TABLE 3.

Adjusted hazard ratios (and 95% confidence intervals) from multivariate analyses assessing rates of contraceptive discontinuation and pregnancy over one year, by selected participant characteristics

Characteristic	Contraceptive discontinuation	Pregnancy
Perceived benefits of childbearing	1.07 (0.99–1.14)	1.17 (1.02–1.34)*
Feeling if became pregnant
Very upset (ref)	1.00	1.00
Somewhat upset	1.00 (0.85–1.18)	1.66 (1.23–2.22)***,§
Wouldn’t care/don’t know	1.01 (0.85–1.22)	1.43 (1.02–2.00)*
Somewhat pleased	0.93 (0.73–1.18)	1.02 (0.63–1.64)
Very pleased	1.19 (0.84–1.69)	2.09 (1.21–3.62)**
Age	0.96 (0.93–0.99)**	0.98 (0.93–1.03)
Race/ethnicity
Black (ref)	1.00	1.00
Latina	0.92 (0.78–1.08)	0.85 (0.63–1.13)
White	0.73 (0.59–0.92)**,†	0.67 (0.42–1.07)
Asian/Pacific Islander	0.87 (0.69–1.09)	0.35 (0.19–0.63)***,††
Mixed/other	1.05 (0.79–1.38)	1.02 (0.61–1.70)
Mother’s education >high school	1.12 (0.98–1.28)	1.19 (0.93–1.53)
Has any children	1.13 (0.95–1.34)	1.60 (1.19–2.14)**
In school or employed	1.15 (0.96–1.38)	1.16 (0.85–1.57)
Baseline contraceptive method
Pill (ref)	1.00	1.00
Patch	1.97 (1.66–2.34)***,‡	1.74 (1.27–2.39)***,‡‡
Ring	1.24 (1.02–1.52)*	2.00 (1.40–2.84)***,‡‡
Injectable	1.29 (1.08–1.55)**	0.82 (0.55–1.22)

^*p≤.05.

^**p≤.01.

^***p≤.001.

^†Differs from mixed/other at p≤.05.

^‡Differs from ring and injectable at p≤.001.

^§Differs from somewhat pleased at p≤.05.

^††Differs from Latina and mixed/other at p≤.01.

^‡‡Differs from injectable at p≤.001.

Notes: Tests of significance are two-tailed. The models control for recruitment clinic and interview mode.