Table 3

Multilevel logistic regression model of determinants of syphilis prevalence among clients of FSWs, IBBA round 1 and round 2

CharacteristicsWithout programme variable and interaction termsWith programme variable and interaction terms
Fixed part of the modelAORp Value95% CIAORp Value95% CI
Constant0.01<0.0010.01 to 0.020.010.0300.01 to 0.03
IBBA round
 Round 1
 Round 20.49<0.0010.38 to 0.620.330.0660.10 to 1.08
Age of respondent (years)
 <25
 25–341.430.0091.09 to 1.881.470.0061.12 to 1.92
 ≥352.32<0.0011.76 to 3.042.31<0.0011.76 to 3.03
Can read and write
 No
 Yes0.66<0.0010.54 to 0.810.65<0.0010.53 to 0.80
Typology of FSWs
 Public place-based1.140.4360.83 to 1.560.940.7340.67 to 1.33
 Brothel-based1.340.0920.95 to 1.881.110.5750.78 to 1.58
 Home-based
 Other1.090.7330.67 to 1.771.270.4490.69 to 2.33
Anal sex with a man/transgender in past 6 months
 No
 Yes1.580.0021.18 to 2.111.530.0051.14 to 2.06
Consistent condom use with FSWs
 No
 Yes0.800.0340.64 to 0.980.850.1410.69 to 1.05
HIV infection status
 Negative
 Positive6.88<0.0015.51 to 8.594.14<0.0013.04 to 5.64
Baseline syphilis prevalence1.19<0.0011.10 to 1.301.18<0.0011.10 to 1.26
Programme indicators
 FSW programme coverage at baseline1.000.6850.99 to 1.01
 Difference in coverage between the two rounds1.000.3271.00 to 1.01
Interaction between round and
 Difference in coverage between the two rounds0.980.0230.97 to 0.99
 Baseline HIV prevalence0.850.0030.76 to 0.94
 Status of HIV3.57<0.0012.26 to 5.64
Typology of FSWs
 Public place-based4.610.0031.67 to 12.72
 Brothel-based4.040.0061.48 to 11.04
 Other2.330.1860.67 to 8.12
Random part of the model
Estimated district-level variance ofEstimated variance95% CI of varianceEstimated variance95% CI of variance
Round0.070.03 to 0.170.040.01 to 0.10
Intercept0.00Not significantly different from 00.00Not significantly different from 0
  • AOR, adjusted OR; FSW, female sex worker; IBBA, integrated behavioural and biological assessment survey.