Table 4.

Effects of potential confounders and potential modifiers of the risk estimates

	All hematological malignancies				Lymphoid malignancies				Myeloid malignancies
	#	ERR /100 mGy	95% CI		#	ERR /100 mGy	95% CI		#	ERR /100 mGy	95% CI
Main results1	790	1.96	1.10	3.12	578	2.01	1.02	3.42	203	2.02	0.47	4.77
Potential confounders analysis:
No adjustment for birth cohort
	790	1.92	1.08	3.05	578	1.99	1.02	3.37	203	1.92	0.43	4.53
a) SES2
Unadjusted	210	1.40	0.08	3.83	161	0.99	−0.16	3.37	47	4.22	−0.17	30.6
Adjusted	210	1.44	0.10	3.90	161	1.03	−0.15	3.45	47	4.16	−0.17	29.6
Effect modification analysis:
a) Sex
Males	466	1.45	0.55	2.80	343	1.91	0.71	3.85	118	0.65	-0.42	2.89
Females	324	2.82	1.27	5.32	235	2.14	0.71	4.64	85	6.09	1.62	19.1
Het. P value			0.20				0.85				0.03
b) Age at exposure category (note: one individual can enter in more than one category if they had several CTs)
<5	219	0.78	0.06	1.78	165	0.74	−0.05	1.93	52	1.12	−0.29	3.60
5 to <10	156	1.81	0.57	3.39	115	1.87	0.48	3.74	40	1.72	−0.71	5.41
10+	466	4.02	2.48	5.99	336	4.25	2.41	6.71	124	3.48	1.05	7.35
Het. P value			0.001				0.002				0.32
c) Time since exposure (years) (note: one individual can enter in more than one category if they had several CTs)
2 to <5	303	3.56	1.96	5.57	222	3.09	1.37	5.37	76	4.88	1.66	9.87
5 to <10	291	2.82	1.58	4.33	216	2.90	1.46	4.70	74	2.98	0.66	6.40
10+	260	1.24	0.42	2.29	184	1.46	0.49	2.75	72	0.45	−0.80	2.56
Het. P value			0.07				0.21				0.04

¹Stratified on sex, birth cohort and country—attained age is used as the underlying time variable.

²Analysis restricted to countries where SES data were available: Belgium, France, the Netherlands and Spain.

#, number of cases; Het., heterogeneity.