Table 8.

Excess relative rate (ERR) of leukemia^* per gray for restricted 5-year lagged dose ranges in subsets of the Life Span Study and INWORKS

	Dosea ranges
	0–100 mGy	0–200 mGy	0–300 mGy	0–500 mGy	0–1000 mGy	Whole
Life Span Study
Mean RBM dose	16.2	28.7	39.2	57.6	88.6	134.3
Person-years	1,158,870	1,266,440	1,320,560	1,382,440	1,442,100	1,480,340
Observed deaths	120	127	137	146	161	196
ERR/Gy	−2.18	−1.93	0.43	0.59	1.15	2.75
90% CI	−5.56; 3.25	< 0; 0.27	−0.93; 2.44	−0.43; 2.03	0.37; 2.22	1.73; 4.21
P (vs. null model)b	0.455	0.139	0.652	0.381	0.008	< 0.001
Fitted excess deathsc	−5.1	−8.7	2.7	5.6	17.0	57.6
INWORKS
Mean RBM dose	8.6	11.7	13.3	14.5	14.9	14.9
Person-years	5,943,550	6,104,410	6,150,100	6,173,470	6,178,150	6,178,320
Observed deaths	422	446	457	463	464	464
ERR/Gy	4.24	3.61	3.79	3.46	3.21	3.15
90% CI	−0.71; 10.47	0.47; 7.54	1.25; 6.99	1.29; 6.19	1.16; 5.80	1.12; 5.72
P (vs. null model)b	0.166	0.054	0.009	0.004	0.005	0.006
Fitted excess deathsc	18.1	23.2	29.4	30.6	29.5	29.1

^*Excluding chronic lymphocytic leukemia in INWORKS

^aCumulative dose in INWORKS

^bLikelihood ratio test

^cUnder a linear ERR model (no modifying effect of age)

RBM: red bone marrow. CI: likelihood-based confidence interval

The LSS ERR and fitted excess estimates are weighted averages over sex with a weight of 0.88 for males and 0.12 for females; sex-averaged estimates for the restricted dose ranges were obtained by fixing the parameter for effect modification by sex to the value for that parameter obtained when fitting the model to the full-dose range