UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                            WASHINGTON, D-C, 20460
EPA-SAB-RAC-LTR-92-003

January 9, 1992
                                                                 OFFICE OP
TT     , ,  ,17.,,.    rr n  -11                                     THE ADMINISTRATOR
Honorable William K. Reilly
Administrator
U.S. Environmental Protection Agency
401 M Street, SW
Washington, D.C.  20460

      Re;   Revised Radon Eisk Estimates and Associated Uncertainties

Dear Mr.  Reilly:

      With this letter we are responding to the request of the Office of Radiation
Programs (ORP) dated April 15, 1991 for review of two documents supporting the
ORP reassessment of radon associated risks to the general population.  This
review is based on the  documents entitled "EPA's National Residential  Radon
Survey Preliminary Results" and "Proposed Revisions in EPA Estimates of Radon-
Risks and Associated Uncertainties." These documents were provided to the
Radiation Advisory Committee in April, and the Committee was briefed on them
by ORP staff at the committee meeting in Montgomery, Alabama, on May 20,
1991.

      These documents represent an important step forward by the Agency in
assessing  the health risks of exposure to radon and its decay products,  in review-
ing and utilizing the recent scientific results and deliberations on issues affecting
the dose and risk, and  in attempting to quantify the attendant uncertainties.  The
National Residential Radon Survey is a success story, since it represents the first
significant nationwide survey of an indoor air pollutant with a firm statistical
basis for its design and implementation.  The original design of this survey was
brought to the Board for review in October 1987 [SAB-RAC-88-002],  The Agency
adopted many of the suggestions made by the Board, and has provided  periodic
status reports on its progress.  The initial analyses have been completed. It
appears the  survey will provide a wealth of important data on annual average
radon concentrations and on housing stock characteristics.

      In its letter of transmittal, ORP requested that the Board address seven
issues.  These questions and the Board responses are detailed below. Overall, the
Board has found that the methods and analyses used by the Agency for the
assessment of radon risk are generally appropriate.  The nationwide average of
radon concentrations in U.S. homes, based on the National Residential  Radon
Survey, represents the  best available data. The Agency's  proposed adjustment  of
the exposure/dose relationship between miners and the general population,
obtained from the recent EPA-sponsored National Research Council report,
                                                                     PriiUtd on Riti'.

-------
"Comparative Dosimetry of Radon in Miners and Homes," (National Academy
Press, Washington, DC, 1991) is appropriate as is the proposed value for the
residential radon- equilibrium fraction.  The average home occupancy factor is   -
reasonable and the Agency's method for deriving an estimate of uncertainty
associated with the number of annual radon-related lung cancers is deemed
acceptable,

      The Board has some reservations and comments regarding the dose-response
model used for the estimation of overall risk.  Notwithstanding the SAB's earlier
recommendation [SAB-RAC-88-042] that an average of the relative risk models
presented in the National Research Council's 1990 report, "Health  Effects of
Radon and Other Internally Deposited Alpha-Emitters-BEIR IV," and in  the
International Commission on Radiological Protection report ICRP-50 be used for
making this estimate, the sole use of the BEIE IV model is now recommended
with no adjustment for a potential increased risk for individuals exposed  as
children and young adults, OEP staff concurs with this recommendation. The
effect of implementing this recommendation would be to lower the  central estimate
of the overall risk.  The possible variation of risk with age it exposure could be
incorporated in the uncertainty analysis.

      The Agency's estimate of overall U.S. risk of lung cancer from radon
exposure is calculated for the total population  consisting of both smokers and non-
smokers.  Since smokers are significantly more sensitive to the  effects of radon
exposure than are non-smokers and former smokers, the Board recommends that
the Agency review periodically the national lung cancer estimates in view of
changes in smoking habits as well as changes in the underlying lung cancer
incidence rate.  Detailed discussion of the various issues is provided below.

      a.    Is the methodology  which will be  used to derive the nationwide
            average of radon concentrations in U.S. homes appropriate?

      Yes. The Board had previously reviewed and provided input to the design
for the National Eesidential Radon Survey, Based on the preliminary results, the
Survey appears to have met its original objectives.  The initial survey result  of use
to the task of estimating the nationwide risks  associated with radon exposure is
the average radon concentration  in U.S. homes* Since data were collected from
multiple locations  in each house, four methods of describing the average radon
concentration for a particular  house were presented.  The Board feels that the
method selected by QRP, that of averaging overall frequently occupied levels, is a
reasonable approach.

