Table 1.

Probability distributions developed in previous studies to represent uncertainty in DDREF for solid cancers induced by low-LET radiation.

Reference	Cancer outcomes	DDREF
		90% CI	Median	Mean
NCRP (1997)	Solid cancer mortality	(1.25, 4.1)a	2.3	2.5
U.S. EPA (1999)	Solid cancer mortality or incidence	(1.1, 4.3)b	2.0	2.25
Grogan et al. (2000)	Solid cancer mortality excluding lung	(1.0, 4.1)c	2.3	2.4
	Lung cancer mortality	(1.5, 8.3)d	4.5	4.7
Land et al. (2003)	Solid cancer incidence excluding breast and thyroid	(1.0, 3.0)e	1.5	1.8
	Breast and thyroid cancer incidence	(1.0, 3.0)f	1.5	1.6
BEIR VII report (NRC 2006)	Solid cancer incidenceg	(0.9, 2.5)h	1.5	1.6
Jacob et al. (2009)	Solid cancer mortality or incidence	(0.53, 1.96)i	0.83j	NAk
		(0.33, 0.86)	0.48
		(0.65, 2.43)	1.02
U.S. EPA (2011)		Distribution from BEIR VII report

^aPiecewise-linear distribution with most probable value at 2 and lower and upper bounds of (1, 5) based on assessment of curvature in dose-responses for solid cancer mortality or incidence in LSS cohort assuming linear-quadratic models, comparison of risks of lung cancer mortality in medical patients exposed to x rays with risks in LSS cohort, and data on cancer induction in laboratory animals.

^bUniform distribution at values of 1–2 and exponential distribution at values >2 based on distribution recommended by NCRP (1997) and other data in LSS cohort, medical patients exposed to x rays, and laboratory animals.

^cPiecewise-linear distribution with most probable value at 2 and lower and upper bounds of (0.2, 5) based on modification of distribution recommended by NCRP (1997) to include possibility of supralinear response at low acute doses (DDREF <1), as suggested by dose-response for cancer mortality in LSS cohort.

^dPiecewise-linear distribution with most probable value at 4 and lower and upper bounds of (0.2, 10) based on distribution for solid cancer mortality excluding lung and comparison of risks of lung cancer mortality in medical patients exposed to x rays with risks in LSS cohort.

^eDiscrete distribution with most probable values at 1.5 and 2 and lower and upper bounds of (0.5, 5) based mainly on assessment of nonlinearity in dose-response for solid cancer mortality in LSS cohort.

^fDiscrete distribution with most probable value at 1.0 and lower and upper bounds of (0.5, 4) based on distribution for solid cancer mortality excluding breast and thyroid and conclusion that dose-responses for breast and thyroid cancer in LSS cohort showed greater tendency toward linearity.

^gExcluding thyroid and nonmelanoma skin cancers but including lymphomas and multiple myeloma.

^hLognormal distribution with 95% CI of (0.83, 2.7) was based on analysis of curvature in dose-response for solid cancer incidence in LSS cohort assuming linear-quadratic model modified to take into account dose-responses for cancer induction in laboratory animals. Distribution essentially represents an LDEF.

ⁱEntries are reciprocals of upper and lower bounds of 90% CIs of combined ratios of estimated risks in workers or members of the public from chronic exposure to low-LET radiation at low dose rates to estimated risks from acute exposure in LSS cohort, referred to as “risk ratios.” First entry represents main result based on estimated risks from seven studies of cancer mortality in workers or members of the public; second and third entries were based on estimated risks from four studies of cancer mortality and three studies of cancer incidence, respectively. Distributions represent DREFs.

^jEntries are reciprocals of central estimates of combined risk ratios.

^kMeans of probability distributions of combined risk ratios were not reported, and means of corresponding DDREFs are not estimated.