Tuberculosis control programmes are faced with an increased burden of
cases, a shift towards diagnostically more difficult categories of pat
ients such as extrapulmonary and smear-negative cases, and the emergen
ce of multidrug-resistant strains of M. tuberculosis. Improved diagnos
is would be a valuable contribution in the struggle to solve this glob
al public health emergency. Nucleic acid amplification reactions promi
se to reduce the time for diagnosis from weeks to hours, while surpass
ing the sensitivity and specificity of the classical methods. Besides
their potential value in diagnosis, amplification reactions offer the
possibility of a rapid identification and drug susceptibility determin
ation. Their application in other fields, such as epidemiology, could
benefit the control of mycobacteriosis indirectly. Although critical a
ssessment is warranted, several features, such as the availibility of
standardized kits, strongly justify the evaluation of amplification re
actions for the detection of mycobacteria for use in the clinical labo
ratory. In fact, evaluation of the technology now, at its present leve
l of development, even in an endemic setting in a developing country,
would yield useful recommendations towards its improvement for applica
tion in different settings, in terms of efficiency, simplicity and cos
t.