-------
      b.     Given the finding to date of higher risk to atomic bomb survivors ex-
            posed as children, is it prudent for EPA to continue to follow the
            previous SAB advice to average the risk estimates from BEIR IV and
            ICRP 50 to reflect a potential higher risk to children?

      No,  The earlier Science Advisory Board advice [SAB-RAC-88-042] should be
revised at this time.  The Radiation Advisory Committee (RAC)  unanimously
recommends that the Agency now use only the BEIR IV Model for its risk
assessment for residential exposure to radon.  This recommendation differs from
the Committee's previous advice to average the estimates based  on the BEIR IV
and the ICRP 50 models. This current recommendation is based on the following
rationale:

            1,  The  earlier recommendation that projections of the BEIR IV
model be averaged with those of ICRP 50 was  based, in part, on the assumption
by the International  Commission on Radiological Protection (ICRP) committee that
radon exposures of children and young adults below 20 years of age may produce
a greater risk for cancer of the lung than do exposures of adults.  This  assumption
was based largely on earlier (1950-1978) Life Span Study data on lung cancer from
whole-body exposure to low-LET radiation from the atomic bomb explosions in
Hiroshima and Nagasaki as reported by  Kato and Schull (1982).  However, the
subsequently published (1990)  National Research Council report, "Health Effects of
Exposure to Low Levels of Ionizing Radiatioji-BEIR V," using Life Span Study
data through 1985 concluded that there  was no dependence on age at exposure for
lung cancer,

            A report on lung cancer in  Chinese tin miners exposed to radon gas
(published in Cancer Research, January  1, 1990, by Jay H. Lubia, You-lin Qiao, et.
al.) also failed to show a dependence on  age at exposure for individuals  less  than
13 years  of age.  The recent (1991) National Research  Council analysis "Compara-
tive Dosimetry of Radon in Mines and Homes"  states that "although the K-factors
[the ratio of the exposure-to-dose relationship in homes to that of mines] for
children and  infants  were somewhat greater than those for adults, none of the
values was above unity," The  K-factor values of 0,7  for adults and 0.8 for children
do not support the greater risk to children assumed  in the ICRP 50 model.

            2,  The  BEIR IV model is based on four reasonably well-documented
cohorts of underground miners and represents  the consensus of  a National
Academy of Sciences/National Research Council Committee of well-established and
qualified scientists.

            3.  The  BEIR IV model incorporates a decrease in lung cancer risk
with time after exposure, a phenomenon that was also accounted for in  the 1984
NCRP model and which has been further demonstrated in epidemiological studies
published subsequent to the BEIR IV report. The ICRP 50 model does  not
incorporate this  decrease in risk with time after exposure,
                                      3

-------
      c.     Is the EPA proposed adjustment of the exposure-dose relationship
            between miners and persons In the general population, based on the
            1991 NAS  report, appropriate?

      Yes, the adjustment of the risk coefficient using a K-factor of 0,7 to account
for the estimated lower bronchial dose per unit exposure in homes compared with
that  in mines  is appropriate. The Board notes that recent data on nasal deposi-
tion  of ultra-fine aerosols suggest that the S-factor may decrease to about 0.6.

      d.     Is the proposed EPA value for the residential radon equilibrium ratio
            reasonable?

      The Agency's  evaluation and selection from the available data set of the
value of 0.5 is reasonable.  The uncertainty range of 0.35-0.55 is  appropriate.
This value of the equilibrium factor may turn out to be lower as further data are
collected from U.S. homes.

      e.     Is the proposed occupancy factor reasonable in view of available data?

      Yes, the Agency has selected a value of 75%, with an associated uncertainty
range of 65 to 80%,

      f.     Is BPA*s method for deriving an estimate of uncertainty associated
            with the number of annual radon-related lung cancer deaths appropri-
            ate?

      Yes, although  the Board does recommend that, because smoking has a
significant effect on risk, the Agency attempt to incorporate smoking into the
uncertainty analysis,  The uncertainty analysis conducted by the Agency begins
with a central estimate  of lung cancer deaths based on the BEIR IV risk model
and the results of the residential radon survey.  Following an approach utilized by
The  National Institutes of Health for quantifiable uncertainties, the uncertainty
ranges were distributed log-normally without correlations and the log-transformed
variances were then  added.  The quantified uncertainties included the variability
in risk coefficients derived from  different cohorts, uncertainty about the decrease
in relative risk following cessation of exposure, and uncertainty about the relation-
ship  between exposure and risk in homes versus that in underground mines. The
quantified uncertainties about exposure included the statistical reliability of the

-------
survey, uncertainty about the equilibrium factor, and uncertainty about the percent
of time people spend in their residences.  Variables not treated quantitatively
include estimates of exposure  in mines, the effect of exposure rate on risk and the
effects of smoking status and  gender.  The knowledge base about these uncertain-
ties is not well developed,

      g.     Is EPA's methodology for deriving the central estimate and range of
            radon-related lung cancer deaths in the U.S. population appropriate?

