. Author manuscript; available in PMC: 2018 Oct 1.

Published in final edited form as: J Urol. 2017 May 18;198(4):848–857. doi: 10.1016/j.juro.2017.05.065

Table 3.

Ordinal Logistic Regression Models: Urinary Severity Interactions (Adjusted for Baseline Severity)^a

Model	Dependent Variable	n	Variable	Estimate	p-value	Odds Ratio (CI)	C Statistic
1	Urinary Severity	377	Urinary Severity at Baseline	0.140	<.0001	1.151 (1.104–1.199)	0.711
			Duration of symptom: 0:<2yr,1:>=2yrs	−0.420	0.060	0.657 (0.424–1.017)
			SF12 - PCS (Duration < 2yrs)	0.068	<.0001	1.070 (1.032–1.110)
			SF12 - PCS (Duration ≥ 2yrs)	0.013	0.374	1.013 (0.985–1.041)
2	Urinary Severity	397	Urinary Severity at Baseline	0.132	<.0001	1.141 (1.099–1.185)	0.714
			Duration of symptom: 0:<2yr,1:>=2yrs	−0.508	0.021	0.602 (0.392–0.925)
			BPI Sites, Non-pelvic (Duration < 2yrs)	−0.113	<.0001	0.893 (0.845–0.944)
			BPI Sites, Non-pelvic (Duration ≥ 2yrs)	−0.013	0.595	0.987 (0.942–1.035)

SF12 PCS and BPI sites variables were centered prior to analysis to facilitate interpretation of the direction of the interactions.