Breaking down the CLABSI bundle

There is an interesting CLABSI prevention study out this week in PLoSONE (full disclosure: tireless co-blogger and pal Eli is one of the authors). In order to determine how well the CLABSI bundle was being implemented in U.S. ICUs, and which individual elements (or subsets of elements) were most strongly associated with CLABSI reductions, the authors surveyed NHSN hospital practices. They used quarterly ICU-specific CLABSI rates as their outcome measure.

The bottom line: the bundle was associated with lower CLABSI rates only for units that monitored and reported high rates of compliance with at least one element of the bundle out of three (maximum sterile barrier precautions, optimal site selection, and daily assessment of need).

Why would meticulous adherence to any one of these three bundle elements be significantly associated with CLABSI reduction, while meticulous adherence to all the elements was not? I believe it was a simple power issue: too few ICUs (only 38%) had high rates of adherence to the whole bundle. Lack of power could also explain why no single element was statistically-significantly associated with CLABSI reduction.

This study is a good first step toward “breaking down the bundle”, to determine which elements are most important for infection prevention, what compliance measurements are most useful, and (eventually) what components should be added, or subtracted, from existing bundles.