SETTING: City of Manaus, Amazonas, Brazil. OBJECTIVE: To explore the relationship between positivity to tuberculin and other environmental mycobacteria sensitins, according to a range of criteria and presence of BCG scar. DESIGN: Dual skin testing with tuberculin and four mycobacterial sensitins, and BCG scar recording of 1070 schoolchildren aged 7-14. Four criteria for positivity were used: simple and dominant, with 5 and 10 mm cut-off points. RESULTS: The standardised prevalence of reactions > or = 5 mm for BCG scar negative children was 58.3% for Mycobacterium avium, 54.2% for M. scrofulaceum, 26.8% for M. fortuitum, 17.9% for M. tuberculosis and 7.6% for M. kansasii. Correlations between tuberculin and each sensitin, for BCG scar negative children, were 0.47 for M. avium, 0.53 for M. scrofulaceum, 0.60 for M. kansasii and 0.22 for M. fortuitum (all with P < 0.01). BCG effect was particularly significant for tuberculin (odds ratio = 3.44 for reactions > or = 5 mm, P < 0.001) and influenced the balance between dominant/non-dominant reactions for all sensitins. CONCLUSION: The correlation between tuberculin and each sensitin confirmed the separation of the rapidly (M. fortuitum) and slowly growing mycobacteria (M. tuberculosis, M. avium, M. scrofulaceum and M. kansasii). The influence of BCG on tuberculin reactions was more marked than on other mycobacterial sensitins.
SETTING: City of Manaus, Amazonas, Brazil. OBJECTIVE: To explore the relationship between positivity to tuberculin and other environmental mycobacteria sensitins, according to a range of criteria and presence of BCG scar. DESIGN: Dual skin testing with tuberculin and four mycobacterial sensitins, and BCG scar recording of 1070 schoolchildren aged 7-14. Four criteria for positivity were used: simple and dominant, with 5 and 10 mm cut-off points. RESULTS: The standardised prevalence of reactions > or = 5 mm for BCG scar negative children was 58.3% for Mycobacterium avium, 54.2% for M. scrofulaceum, 26.8% for M. fortuitum, 17.9% for M. tuberculosis and 7.6% for M. kansasii. Correlations between tuberculin and each sensitin, for BCG scar negative children, were 0.47 for M. avium, 0.53 for M. scrofulaceum, 0.60 for M. kansasii and 0.22 for M. fortuitum (all with P < 0.01). BCG effect was particularly significant for tuberculin (odds ratio = 3.44 for reactions > or = 5 mm, P < 0.001) and influenced the balance between dominant/non-dominant reactions for all sensitins. CONCLUSION: The correlation between tuberculin and each sensitin confirmed the separation of the rapidly (M. fortuitum) and slowly growing mycobacteria (M. tuberculosis, M. avium, M. scrofulaceum and M. kansasii). The influence of BCG on tuberculin reactions was more marked than on other mycobacterial sensitins.