Article Text


Epidemiology poster session 2 : Population: Commercial sex worker
P1-S2.17 Modelling the impact of rapid test syphilis screening among female sex workers in China
  1. Kate Mitchell,
  2. Andrew Cox
  1. LSHTM, London, UK


Background In China, female sex workers (FSWs) are hard to reach for public health interventions. STIs and syphilis (Tp) in particular, are now a major public health concern in China. Currently, diagnosis of active Tp requires laboratory tests, which have high sensitivity, are capable of diagnosing active Tp infection, but involve delays between testing and treatment. Such delays lead to a low proportion of FSWs returning for treatment. In contrast, rapid Tp tests can provide instant results, thereby offering the possibility of immediate treatment, but are generally lower sensitivity and cannot differentiate between active and previous Tp infection. In this study, we estimate the impact of using rapid tests to screen FSWs for Tp in a Chinese setting and compare the performance with lab-based diagnosis.

Methods A deterministic model of Tp transmission among FSWs and clients was parameterised and fitted to data from the medium prevalence setting (Kaiyuan City). Random sampling of parameter uncertainty ranges was used to find multiple fits. Model fits were used to explore the epidemiological impact of providing rapid Tp testing and treatment for FSW at different annual coverage levels for 3 years, in comparison to using a high sensitivity lab-based test. Because of evidence for Tp rebounding following treatment, the impact of the intervention after it has ended was evaluated.

Results The model projected the 3-year impact of annually screening 40% or 80% of FSWs with a rapid Tp test (sensitivity 87%) would result in a 46% (43–54%) or 69% (66–77%) relative reduction in Tp prevalence in this setting, respectively. The corresponding decrease in Tp prevalence among clients was estimated to be 2 to 3-fold less but more infections were averted among them. The Asbtract P1-S2.17 figure 1 shows that for a lab-based testing strategy to achieve the same impact as using a rapid test with 87% sensitivity would require a return rate of 92% or 98% if the delay in treatment was 1 or 2 weeks, respectively. After screening ceases, the model projects the percentage of infections averted would continue increasing for a number of years in this setting.

Asbtract P1-S2.17 Figure 1

Required return rate for lab-based gold standard test to have same impact as a rapid test for different delays in treatment. Different lines are for different delay periods between testing and treatment for the lab test - thick line, 0 weeks; dashed line, 1 week; dotted line, 2 weeks; narrow line, 4 weeks.

Conclusions In a moderate Tp prevalence setting, a rapid test point of care intervention could have considerable impact on Tp transmission. Unless there is very low sensitivity for the rapid test and a very high return rate for the lab-based method of screening, the rapid test intervention will provide a greater relative reduction in Tp prevalence.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.