 |
 |
 |
 |
 |
 |
 |
 |
 |
||||
|
An evaluation of the syndromic approach to identifying Chlamydia trachomatis, Neisseria gonorrhoeae, and Trichomonas vaginalis among HIV-infected pregnant women in Botswana Authors: Carolyn Wester, Ibou Thior, Lisa Stocking, Shahin Lockman, Roger Shapiro, Soyeon Kim, Trevor Peter, and Max Essex. Background: In 1992 Botswana implemented a symptom-driven syndromic approach to identify and treat non-HIV sexually transmitted infections (STIs). We evaluated the usefulness of the syndromic approach for identifying Chlamydia trachomatis (CT), Neisseria gonorrhoeae (NG), and Trichomonas vaginalis (TV). Methods: HIV-infected pregnant women at 1 urban and 3 rural sites in Botswana were evaluated for STIs at 34 weeks of gestation. All participants were screened for STIs by symptoms (pelvic pain, vaginal discharge); physical exam using a speculum (PE) (abnormal vaginal or cervical discharge, cervical pain, or pelvic pain) and laboratory testing for CT (PCR), NG (PCR), and TV (wet prep). Results: From August 1, 2001 to January 3l, 2003, 717 HIV-infected pregnant woman were evaluated, and 707 (98.6%) had laboratory testing performed for at least one pathogen. The laboratory prevalence of CT was 62/636 (9.8%), GC 31/635 (4.9%), TV 149/665 (20.8%), and any of the three 207/707 (29.3%). One hundred sixteen (16.4%) women had symptoms of an STI. Per the syndromic approach, 103/707 (14.6%) had positive symptoms confirmed by PE. The sensitivity and specificity of the syndromic approach was 12.6% and 84.6%, respectively, with a positive predictive value (PPV) of 25.2% and a negative predictive value (NPV) of 70.1%. However, 378/589 (64%) of asymptomatic women also had positive findings on PE. Using PE alone in all women, 481/705 (68.0%) had positive findings suggestive of an STI. The sensitivity and specificity of PE findings alone were 88.3% and 40.1%, respectively, with a PPV of 37.8% and a NPV of 89.3%. Conclusions: A symptom-driven syndromic approach is less sensitive for identifying CT, GC, and/or TV among HIV-infected pregnant women than PE using speculum alone, although the potential for over-treatment with PE alone remains a concern. PE using speculum should be considered as an alternative to the syndromic approach for detecting STIs in areas where laboratory-based screening is not feasible. |