Reproductive-tract infections in women in low-income, low-prevalence situations: assessment of syndromic management in Matlab, Bangladesh

In the control of reproductive-tract infections, including sexually transmitted infections (STIs), in low-income and middle-income countries, WHO recommends syndromic management for individuals with symptoms. This intervention was initially developed in areas where prevalence of such infections is high. We investigated the clinical effectiveness and cost of this approach among a group of women with a low prevalence of infection. During a 5-month period, we investigated all women complaining of abnormal vaginal discharge and seeking care at maternal and child health/family-planning centres in Matlab, Bangladesh, for the presence of laboratory-diagnosed reproductive-tract infections and STIs. Syndromic diagnoses made by trained health-care workers were compared with laboratory diagnosis of infection. We then calculated the costs of treating women by means of the recommended WHO algorithm and an adapted algorithm incorporating use of a speculum and simple diagnostic tests. The prevalence of endogenous infections among 320 women seen was 30%. Cervical infections (Neisseria gonorrhoeae and Chlamydia trachomatis) were found in only three women. The WHO algorithm had a high sensitivity (100%) but a low specificity (zero for bacterial vaginosis, candida, and Trichomonas vaginalis). The speculum-based algorithm had a low sensitivity (between zero and 59%) but a higher specificity (79–97%). Between 36% and 87% of costs would have been spent on uninfected women. The high rate of overtreatment in the population studied carries both financial and social costs- the latter in potentially exposing women misdiagnosed as having an STI to threats of domestic disruption or even violence. We make recommendations for management programmes in areas of low STI prevalence and low income.