Table 4.

GEE empirical significance tests, Odds Ratios (OR), and 95 % confidence intervals (CI) for the overall intervention effect unadjusted and adjusted for baseline prevalence and significant covariates, African American MSM, Philadelphia, PA 2008–2012

Outcome	Unadjusted			Adjusted for baseline
	OR (95 % CI)	p value	d	OR (95 % CI)	p value	d
Consistent condom use	1.03 (0.73, 1.45)	0.8658	0.02	1.01 (0.71, 1.44)	0.9504	0.01
Proportion condom-protected intercourse	1.55 (0.87, 2.77)	0.1362	0.27	1.58 (0.89, 2.84)	0.1210	0.28
Unprotected intercourse	0.96 (0.68, 1.35)	0.8098	−0.03	0.99 (0.69, 1.42)	0.9612	0.00
Multiple sexual partners	1.07 (0.80, 1.43)	0.6641	0.04	0.96 (0.71, 1.30)	0.7949	−0.02
Insertive anal intercourse	1.11 (0.81, 1.51)	0.5244	0.06	1.12 (0.81, 1.54)	0.4982	0.05
Receptive anal intercourse	0.80 (0.57, 1.13)	0.2016	−0.14	0.64 (0.44, 0.94)	0.0218	−0.27

The intervention effect is averaged over the 6-month and 12-month post-intervention assessments. All adjusted analyses adjust for baseline of the criterion. Proportion of condom-protected, insertive anal intercourse, and receptive anal intercourse also adjusted for sexual self-identification. Multiple partners also adjusted for self-reported HIV status. Insertive anal intercourse also adjusted for age group. d is the effect size estimate in standard deviation units based on Cox transformation of the odds ratio [61]

GEE generalized estimating equations, MSM men who have sex with men