Abstract

BACKGROUND--BCG vaccination alters the response to tuberculin testing and influences the potential validity of the Heaf test in the diagnosis of tuberculosis. This study used a purified protein derivative 1 tuberculin unit (TU) Mantoux test with a cut off of 5 mm induration as an indicator of tuberculous infection in high risk children to determine whether this would distinguish infection from previous neonatal BCG vaccination. METHODS--Children at high risk of tuberculosis on chest radiography, Heaf test, or contact history who had been screened in the contact tracing clinic and referred for further assessment were included in the study. After clinical examination, chest radiography, and Mantoux testing they were assigned to three groups (tuberculous disease, chemoprophylaxis, or no treatment) and followed up for 6-24 months in the outpatient clinic and subsequently by postal questionnaire. RESULTS--Comparison of the Heaf and Mantoux tests showed a difference in the results with 82% of cases positive by the Heaf test and 59% positive by the Mantoux test. Using the Mantoux test result in combination with clinical and radiographic findings 194 children were allocated to the three groups as follows: primary tuberculosis (5), chemoprophylaxis (101), no treatment (88). During follow up for a mean (range) time of 46 (11-102) months four additional cases received treatment for primary tuberculosis, two in the chemoprophylaxis group and two in the untreated group. CONCLUSIONS--The use of the 1 TU Mantoux test after neonatal BCG vaccination reduced the number of children receiving treatment from 129 to 93-that is, by 36%. Although the numbers are small, there was no increase in the later development of tuberculosis.