Apparent and true prevalence of risky behaviours by naïve sensitivity and Bayesian methods in female sex workers participating in a national behavioural survey (N=872), Iran, 2010
| Risky behaviours prone to social desirability bias (categorical variables) | Apparent prevalence | True prevalence (naïve sensitivity) | True prevalence (Bayesian) | |||
|---|---|---|---|---|---|---|
| % | 95% CI | % | 95% CI | % | 95% CRI | |
| Ever married | 81.5 | 71.5 to 88.6 | 87.5 | 73.7 to 101.7* | 87.7 | 77.7 to 97.9 |
| Ever used drugs | 71.6 | 59.7 to 81.1 | 79.4 | *−787.5 to 3150.0* | 69.6 | 22.9 to 87.5 |
| Not tested for HIV, last 12 months | 64.1 | 51.9 to 74.7 | 62.6 | 48.7 to 80.8 | 64.0 | 51.6 to 78.4 |
| Not receiving the result of the HIV test | 21.0 | 9.7 to 39.8 | 31.5 | 9.0 to 93.6 | 36.8 | 14.0 to 82.0 |
| History of genital ulcer or discharge, last 12 months | 52.9 | 39.3 to 66.1 | 82.8 | 61.3 to 120.1* | 82.1 | 65.3 to 98.0 |
| Non-condom use last sex act | 36.7 | 28.0 to 46.8 | 36.1 | 19.4 to 52.5 | 35.8 | 22.9 to 48.5 |
| Ever associated with a venue to sell sex | 35.5 | 25.8 to 46.6 | 66.3 | 37.1 to 133.7* | 68.0 | 44.9 to 95.7 |
*Impossible values.
CRI, credible interval.