Objectives: To determine the prevalence of sexually transmitted diseases (STDs) and to assess the validity of STD screening approaches among family planning clients in Dar es Salaam, Tanzania.
Methods: Between March and September 1995, information about sociodemographic characteristics, contraceptive use, sexual behaviour, and medical history was obtained from consenting women (n = 908). After interview, blood and genital specimens were collected for diagnosis of STDs and HIV. Based on the information obtained at interview and clinical examination, STD diagnostic algorithms were developed and validated.
Results: The prevalence of STDs was HIV (16.9%), gonococcal and/or chlamydial cervicitis (8.2%), and Trichomonas vaginalis and/or Candida albicans (27.2%). The risk of cervicitis was increased among unmarried women and among women with a husband < or = 25 years of age and women having more than one sex partners in the past 3 months or a new sex partner during the past month. Most women with cervicitis (62.2%) and vaginitis (67.6%) were asymptomatic. A screening strategy for cervicitis based on symptoms had a sensitivity of 29.7%, a specificity of 84.1%, and a positive predictive value (PPV) of 15.9%. The corresponding figures for an algorithm based on clinical signs were 20.3%, 90.2%, and 15.6%. The sensitivity of a simple risk assessment algorithm ranged from 20.3% to 73%. An approach based on both risk assessment (risk score > or = 1) and clinical signs (cervical mucopus and friability) had a sensitivity of 37.8%, a specificity of 87.5%, and a PPV of 21.4%. A syndromic approach for vaginitis resulted in a higher sensitivity than the approach based on the type of vaginal discharge.
Conclusion: Although there is no single screening strategy for cervicitis which can be advocated for large scale application, risk assessment might be the only cost effective strategy for identifying women with cervicitis in family planning clinics in Tanzania.