Table 4

Adjusted odds ratios for current perception of high risk of acquiring HIV among mobile FSWs in Southern India

Characteristics

Model I

Model II

Model III

Model IV


Adjusted OR

(95% CI)

Adjusted OR

(95% CI)

Adjusted OR

(95% CI)

Adjusted OR

(95% CI)


Condom use in past one week with occasional clients


Consistent

1

1

Inconsistent

2.85 (2.31 -3.51)

2.07 (1.65 -2.60)

No client

0.84 (0.41 -1.71)

0.95 (0.45 -1.97)


Condom use in past one week with regular clients


Consistent

1

1

Inconsistent

1.16 (0.95 -1.42)

0.99 (0.79 -1.22)

No client

0.92 (0.66 - 1.27)

0.91 (0.64 -1.29)


Condom use in past one week with non paying partners


Consistent

1

1

Inconsistent

0.56 (0.43 -0.72)

0.65 (0.49 -0.86)

No partner

0.32 (0.26 -0.38)

0.52 (0.42 -0.64)


STI symptoms in last six months & use of condom


No STI symptom

1

1

No sex during STI symptom

2.68 (2.31 -3.11)

2.36 (2.02 -2.75)

Continued sex during STI symptom

6.76 (5.56 - 8.23)

5.70 (4.66 - 6.98)


Used alcohol prior to sex


No

1

1

Yes

2.36 (2.04 – 2.71)

2.20 (1.90 – 2.54)


Reason for entering in to sex work


Choice/tradition

1

1

Economic/force

1.39 (1.12 -1.72)

1.39 (1.12 -1.72)


Sex work


Full time

1

1

Part time

1.51 (1.31 -1.75)

1.37 (1.18 -1.59)


State


Tamil Nadu

1

1

1

1

Andhra Pradesh

2.84 (1.43 - 5.66)

2.97 (1.50 -5.89)

1.91 (0.89 - 4.10)

2.16 (1.02 - 4.60)

Karnataka

1.98 (0.99 - 3.93)

1.30 (0.65 - 2.60)

1.59 (0.73 - 3.44)

1.28 (0.59 - 2.78)

Maharashtra

0.37 (0.19 - 0.71)

0.48 (0.25 - 0.93)

0.38 (0.18 - 0.79)

0.45 (0.22 - 0.93)


Random component


District (SD)

0.54 (0.38 - 0.75)

0.54 (0.38 - 0.74)

0.59 (0.42 - 0.83)

0.58 (0.42 - 0.81)

Regression statistics

Log likelihood

-3288.778

-3128.85

-2896.539

-2854.597

Districts

22

22

22

22

N

5413

5413

5413

5413


    Note:

1. The estimated standard deviation (SD) of the district variable without the state variable was 0.99.

2. Experience of sexual violence, living arrangements, currently in debt, degree of mobility, age, education, and marital status were also included as covariates in Models II and IV. Adjusted odds ratios for these covariates (except age) were not significant at 5% level of significance.

Jain et al. BMC Public Health 2011 11(Suppl 6):S5   doi:10.1186/1471-2458-11-S6-S5

Open Data