (because of quantum 
     realities) be overcome are a bar to their recovery of 
     damages. The ``more probable than not'' standard would send 
     about half of the legitimate radiation victims away 
     emptyhanded. The releaser of the radiation would get a 
     windfall by not having to pay any of the claims for lesser 
     radiation exposures when it is quite possible that among 
     those claims are real victims that actually get cancer and 
     die.
       It is appropriate to note that Fair Jury Value (FJV) means 
     just that. It's not just medical special damages, but it can 
     include pain and suffering, loss of consortium, loss of 
     earnings, and even an adjustment to compensate for the moral 
     circumstances under which the exposure occurred. An innocent 
     and hapless victim of medical experimentation that violates 
     international law and all peremptory norms of human conduct 
     (i.e. the Nuremburg and Helsinki Accords), might get more in 
     a Fair Jury Value than a similarly injured worker in a 
     nuclear power plant or radiation laboratory. FJV is what fair 
     jury, or administrative panel, would or does award in the 
     state where the claimant makes his/her claim. There 
     legitimately might be federal guidelines for FJV's in cases 
     against the government where the entire compensation scheme 
     arises out of a single piece of Federal legislation.

                          ____________________