Table 4.

Univariate and multivariate models of the association between specific vaginal practices at baseline and follow-up, and consistent condom use among MIRA participants (both study groups combined) (4874 participants, 25,905 persons-visits)

Vaginal practices	Univariate analyses				Multivariate analyses
	OR estimate	Lower 95% CI	Upper 95% CI	Chi-square P-value	AOR estimate	Lower 95% CI	Upper 95% CI	Chi-square P-value
Baseline measures
Baseline intravaginal washing	0.89	0.79	0.99	0.033	na
Baseline intravaginal wiping	0.84	0.77	0.91	<.0001	na
Baseline intravaginal insertion	0.91	0.82	1.01	ns	na
Baseline used product in vaginaa	0.88	0.81	0.97	0.0064	0.90	0.82	0.98	0.021
Repeated-measures at follow up
Intravaginal washing at follow-upa	0.83	0.78	0.89	<.0001	0.88	0.82	0.94	0.0002
Intravaginal wiping at follow-upa	0.89	0.85	0.94	<.0001	0.94	0.89	0.99	0.0327
Intravaginal insertion at follow-upa	0.86	0.80	0.93	0.0002	0.89	0.82	0.97	0.0067

^aIn univariate analyses, washing, wiping and insertion at follow-up visits were controlled for at study site

Multivariate analyses controlled for baseline factors and potential confounders that remained significant at P < 0.05: study group, study site, age, married, educational status, women’s behavioral risk, high risk partner, cohabitation, postive curable STI at baseline

OR odds ratio, AOR adjusted OR, CI confidence interval, na not applicable, ns non significant