Table 2. Univariable (crude) and multivariable Poisson regression models of the association between the PUCP and the KRT incidence.

		Rate ratio (95% CI)	P value
Crude
	Uncoded CKD patients in % (continuous)	1.000 (0.997 to 1.003)	0.886
	Uncoded CKD patients in % (quintiles)
	Lowest	1	<0.001
	2	1.221 (1.167 to 1.277)
	3	1.123 (1.074 to 1.174)
	4	1.075 (1.029 to 1.123)
	Highest	1.099 (1.051 to 1.149)
Model 1 a
	Uncoded CKD patients in % (continuous)	1.002 (0.999 to 1.005)	0.149
	Uncoded CKD patients in % (quintiles)
	Lowest	1	0.062
	2	1.033 (0.982 to 1.087)
	3	1.051 (1.003 to 1.102)
	4	1.000 (0.954 to 1.048)
	Highest	1.044 (0.996 to 1.093)
Model 2 b
	Uncoded CKD patients in % (continuous)	1.003 (1.000 to 1.006)	0.093
	Uncoded CKD patients in % (quintiles)
	Lowest	1	0.360
	2	1.020 (0.969 to 1.075)
	3	1.030 (0.982 to 1.082)
	4	1.000 (0.953 to 1.050)
	Highest	1.037 (0.990 to 1.087)
Model 3 c
	Uncoded CKD patients in % (continuous)	1.004 (1.000 to 1.008)	0.029
	Uncoded CKD patients in % (quintiles)
	Lowest	1	0.030
	2	1.080 (1.021 to 1.143)
	3	1.061 (1.009 to 1.116)
	4	1.035 (0.982 to 1.090)
	Highest	1.077 (1.020 to 1.138)

^aAdjusted for sociodemographic covariates. ^bAdjusted for sociodemographic and healthcare quality covariates. ^cAdjusted for sociodemographic, healthcare quality and health status covariates. P values are calculated using likelihood-ratio tests. n = 106. CI = confidence interval. CKD = chronic kidney disease