Early detection of bacteria in bloodstream infections using molecular diagnostics: implications for patient care

Bloodstream infections (bacteraemia) have a mortality of up to 25%. Starting
appropriate therapy empirically has been found to reduce mortality. Therefore,
(combinations of) broad spectrum antibiotics are used. However, broad spectrum
antibiotics have the disadvantage of not covering all bacteria, selection of
antibiotic resistant bacteria and increased drug toxicity.

In this study we aim to determine whether the use of a novel molecular
technique (real-time PCR) decreases the time during which empirical
broad-spectrum antibiotics are used in comparison to the conventional
identification techniques of blood cultures.

Randomised, blinded clinical trial

After randomisation, test results obtained by either the PCR test or the
conventional methods will be conveyed and antibiotic therapy will be adjusted
accordingly. Demographic and follow-up data will be monitored in a standardised
way.

The burden from participation in this study is very low, no extra blood has to
be drawn and in order to obtain information about the patient the chart will be
used. Patients randomised for the conventional method (the control group) will
experience no benefits or disadvantages from their participation in this study.
Patients randomised for the PCR-based test may profit from participation, since
they can be treated with an adequate antibiotic sooner. The risk of the new
test being less accurate than the conventional method is small. We found that
the agreement of the PCR-based test with the results of the gold standard is
95% for Gram-negative rods and 96% for Gram-positive cocci. The agreement of
the conventional method with the gold standard is about 97%. To minimise the
risk of a false diagnosis, blood cultures in the group of the PCR-based method
will also be tested with the conventional method, so treatment can be adjusted
in case of discrepant results.