Attachment 12


            FEASIBILITY OF AN EPIDEMIOLOGIC STUDY OF A COHORT OF
           SUBMARINERS WHO RECEIVED RADIUM IRRADIATION TREATMENTS

     In a retrospective cohort study also called a historical cohort 
study, a cohort of individuals is identified based on characteristics 
of exposure in the past and their subsequent disease experience is 
evaluated to a defined point in time.  A cohort is identified solely 
based on the exposure status independent of disease status.  The 
cohort selected for this study must be one in which a large number of 
individuals have been irradiated with radium.  Also, reasonably 
reliable data on exposure of individual cohort members should be 
available.

     The roster of 15,000 submariners who allegedly received the 
radium treatment is a roster of a self-selected group, most of whom 
have health problems (personal communication with Mr. James Garrity, 
Submarine Survivors Association).  This group of veterans may not be 
appropriate for a cohort study for several reasons.  First, they may 
or may not represent the entire group of veterans who received the 
radium treatments in the 40's and 50's.  Second, they are 
self-selected into the group because of their health problems.  
Therefore, a study of this group of veterans will likely to be biased 
toward an overestimate of specific health risks.  Third, because 
people with more serious health problems would have died already, a 
study of the survivors could underestimate the true risk.  As 
alternative approach to study veterans who were treated with radium is 
discussed below.

I.   Identification of Study Subjects

     As a way of evaluating the feasibility of identifying a large 
number of Navy veterans who have received irradiation treatment with 
radium in the military during the 1940's and 50's and a group of 
controls, the following procedures are proposed:

     1.   Selection of a study group

          1)   obtain a dozen names and social security numbers (or 
               service numbers) of Navy veterans who contacted the 
               Submarine Survivors Group, Inc.,

          2)   request their military personnel records and medical 
               records from the National Personnel Records Center,

          3)   review their records for evidence of radium irradiation 
               treatments

          4)   for those veterans with documented evidence of 
               treatment, search for their class roster at the New 
               London US Naval Submarine Base,


				  1


          5)   if the class roster is not available, look for the name 
               of the submarine to which he was assigned subsequent to 
               his training,

          6)   obtain microfilmed ship log of selected submarines from 
               the Bureau of Naval Personnel,

          7)   list all submariners by their names, service numbers, 
               and rank,

          8)   search their records at NPRC for evidence of radium 
               treatment,

          9)   calculate the proportion of submariners with documented 
               exposure history (to be used in the sample site 
               estimate),

          10)  repeat the above steps until the desired number of 
               individuals is accumulated.

2.   Selection of a comparison group

     Because of a "health veteran effect," the general US male 
population is not an ideal comparison group.  Veterans, on the 
average, are healthier than the general population for several reasons 
including pre-induction physical screening, requirements to maintain 
certain physical standards, lifestyle changes that are conducive to 
better health, and better access to medical care.  Therefore, a 
veteran group that was susceptible to the same selective factors as 
the study group would be an appropriate comparison group.  Another 
group of submariners who were on active duty during the same period 
but did not receive the radium treatment will be an ideal comparison 
group.  However, it is possible that some submariners received the 
treatment without the event being recorded in his military records.  
Because of this possibility of misclassification of individuals, other 
submariners are not suitable as controls.  The next group of choice 
would be a group of sailors who served on surface ships during the 
1940's and 1950's.  Navy ships that were active during this period can 
be identified and ship logs of these vessels can be obtained from the 
Bureau of Naval Personnel.  Names, service numbers and ranks of an 
appropriate number of sailors could be abstracted.  As an additional 
means of guarding against misclassification false negative) on 
exposure, their records can be searched at the NPRC for evidence of 
receiving the radium treatment and their names can be matched against 
the roster maintained by the Submarine Survivors Group.

				   2



II.  Study Design Consideration

     These two groups (exposed and control) (can be studied for 
morbidity and mortality outcomes.  A mortality follow-up study may be 
the most appropriate study design for several reasons.  The suspected 
health outcomes of concern are cancers resulting from the irradiation 
of nasopharyngeal areas with radium, and most of the cancers of the 
head and neck areas are eventually fatal.  Also, at least three 
decades have elapsed since the treatment, a period which is considered 
sufficient for the manifestation of radiogenic cancer.  Finally, a 
mortality study can be completed quicker and is generally less costly 
than a morbidity study.  If the mortality study results suggest any 
health effects related to the radium treatments, a morbidity study can 
be designed to further evaluate the adverse health outcomes.

Prepared by
Environmental Epidemiology Service (116E)
Veterans Health Administration
(202) 606-5420

________________________ 	_________
Han K. Kang, Dr. P.H.    	  Date