Objective: To evaluate the effectiveness of the syndromic approach in management of vaginal infections among pregnant women in Moshi urban, Tanzania.
Methods: Between June 2002 and March 2004, 2654 pregnant women in their 3rd trimester were recruited from two community primary health care clinics. They were interviewed using a structured questionnaire, examined and genital samples collected. Vaginal secretions were evaluated for bacterial vaginosis (BV) by Amsel's criteria, for trichomoniasis (TV) and candidiasis (CA) by wet mount microscopy. Sensitivity, specificity, and positive and negative predictive values were used to assess the effectiveness of syndromic management for vaginal infections.
Results: The prevalence of TV, BV and CA were 5.0%, 20.9% and 11.4% respectively. Nearly 70% of the vaginal infections with trichomoniasis and/or bacterial vaginosis were asymptomatic. Self reported vaginal discharge or the presence of vaginal discharge during examination were poor predictors of vaginal infections with a sensitivity of 29%-54% and 26%-50% respectively in detecting TV and BV. Using them would lead to under-diagnosis of those with these infections and overtreatment of those without the infections.
Conclusions: Diagnosis and treatment of TV and BV among pregnant women remains a challenge in this setting. The current symptom-directed treatment would miss a substantial proportion of TV and/or BV infections because of the poor performance of the test, and because most infections were asymptomatic or unrecognized. The introduction of simple, point of care laboratory screening tests for vaginal infections into routine antenatal care at primary health clinics with laboratory facilities should be considered.