Skip to main content
. 2013 Aug 5;16(1):18048. doi: 10.7448/IAS.16.1.18048

Table 5.

Final multivariate logistic regression model for knowing partner's HIV status among 487 relationships that were on-going at baseline

Variable N (% knowing partner's status) Unadj OR, 95% CI Adj OR, 95% CIa p b
Study group
 ART 287 (42) 1.0 1.0 0.009
 MON 200 (49) 1.34 (0.92, 1.94) 1.95 (1.18, 3.22)
Disclosed HIV status to their partner
 No 141 (9) 1.0 1.0 <0.0001
 Yes 346 (59) 14.32 (7.79, 26.31) 13.42 (7.19, 26.93)
Partner resides
 With member 226 (61) 1.0 1.0 0.005
 In isigodi 49 (24) 0.21 (0.10, 0.44) 0.30 (0.13, 0.69)
 Outside isigodi 212 (33) 0.31 (0.21, 0.47) 0.51 (0.30, 0.87)
Gender norm quartiles: 0.01
Least support for equitable norms (31–37) 1 137 (45) 1.0 1.0
               (38–41) 2 154 (51) 1.24 (0.78, 1.99) 1.52 (0.84, 2.77)
               (42–45) 3 84 (29) 0.48 (0.27, 0.88) 0.47 (0.22, 1.01)
Most support for equitable norms (46–57) 4 112 (48) 1.13 (0.67, 1.88) 1.42 (0.72, 2.77)
Has partner had sex with others in last 6 months
 No 175 (59) 1.0 1.0 0.0006
 Do not know 157 (47) 0.59 (0.38, 0.92) 0.88 (0.51, 1.52)
 Yes, I think so/yes, definitely 155 (26) 0.25 (0.15, 0.39) 0.34 (0.19, 0.61)
Condom use at last sex
 No 190 (30) 1.0 1.0 0.004
 Yes 297 (54) 2.76 (1.86, 4.10) 2.04 (1.25, 3.33)
Discuss having another child with partner
 No 314 (43) 1.0 1.0 0.05
 Yes 173 (49) 1.27 (0.87, 1.85) 1.62 (1.0, 2.65)
a

Final model also adjusted for gender (p=0.10) and age as a categorical variable (p=0.44).

b

Adjusted Wald test p-value.

c

Variance adjusted for cluster correlation among participants in multiple on-going relationships.