Table 1.
HIV DHS 2001: 3,949 | HIV DHS 2007:11,138 | |||||||
---|---|---|---|---|---|---|---|---|
Variables | Cases | P-value** | Marginal OR & 95%CI*** |
Posterior OR & 95% CI† |
Cases | P-value** | Marginal OR & 95%CI*** |
Posterior OR & 95% CI |
Number | 565(14.3) | 1596(14.7) | ||||||
Gender | ||||||||
Male | 212(11.7) | 1.00 | 649(12.6) | 1.00 | ||||
Female | 353(16.6) | <0.001 | 1.59(1.32, 1.89) | 353(16.6) | <0.001 | 1.43(1.43, 1.59) | ||
Place of residence | ||||||||
Urban | 290(23.0) | 2.73(2.13, 3.28) | 934(19.8) | 2.35(2.11, 2.63) | ||||
Rural | 275(10.2) | <0.001 | 1.00 | 662(10.8) | <0.001 | 1.00 | ||
Age at diagnosis | ||||||||
<24 years | 122(7.5) | See figure 3a | 299(6.9) | See figure 3a | ||||
25 & 35 years | 273(21.2) | 728(19.6) | ||||||
36 &49 years | 159(17.7) | 518(21.2) | ||||||
50 & 59 years | 11(8.0) | <0.001 | 51(12.6) | <0.001 | ||||
Region of residence*** | ||||||||
Central(4)(2) | 85(15.1) | 2.06(1.42, 2.99) | 1.22(0.97, 1.55) | 177(18.4) | 2.39(1.85, 3.10) | 1.39(1.06, 1.82) | ||
Copperbelt(2)(4) | 99(19.8) | 2.87(1.99,4.13) | 1.02(0.74, 1.43) | 196(16.7) | 2.13(1.65, 2.74) | 0.99(0.73, 1.31) | ||
Eastern(5)(7) | 53(13.6) | 1.83(1.21, 2.76) | 1.11(0.84, 1.43) | 180(11.8) | 1.42(1.10, 1.83) | 0.91(0.68, 1.13) | ||
Luapula(7)(6) | 37(11.1) | 1.45(0.92, 2.27) | 0.86(0.61, 1.17) | 155(14.5) | 1.80(1.38, 2.34) | 1.03(0.83, 1.28) | ||
Lusaka(1)(1) | 95(22.1) | 3.29(2.27, 4.76) | 1.09(0.78, 1.41) | 297(19.8) | 2.62(2.06, 3.32) | 1.29(0.99, 1.56) | ||
Northern(9)(9) | 49(7.9) | 1.00 | 0.69(0.52, 0.87) | 104(8.6) | 1.00 | 0.61(0.46, 0.81) | ||
North-western(8)(8) | 41(9.3) | 1.18(0.77, 1.82) | 0.81(0.60, 1.03) | 87(8.6) | 1.01(0.75, 1.35) | 0.61(0.44, 0.85) | ||
Southern(3)(5) | 69(17.3) | 2.43(1.64, 3.59) | 1.33(1.06, 1.91) | 202(15.9) | 2.01(1.56, 2.58) | 1.25(0.90, 1.60) | ||
Western(6)(3) | 37(13.4) | <0.001 | 1.79(1.14, 2.81) | 1.13(0.82, 1.57) | 198(17.3) | <0.001 | 2.23(1.73, 2.87) | 1.34(1.01, 1.78) |
Data are expressed as number (percentages) of cases that are HIV positive.
P-value comparison across HIV positive and negative cases using the chi-squared test for categorical variables.
Crude unadjusted marginal odds ratio (OR) obtained from the standard logistic regression model and ranked from the highest to the lowest OR. The Northern province is taken as reference because its low HIV prevalence in both surveys.
Rank of marginal OR in 2001: Lusaka(1), Copperbelt(2), Southern(3), Central(4), Eastern(5), Western(6), Luapula(7), Northwestern(8), Northern(9).
Rank of marginal OR in 2007: Lusaka(1), Central(2), Western(3), Copperbelt(4), Southern(5), Luapula(6), Eastern(7), Northwestern(8), Northern(9).
Spatially adjusted posterior odds ratio (OR) obtained from the Bayesian geo-additive regression model after controlling for nonlinear effect of age, categorical variables and the province of residence (spatial effects).