Table 2.

Latencies, LRT, fitted parameters, ERR and bootstrap confidence intervals for the two-phase model

Site	ϕmabs	ϕm						ERR1	ERR0.025
Stom	11.89	11.89	21.19	19.64	0.46	120.01	82.67	0.46 (0.21, 0.80)bca (0.19, 0.79)per	0.39 (0.15, 0.70)bca (0.15, 0.71)per
Liver	36.90	36.90	34.87	27.64	1.43	291.22	76.76	1.43 (0.70, 2.41)bca (0.72, 2.39)per	1.10 (0.49, 1.90)bca (0.40, 1.88)per
Lung	36.99	13.60	16.04	14.74	0.44	37.90	4.46	0.88 (0.34, 1.50)bca (0.27, 2.07)per	0.86 (0.40, 1.49)bca (0.32, 2.02)per
Panc	11.86	11.86	16.22	14.80	9.77	1060.94	43.77	9.77 (3.50a, 14.27)bca (1.72, 55.11)per	9.12 (3.25b, 13.46)bca (1.49, 48.25)per
Leuk	23.66	23.66	25.40	25.29	1.69	3353.15	172.25	1.69 (0.20, 4.81)bca (0.19, 4.77)per	1.17 (0.18, 3.58)bca (0.24, 3.79)per

ϕ_mabs: latency (years) which maximises LRT_2p-con

ϕ_m: optimal latency, which maximises LRT_2p-con with the constraint ERR₁ ≥ 0

: Likelihood Ratio Test for the two-phase model against the control model at ϕ_m

: Likelihood Ratio Test for the two-phase model against the linear model at ϕ_m

: fitted parameters for the two-phase model at ϕ_m

ERR₁ : Excess Relative Risk at 10 mSv (lagged by ϕ_m), as estimated from the fitted two-phase model

ERR_0.025 : Excess Relative Risk at 0.25 mSv (lagged by ϕ_m), as estimated from the fitted two-phase model

^bca Bootstrap BC_a 95%CI (5000 bootstrap replications)

^per Bootstrap percentile 95%CI (1000 bootstrap replications)

^a : minimum bootstrap value, lower tail probability = 0.034

^b : minimum bootstrap value, lower tail probability = 0.024