Table 3.

The odds ratio for the determinants of current use of contraception among 1813 non-pregnant, sexually active women of reproductive age, who say they do not want to get pregnant

			Unadjusted model		Adjusted model
Categories	Total	Current use	Odds Ratio		Odds Ratio
	N = 1813	N = 610	(95% C.I.)	p-value	(95% C.I.)	p-value
Intercept					1.00
HIV statusa
Negative	1651	566	1.00
Positive	155	42	0.71 (0.49,1.03)	0.0714	0.89 (0.60,1.32)	0.5598
Age
15–24	373	144	1.00
25–34	643	312	1.50 (1.16,1.94)	0.0022	1.17 (0.84,1.64)	0.3578
35 and above	797	154	0.38 (0.29,0.50)	<.0001	0.28 (0.19,0.41)	< 0.0001
Martial status
Unmarried	207	55	1.00
Married/Living together	1418	494	1.48 (1.07,2.05)	0.0194	2.23 (1.54,3.23)	< 0.0001
Divorced/Separated/Widowed	188	61	1.33 (0.86,2.05)	0.2008	2.14 (1.33,3.44)	0.0017
Education
None/some primary	593	150	1.00
Completed primary	977	360	1.72 (1.46,2.18)	< .0001	1.67 (1.32,2.13)	< .0001
Secondary+	270	100	2.11 (1.54,2.89)	< .0001	1.95 (1.36,2.80)	0.0003
Number of living children
< 3	539	203	1.00
3+	1274	407	0.78 (0.63,0.96)	0.0187	1.25 (0.91,1.71)	0.1665

Note: Both unadjusted and adjusted odds ratios shown in this table are estimated using logistic regression. Unadjusted odds ratio are based on separate logistic regressions for each predictor variable with that variable as the only predictor variable. Adjusted odds ratios are based on a single logistic regression consisting

of all the predictor variables in the table. For any given predictor variable in the adjusted column, the set of control variables consists of all the other predictor variables in the table

42 women reported to currently using contraceptives whereas they were pregnant/breastfeeding

^aContains 8 missing observations (but less than 10%)