This paper will review the significance of results obtained by DNA amp
lification methods performed on clinical materials for the detection o
f bacterial pathogens. They will be compared with conventional culture
, antigen detection or serological methods with respect to speed, sens
itivity and specificity. PCR has provided promising results in the ide
ntification of Bordetella pertussis, Chlamydia pneumoniae, Legionella
pneumophila, Streptococcus pneumoniae, Mycoplasma pneumoniae and the v
arious pathogroups of diarrheagenic Escherichia coli. PCR and LCR have
also shown encouraging results when used in the diagnosis of sexually
transmitted diseases caused to Chlamydia trachomatis. In patients wit
h Lyme disease, the sensitivity of PCR is still insufficient, when com
pared to serological methods. Here PCR is an adjunct in the diagnosis
and no substitute for clinical materials have also proved to be both p
roblematic and challenging. Problems in using the PCR include determin
ing the optimal target selection, quantifying the sample volume necess
ary for analysis, determining a standard for sample preparation, and o
ptimizing amplification reactions. There are also difficulties with PC
R inhibitors present in the clinical material and with monitoring the
performance of the technique. PCR results are highly reliable and repr
oducible between laboratories when standardized reagents and protocols
are used. An important step in this direction is the commercial avail
ability of PCR kits. Such kits also simplify the handling of PCR, thus
requiring less technical expertise, and allowing broader use for diag
nosis. In the near future, additional studies must provide a correlati
on between PCR results and conventional methods with larger numbers of
samples. Moreover, as a final evaluation, PCR detection methods must
prove their benefit with respect to clinical management.