Skip to main content

Table 6 Predictive models for each different type of infection as outcome variables (N = 1026)1

From: Human immunodeficiency virus, syphilis and hepatitis C virus prevalence trends among cross-border migrant Vietnamese female sex workers in Guangxi, China

 

HIV

Syphilis

HCV

PREDICTORS

Univariate

Multivariablea

Univariate

Multivariableb

Univariate

Multivariablec,#

 

OR (95 % CI)

aOR (95 % CI)4

OR (95 % CI)

aOR (95 % CI)4

OR (95 % CI)

aOR (95 % CI)4

Education

2.68 (1.23,5.83)***

1.82 (0.79,4.20)

2.47 (1.46,4.17)****

0.92 (0.49,1.73)

2.37 (1.03,5.46)***

1.68 (0.59,4.77)

Length of working

0.31 (0.14,0.69)****

0.51 (0.21,1.25)

0.25 (0.14,0.45)******

0.60 (0.30,1.18)

0.41 (0.18,0.95)***

1.53 (0.50,4.67)

Marital status

0.73 (0.36,1.47)

N/A

0.34 (0.20,0.58)******

0.96 (0.49,1.91)

0.41 (0.18,0.95)***

0.78 (0.27,2.26)

No condom use in the last sex act

1.60 (0.55,4.68)

N/A

3.04 (1.62,5.72)****

1.55 (0.68,3.53)

0.88 (0.20,3.77)

N/A

Inconsistent condom use in the past month

1.31 (0.61,2.79)

N/A

3.51 (2.14,5.74)******

2.27 (1.22,4.22)***

1.60 (0.70.3.63)

N/A

HIV knowledge

0.82 (0.46,1.45)

N/A

0.84 (0.56,1.25)

N/A

0.58 (0.30,1.13)

0.63 (0.27,1.48)

Syphilis infection

3.17 (1.26,7.96)***

1.58 (0.56,4.46)

N/A

N/A

9.05 (3.97,20.59)******

3.64 (1.20,11.00)***

HCV infection

5.82 (1.89.17.90)****

4.14 (0.89,19.36) **

9.05 (3.97,20.60)******

4.78 (1.53,14.90)****

N/A

N/A

Age (continuous)2

1.05 (1.01,1.08)****

1.01 (0.98,1.05)

1.09 (1.06,1.11)******

N/A

1.05 (1.01,1.09)****

1.01 (0.95,1.07)

Age-younger (RCS)3

N/A

N/A

N/A

1.18 (1.01,1.37)***

N/A

N/A

Age-older (RCS)3

N/A

N/A

N/A

0.90 (0.76,1.07)

N/A

N/A

Drug use

3.32 (0.96, 11.51)*

0.84 (0.15,4.82)

8.15 (3.75,17.68)******

2.56 (0.87,7.49)**

58.63 (24.14,142.39)******

44.01 (16.34,129.53)******

Venues types (medium vs. low)

0.28 (0.11,0.72)***

0.46 (0.17,1.28)

0.36 (0.20,0.66)****

0.83 (0.41,1.65)

0.56 (0.23,1.34)

0.34 (0.09,1.38)

Venues types (high vs. low)

0.41 (0.05,3.06)

0.91 (0.11,7.46)

N/A

N/A

N/A

N/A

  1. Notes: 1.All predictive models employed multiple imputation to deal with missing values; 2. LRT is non-significant for the RCS terms; 3.RCS-Restricted Cubic Spline with 3 knots for age; 4. aOR (95 % CI): adjusted odds ratios and its 95 % confidence intervals.
  2. Model a’s AUC is 0.84 (0.80-0.88), Model b’ AUC is 0.80 (0.75-0.84), Model c’s AUC is 0.82 (0.70, 0.93). Areas under the ROC curve (AUC) are used to provide the predictive value for each model.
  3. #-this model employed MI to impute missing values of HIV knowledge, as the missing values have more than 20 %
  4. *p < 0.20, **0.05 < p < 0.10, ***p < 0.05, ****p < 0.005, ******p < 0.0001