Table 7.
Logistic regression analysis of variables related to having an HIV test within the last 12 months
| Variables | Crude OR | Adjusted OR | 95% CI of adjusted OR | Mantel-Haenszel X2 |
| Rural location | 1.22 | 1.32 | 1.05–1.66 | 5.52 |
| Male | 0.49 | 0.55 | 0.43–0.69 | 26.45 |
| Less than 40 years old | 1.67 | 1.71 | 1.29–2.25 | 14.28 |
| No school | 0.63 | 0.68 | 0.47–0.97 | 4.48 |
| Believe at risk of HIV | 1.39 | 1.31 | 1.04–1.65 | 5.32 |
| Visited government facility within last 12 m | 2.73 | 2.47 | 1.86–3.27 | 39.80 |
| Believe ARVs can help AIDS | 2.66 | 1.85 | 1.07–3.20 | 4.90 |
| In favour of routine HIV testing in govt clinics | 2.65 | 2.01 | 1.14–3.51 | 5.92 |
The variables included in the original saturated model were: rural/urban location, sex of respondent, age group, education level, whether thought people with HIV should live apart, whether believed themselves at risk of HIV, whether visited a government clinic, whether believed ART could help AIDS, and whether in favour of routine HIV testing