Validation of national algorithms for the diagnosis of sexually transmitted diseases in Brazil: results from a multicentre study.

OBJECTIVE: To validate STD flow charts for the management of genital discharge and genital ulcer currently recommended by the National STD Control Programme in Brazil.
METHODS: A study was conducted in five Brazilian STD clinics from January to June 1995. After an interview, a clinical examination was performed by a physician, who recorded a presumptive diagnosis, based on his/her clinical experience. This diagnosis was compared with a gold standard laboratory diagnosis in order to calculate sensitivity, specificity, and positive predictive value of the clinical diagnosis. The validity of the simulated national flow charts was assessed using the same method.
RESULTS: A total of 607 men and 348 women participated in the study. Gonorrhoea was the aetiology most frequently detected in men with urethral discharge. The sensitivity of the clinical diagnosis was far lower than the sensitivity fo the national flow chart, using the syndromic approach, for both gonococcal and chlamydial urethritis. Adding a simple laboratory test (Gram stain) to the national flow chart increased the specificity and positive predictive value for gonorrhoea. Among the women with vaginal discharge, a cervical infection was detected in 17%, a vaginal infection in 74%, and mixed infection in 9%. The sensitivity of the diagnosis for cervical infection increased from 16% (clinical aetiological approach) to 54% (when adding a syndromic approach) and to 68% when adding a risk assessment, as in the national flow charts. The cure or improved rate of genital ulcers was 96% after 1 week.
CONCLUSIONS: The results of the study will help to convince policy makers and those involved in training healthcare workers in Brazil of the public health advantages of the syndromic approach, as an essential part of STD/HIV control activities.