      As has been discussed in addressing the earlier questions, the Board has
found the methods and data to be appropriate, with the exception of the dose-risk
methodology.  The Board now recommends the sole use of the BEIR IV model for
this purpose.  This model has already served as the basis for the uncertainty
analysis presented by ORP.  Although alternative methods or assumptions could
lead to different central estimates, in general these differences are not large
compared with the range of the uncertainty.  The  present analysis, with the use  of
the single BEIE IV model, will yield a solid, well-documented  and defensible
central estimate of the radon-related lung cancer deaths in the U.S. and the
probable range of uncertainties. The possible variation of risk with age may also
be addressed in the uncertainty analysis.

General Recommendations:

      OEP has identified and utilized most of the recent data or assessments on
the factors influencing the risk assessment, including the K-factor and the equilib-
rium factor.  As pointed out briefly in the discussion of these individual issues,
both are topics of (or their values  are influenced by) current research.

      The Agency's estimate of overall U.S. risk of lung cancer from radon
exposure is calculated for the total population consisting of both smokers and  non-
smokers.  It is  well-established that smokers are significantly more sensitive to the
effects of radon exposure than are non-smokers and former smokers.  As noted in
the ORP discussion of uncertainties, since the dose-risk estimate is based on a
relative risk model, changes in the underlying cancer  incidence rate, including the
results of changes in smoking habits, will cause changes in the number of lung
cancer deaths attributable to radon.  The Board recommends that the Agency
review these cancer risk estimates and the underlying data periodically to update
the estimates.  The Board believes the current methodology and the reliance on
BEIR IV are  sufficiently robust, so that these  reassessments would primarily need
to focus on updating the various factors.

-------
      We appreciate the opportunity to review this issue and are ready to provide
review comments on any significant revisions to the subject documents. We look
forward to your response on the major points we have raised.

                                  Sincerely,
                                         Raymond C,
                                         Science Advisory Board
                                         Oddvar F, Nyga^^f Chairman
                                         Radiation Advisory Committee
Enclosure:  FY91 Roster of the Radiation Advisory Committee

cc;    Assistant Adminiitrators
      Director, Office of Radiation Programs
      Director, Office of Drinking Water

-------
               U.S.  ENVIRONMENTAL PROTECTION AGENCY
                     SCIENCE ADVISORY BOARD
                   RADIATION ADVISORY COMMITTEE

                              ROSTER

CHAIRMAM

Dr. Oddvar F, Nygaard
     Division of Biochemical Oncology
     Department of Radiology
     University Hospitals of Cleveland
     2058 Abington Road
     Cleveland, Ohio  44106

MEMBERS

Dr. Kelly H. Clifton
     Department of Human oncology and Radiology
     University of Wisconsin Clinical Cancer center
     K4/33Q, Clinical Science Center
     600 Highland Avenue
     Madison, Wisconsin 53792

Dr. James E. Martin
     Assistant Professor of Radiological Health
     University of Michigan
     School of Public Health
     Ann Arbor, Michigan  48109

Dr. Genevieve M. Matanoski
     Professor of Epidemiology
     The Johns Hopkins University
     School of Hygiene and public Health
     Department of Epidemiology
     624 North Broadway, Room 230
     Baltimore, Maryland  21205

Dr. H.  Robert Meyer
     C.N.S.I.
     750 East ParK Drive
     Suite 200
     Harrisburg, Pennsylvania  17111
Dr. Richard G. Sextro
     Building Ventilation and
       Indoor Air Quality Program
     Lawrence Berkeley Laboratory
     Building 90, Room 3058
     Berkeley, California  94720

-------
Mr. Paul G. Voilleque
     MJP Risk Assessment, Inc.
     Historic Federal Building
     591 Park Avenue
     Idaho Falls, Idaho  83405-0430

DESIGNATED FEDERAL .OFFICIAL

Mrs. Kathleen W. Conway
     Science Advisory Board
     U.S, Environmental Protection Agency
     401 M Street, S.W., A-101P
     Washington, D.C. 20460

STAFF SECRETARY

Mrs, Dorothy M. Clark
     Secretary, Science Advisory Board
     U.S. Environmental Protection Agency
     401 M Street, s.w,, A-101F
     Washington, D.C.  20460

-------