Skip to main content
. Author manuscript; available in PMC: 2019 Apr 26.
Published in final edited form as: Int J Gynaecol Obstet. 2015 May 7;130(Suppl 3):E31–E36. doi: 10.1016/j.ijgo.2015.05.001

Table 4.

Multivariate parsimonious model of predictors of contraceptive use and unmet need among sexually active individuals with HIV/AIDS living in the Nairobi slums (n = 450).a

Characteristic Contraceptive use Unmet need for contraception


Adjusted OR
(95% CI)
P valueb Adjusted OR
(95% CI)
P valueb
Sex
    Female (reference) 1.00 1.00
    Male 1.01 (0.63–1.64) 0.956 0.66 (0.41–1.07) 0.091
Education level
    None (reference) 1.00 1.00
    Primary 3.40 (1.16–9.93) 0.026 0.51 (0.18–0.74) 0.020
    Secondary or higher 3.12 (1.04–9.40) 0.043 0.53 (0.19–0.78) 0.025
Marital status
    Single (reference)c 1.00 1.00
    Married or co-habiting 5.25 (3.30–8.27) <0.001 0.24 (0.16–0.38) <0.001
No. of living children
    0 (reference) 1.00 1.00
    1–2 3.37 (1.08–10.52) 0.036 2.35 (0.82–6.76) 0.111
    3–4 4.04 (1.30–12.57) 0.016 2.28 (0.79–6.59) 0.125
    ≥5 3.85 (1.20–12.31) 0.023 3.12 (1.05–9.24) 0.040
Discussion with a health provider
    Never (reference) 1.00 1.00
    Ever 2.54 (1.60–4.04) <0.001 0.65 (0.41–0.99) 0.049
    Social support 1.55 (1.08–2.24) 0.017 0.76 (0.53–1.09) 0.132

Abbreviations: CI, confidence interval; OR, odds ratio.

a

Logistic regression. The model was adjusted for age, ethnicity, duration of HIV infection, duration of antiretroviral therapy, disclosure of HIV status, stigma, and depression.

b

0.05.

c

Included all women who were divorced, separated, never married, and never cohabited.