Table III.

Cox proportional hazards models for the influences of the number of ultraviolet B (UVB) phototherapy treatments on study outcomes

	Skin cancer			Non-melanoma skin cancer			Cutaneous melanoma
	HR	95% CI	p-value	HR	95% CI	p-value	HR	95% CI	p-value
Model 1a  Number of UVB phototherapy treatments	1.002	0.985–1.019	0.85	0.998	0.974–1.023	0.90	1.010	0.989–1.031	0.35
Model 2b  Number of UVB phototherapy treatments	1.001	0.984–1.017	0.95	0.998	0.974–1.022	0.84	1.009	0.988–1.031	0.40
Age, years	1.048	1.028–1.069	< 0.001	1.046	1.026–1.067	<0.001	1.071	0.995–1.153	0.07
Male sex	1.676	1.055–2.662	0.03	1.571	0.976–2.531	0.06	4.525	0.519–39.416	0.17
Model 3c  Number of UVB phototherapy treatments	1.001	0.984–1.018	0.90	0.998	0.974–1.022	0.87	1.010	0.987–1.033	0.42
Age, years	1.065	1.041–1.089	< 0.001	1.061	1.037–1.086	<0.001	1.115	1.017–1.222	0.02
Male sex	1.739	1.076–2.810	0.02	1.629	0.993–2.672	0.05	4.345	0.491–38.451	0.19
Outdoor occupation	2.102	1.292–3.422	0.003	1.964	1.180–3.269	0.01	4.674	0.874–24.991	0.07
High UV indexd	1.111	0.696–1.775	0.66	1.257	0.772–2.047	0.36	0.209	0.023–1.865	0.16
Organ transplantation	5.896	3.171–10.962	< 0.001	5.379	2.811–10.293	<0.001	17.139	1.961–149.772	0.01
Hydrochlorothiazide	1.373	0.865–2.181	0.18	1.328	0.819–2.154	0.25	2.060	0.407–10.411	0.38

^aModel 1: Univariate Cox proportional hazards model adjusted for the number of UVB phototherapy treatments received by each patient.

^bModel 2: Multivariate Cox proportional hazards model adjusted for the variable in model 1 and for age and sex.

^cModel 3: Multivariate Cox proportional hazards model adjusted for all variables in model 2 and for occupation, high UV index, organ transplantation, and the prescription of hydrochlorothiazide.

^dMean UV index > 6 at location of residence.

The number of UVB phototherapy treatments was analysed as a continuous variable, and the HR indicates the effect of every 1 additional UVB phototherapy treatment received by the patient.

CI: confidence interval; HR: hazard ratio.