Skip to main content

Table 3 Risk factors associated with ‘never testing’ for hiv

From: ‘Never testing for HIV’ among Men who have Sex with Men in Viet Nam: results from an internet-based cross-sectional survey

 

Univariate logistic regression analysis

Multivariate logistic regression (final models for each independent variable)

Factors

 

OR

  

aOR

 
 

Beta

(95% CI)

p-value

Beta

(95% CI)

p-value

Age ≥ 25 years

−0.84

0.43*

 

−0.17

0.84

 
 

(0.34 - 0.55)

<0.001

 

(0.52–1.38)

0.50

Homosexual (self-reported)

0.22

1.25

 

0.41

1.51

 
 

(0.97 - 1.61)

0.09

 

(0.94–2.40)

0.09

Bong Kin (self-reported)

−0.14

0.87

 

0.09

1.09

 
 

(0.67 - 1.12)

0.28

 

(0.69–1.73)

0.72

Bisexual

−0.09

0.92

 

0.10

1.10

0.65

 

(0.73 - 1.16)

0.48

 

(0.73–1.65)

 

Education (university, vocational school or higher)

−0.66

0.52*

 

−0.33

0.72

 
 

(0.39 - 0.69)

<0.001

 

(0.42–1.23)

0.23

Marital Status (Married)

−0.44

0.64

 

0.31

1.36

 
 

(0.40 - 1.03)

0.07

 

(0.50–3.69)

0.54

Monthly income (≤ 5 million VND)

1.12

3.07*

 

0.69

1.99*

 
 

(2.41 - 3.93)

<0.001

 

(1.22–3.23)

0.006

Student

0.96

2.62*

 

0.63

1.88*

0.02

 

(2.03 - 3.38)

<0.001

 

(1.09–3.24)

 

Urban Resident

−0.51

0.60*

 

−0.23

0.79

 
 

(0.44 - 0.84)

0.002

 

(0.46–1.38)

0.41

Internet Use (≥ 15 hrs/wk)

−0.33

0.72*

 

−0.50

0.61*

 
 

(0.57 - 0.91)

0.007

 

(0.40–0.92)

0.02

Internet Solicitation for Casual Sex Partners

0.14

1.15

 

0.02

1.02

 
 

(0.88 - 1.49)

0.32

 

(0.63–1.65)

0.95

Anal Sex (≤ 6 mos)

−0.46

0.63*

 

−0.41

0.67

 
 

(0.48 - 0.84)

0.001

 

(0.39–1.14)

0.14

Alcohol Consumption (before and/or after sex)

−0.33

0.72*

 

0.05

1.05

 
 

(0.56 - 0.92)

0.01

 

(0.68–1.62)

0.84

Medically Circumcised

0.20

1.22

 

0.53

1.69

 
 

(0.83 - 1.80)

0.31

 

(0.91–3.14)

0.10

Accessed free condoms (≤ 12 mos)

−0.80

0.45*

 

−0.31

0.73

0.32

 

(0.32 - 0.63)

<0.001

 

(0.40–1.35)

 

Accessed free water-based lubricant (≤ 12 mos)†

 

0.50*

  

1.37

 

−0.70

(0.33 - 0.75)

0.001

0.31

(0.54–3.45)

0.51

Received HIV Education Materials (general ≤ 12 mos)

−0.41

0.66*

 

−0.43

0.65

0.11

 

(0.50 - 0.87)

0.003

 

(0.39–1.10)

 

Accessed HIV Education Materials (MSM brochures ≤ 12 mos)

−0.63

0.53*

 

0.29

1.33

 
 

(0.38 - 0.74)

<0.001

 

(0.66–2.66)

0.42

Contacted by MSM Peer Educator (≤ 12 mos)

−0.91

0.40*

 

−0.45

0.64

0.25

 

(0.26 - 0.62)

<0.001

 

(0.30–1.36)

 

Participation in HIV behavioral intervention (≤ 12 mos)

−0.33

0.72*

 

−0.58

0.56*

 
 

(0.53 - 0.98)

0.04

 

(0.35–0.88)

0.01

Perceived moderate- to high-risk of HIV infection

−0.11

0.90

 

−0.27

0.76

0.30

 

(0.65 - 1.24)

0.52

 

(0.46–1.26)

 

Casual, commercial partners (6 mos)

0.22

1.24

0.11

0.50

1.65

0.09

 

(0.95 - 1.62)

  

(0.93–2.90)

 

Multiple sexual partnerships (6 mos)

−0.36

0.70*

0.01

−0.21

0.82

0.33

 

(0.53 - 0.92)

  

(0.54–1.23)

 

Inconsistent condom use (6 mos)

0.42

1.52*

 

0.37

1.45

0.07

  

(1.14 - 2.02)

0.004

 

(0.97–2.16)

 
  1. Beta: Regression Coefficient; OR = Odds Ration; CI = Confidence Interval; aOR = Adjusted Odds Ratio.
  2. *Statistically significant (P-value < 0.05).
  3. †Due to colinearity between “receiving free condoms in the last 12 months” and “receiving free water-based lubricant in the last 12 months”, the independent variable “receiving free water-based lubricant in the last 12 months” was removed from the multiple logistic regression models due to lower number of observations. Note: Colinearity does not bias results, but it produces large standard errors in the related independent variables and an effective remedy for Colinearity is to drop an explanatory variable in order to produce a model with significant coefficient estimates for the remaining explanatory variables [14].
Back to article page

BMC Public Health

ISSN: 1471-2458

Contact us