Friday, February 26, 2010

Preventing nosocomial bloodstream infections in children

As is the case in many areas of health care, there is less published literature on nosocomial infection prevention in children than in adults. So it is nice to see some multicenter data on central-line associated bloodstream infection (CLABSI) prevention in kids. The bottom line: in 29 pediatric intensive care units (PICUs) across the U.S., implementation of bundles for line insertion and maintenance resulted in a 43% reduction in CLABSI rate (from 5.4 to 3.1 infections per 1000 central line days). One of the interesting findings from this study was that adherence to the “maintenance bundle” seemed to be more important for CLABSI prevention than did adherence to the “insertion bundle”. See here for what each bundle included.


  1. There must be a mistake. It doesn't appear that they were able to get their infection rates to zero. Perhaps if they stopped taking care of really sick patients?

  2. I recently learned that some hospitals with very low rates of CLABSIs have modified the CDC definition to exclude those which meet the definition but are probably not line-related in reality (e.g., a positive VRE blood culture in a neutropenic cancer patient with no VRE infection at another site). This raises two issues: (1) CDC needs to actively improve the specificity of HAI definitions, and (2) we need validation to ensure that all hospitals are applying the definitions in exactly the same manner.

  3. I've often wondered if the rigor CMS applies to process measures (hospital audits etc) wouldn't be better for CLA-BSI than antibiotic timing for pneumonia or other measures of questionable impact.

  4. You were right Eli. I think that there could be something wrong. I just hope that they are applying the definitions in exactly the same manner. Thanks for the share. Really learn something from this blog. keep it up!


    disinfecting cap