Detecting cervical infection among family planning clients: difficulties at the primary health-care level in Indonesia.

Abstract

In developing and testing an operational model for the integration of reproductive tract infection/sexually transmitted disease (RTI/STD) management into existing family planning (FP) services in Indonesia, this study allowed for assessment of disease prevalence and evaluation of diagnostic methods for detection of endocervicitis caused by chlamydial infection and/or gonorrhoea. Data were collected over 28 weeks in 1997 at 2 FP clinics in the low-income harbour neighbourhood of North Jakarta. Among 486 consenting female FP clients, prevalence of chlamydial infection was 9.3%, gonorrhoea 1.2%, trichomoniasis 4.5% and syphilis 0.8%. Clinically observed abnormal vaginal discharge, cervical inflammation and vaginal lesions/ulcers were all associated with cervical infection (P<0.05), but insufficiently sensitive (<60%). Clinical diagnosis for cervical infection had 48.8% sensitivity, 75.4% specificity, but only 18.3% positive predictive value (PPV). On-site Gram stains for gonorrhoea were 83.3% sensitive and 94.5% specific, but had only 16.1% PPV. Presence of mucopurulent cervicitis was only 39.6% sensitive for cervical STD, with PPV of only 16.3%. Development of an affordable and accurate detection tool for chlamydial infection remains the main obstacle to effective RTI/STD management in this population.