What Will Happen to the A-Bomb Database? A Pending Proposal
Right now, the A-Bomb Study has reached a crossroad where decisions must be made by RERF and RERF's sponsors on how to preserve the study's legitimacy as a prospective epidemiological inquiry -- with all the normal continuity and safeguards expected of such a study -- and how not to sacrifice it (needlessly) in the quest for a better dosimetry.
We say "needlessly" because this book will show how improved dosimetries can be added to the A-Bomb Study without destroying its continuity and its legitimacy as a first-class prospective inquiry.
Key Role of the 1950-1982 Follow-Up :
The first step in our proposal is to identify the study's valid prospective structure. Chapter 5 demonstrates that there has been substantial continuity of doses, cohorts, and structure of the study during the 1950-1970, 1950-1974, 1950-1978, and 1950-1982 follow-ups (Table 5-A).
Although the "change no input" rule and others were bent in the handling of this database during those follow-ups, the rules were not broken badly enough to nullify the study's credibility as a uniquely valuable prospective inquiry into the effect of ionizing radiation upon human cancer-rates. We reach this conclusion largely because the rule-bending occurred while relatively little of the cancer-outcome was known. Even by late 1982, about two-thirds of the study's population was still alive, and more of the cancer-story lay in the study's future than in its past.
It is clear from Chapter 5 that the 1950-1982 follow-up is currently the last complete follow-up where there is substantial continuity of doses, cohorts, and structure. Therefore, the 1950-1982 follow-up represents the proper base for subsequent follow-ups, if the A-Bomb Study is going to maintain the continuity which is required of all prospective studies.
A Simple Proposal :
Our proposal is that the eight 1950-1982 dose-cohorts of A-bomb survivors (Table 5-A) now be "frozen" and kept intact as the base for all the remaining follow-ups of the Life Span Study. Membership in a cohort would be virtually immutable from 1982 onward, in the normal manner of a prospective study. Thus, there would be no change in any cohort's mean age at the time of bombing, no change in any cohort's initial male-female ratio, and no movement of any cancer-deaths from one cohort to another. Of course, there would also be no further alteration of any cohort's mean T65DR dose.
These eight stable dose-cohorts would be "set in concrete" and observed for health effects for their full remaining lifespans as "constant cohorts." The five existing age-bands would also remain undisturbed. If the study is continued in this manner, no one would be entitled to "raise an eyebrow" skeptically about the objectivity of the T65DR database, or to question its scientific worth due to irregular, unacceptable, or sub-standard handling.
Then, as supplemental information appended to the study, the new DS86 dose-estimates for these same cohorts of persons could be (and should be) provided.
This "dual dosimetry" would permit everyone to benefit from the insights about dosimetry which occurred after the eight cohorts were established. Indeed, the proposal could easily handle the continual "enhancing" of the current DS86 dosimetry which is planned for the future. With this approach, even a possible 24th iteration of DS86 would in no way disrupt the permanent architecture of the T65DR database.
In short, we propose to append the DS86 dosimetry to the existing structure of the A-Bomb Study, without sacrificing the study's T65DR anchor into the world of legitimate prospective research.
Demonstration of the Proposal :
The simplicity of the proposal is demonstrated in Chapters 13, 14, 15, and 16 of this book, where we do analyses in both dosimetries for the same sets of persons. While only the T65DR results have the standing of a prospective study, the supplemental and parallel DS86 results provide a basis on which analysts can properly propose any hypothesis which they regard as justified by the results in DS86.
We repeat a reminder published elsewhere: Under the basic rules of research, the DS86 database can never have legitimacy as an integral part -- much less the only part -- of a prospective inquiry (Go89b).
Everything to Gain, Nothing to Lose :
The "constant-cohort, dual-dosimetry" proposal offers everything to gain, and nothing to lose. By keeping the T65DR dosimetry and database intact, one gains both the scientific credibility which belongs to a well-conducted prospective study, and yet one is able to find out what the results would be with a different dosimetry.
By contrast, one loses everything if the T65DR database is sent to oblivion and if it is replaced by a new dosimetry and a new structure. In effect, the on-going study would be terminated and a new one would begin -- not with a clean slate, but with all the cumulative results through 1982 at hand.
In addition, RERF and NAS have stated the intention to continue altering the study's input indefinitely, even after introducing the DS86 database. Their enthusiasm for an ever-improving dosimetry may have caused them to give insufficient weight to the unwanted consequences of such practices upon the future scientific credibility of the whole effort.
A Pending Decision :
After all the suffering in Hiroshima and Nagasaki, and after all the decades of hard work to extract important biomedical information for humanity from that experience, the A-Bomb Study has now reached the crossroad which we have described in this chapter and the previous one.
No one wants to throw away the chance to continue this uniquely valuable A-Bomb Study as a first-class prospective inquiry. Surely everyone must want to help protect it from being subjected, needlessly and forever, to the criticism that its handling was epidemiologically unacceptable. And skepticism about standards in health research is becoming very common (see Chapter 24, Part 1, "A Distasteful Subject").
Following our correspondence with RERF on this issue in 1988 (see Chapter 5), we opted to leave the matter at rest until we were able to demonstrate our proposal to RERF, NAS, and interested epidemiologists. This book provides the demonstration. Neither we nor anyone else can ever use this "constant-cohort, dual-dosimetry" approach beyond the 1982 follow-up, unless RERF and RERF's sponsors make a favorable decision about our proposal.
The decision may affect more than the A-Bomb Study. Practices affecting a world-famous database are likely also to have international repercussions for the standards of biomedical research in general -- far beyond the single field of ionizing radiation.
For both reasons, we regard the issue as one of the most important topics of this book.