Table 4.
Associations of predictor variables with HIV test refusal (Unadjusted) (N = 1525)
| Variable | HIV test refusers (N = 94) n (%)b |
HIV test acceptors (N = 1431) n (%)b |
OR for refusala | 95% CI | P value |
|---|---|---|---|---|---|
| Socio-demographic characteristics | |||||
| Mean age | 23.9 | 23.6 | 1.02 | 0.98–1.07 | 0.323 |
| Low education (only primary or less education) | 77 (81.9) | 1200 (83.9) | 0.73 | 0.40–1.33 | 0.307 |
| Currently married | 80 (86.0) | 1256 (87.8) | 1.01 | 0.53–1.94 | 0.966 |
| In a polygamous relationship | 23 (24.7) | 371 (25.9) | 0.88 | 0.52–1.48 | 0.639 |
| First pregnancy | 17 (18.1) | 309 (21.6) | 0.89 | 0.50–1.58 | 0.699 |
| Working in farming or manual laborc | 32 (34.4) | 690 (48.3) | 1.12 | 0.69–1.83 | 0.635 |
| Luo ethnic group | 91 (96.8) | 1316 (92.0) | 2.65 | 0.78–8.87 | 0.117 |
| HIV/AIDS stigma measures | |||||
| Anticipated stigma | |||||
| Anticipated stigma from partnerd | 47 (50.0) | 522 (36.5) | 2.07 | 1.29–3.33 | 0.002 |
| Anticipated stigma from familye | 45 (47.9) | 518 (36.2) | 1.53 | 0.96–2.42 | 0.074 |
| Anticipated stigma from othersf | 61 (64.9) | 880 (61.5) | 1.45 | 0.90–2.32 | 0.129 |
| Mean anticipated stigma combined score (n = 1511)g | 0.3499 | 0.2805 | 2.13 | 1.06–4.28 | 0.035 |
| Perceived community stigma | |||||
| Mean negative attitudes sub-scale (n = 1515) | 0.8346 | 0.7878 | 1.21 | 0.70–2.09 | 0.504 |
| Mean perceived discrimination sub-scale (n = 1509) | 0.8784 | 0.8967 | 0.86 | 0.47–1.56 | 0.614 |
| Mean total perceived community stigma score (n = 1504)h | 0.8414 | 0.8540 | 0.92 | 0.45–1.85 | 0.808 |
| Other individual-level predictors | |||||
| Mean months since the study began (Nov 20, 2007) | 8.99 | 10.66 | 0.91 | 0.86–0.97 | 0.003 |
| Knowledge of mother-to-child transmission | 74 (81.3) | 1256 (88.5) | 0.81 | 0.44–1.49 | 0.504 |
| Knows someone with HIV/AIDS | 63 (67.0) | 1109 (77.6) | 0.44 | 0.27–0.73 | 0.001 |
| Does not know if partner has tested for HIVi | 31 (33.0) | 308 (21.5) | 1.95 | 1.19–3.20 | 0.008 |
aEstimates account for clustering by site (health facility), using random effects logistic regression
bResults are presented as n(%) in these columns, unless otherwise indicated to be means. The denominators for percentages differ slightly in some cases, due to small numbers of missing responses
cAs compared to those working in housework, selling things (including fish), or other occupations (combined)
d72 Women refused to answer one or both of the anticipated partner stigma items
e35 Women refused to answer one or both of the anticipated family stigma items
f26 Women refused to answer one or both of the anticipated “other” stigma items
gMean anticipated stigma score was computed for women who had non-missing data for at least 6 of the 9 items
hMean perceived community stigma sub-scale scores were computed for women who had no more than one missing item in that sub-scale, and the mean total score was computed for those who do not have more than 1 item missing on any sub-scale
i22 Women refused to answer the question about her knowledge of her male partner’s HIV testing status