Effectiveness of a care bundle to reduce central line-associated bloodstream infections

In reply:

We thank Reasbeck and Flockhart for their interest in our article. We were indeed aware of the revised Victorian Healthcare Associated Infection Surveillance System (VICNISS) definition of central line-associated bloodstream infection (CLABSI) and its potential to influence the CLABSI rates. To account for this, we applied the revised definition across the entire study period. Additionally, we independently reviewed all the microbiological data in an effort to identify any missed CLABSIs.

We congratulate Ballarat on its excellent CLABSI record. Our comment that this is the first time that a zero CLABSI rate has been achieved and sustained in an Australian intensive care unit was in reference to publications in peer-reviewed journals; this should have been made clearer. The VICNISS data — like those of the Western Australian Department of Health, where exceptional results have also been reported — do not provide details on methods.

We acknowledge that factors other than the intervention may have influenced our results. However, most of the contemporary CLABSI prevention trials have used an unblinded non-randomised design and were similarly susceptible to “a variation of the Hawthorne effect”.