RT Journal
A1 Lin MY, Weinstein RA, Trick WE
T1 SUrveillance quality in reporting nosocomial bloodstream infection rates—reply
JF JAMA
JO JAMA
YR 2011
FD February 23
VO 305
IS 8
SP 779
OP 780
DO 10.1001/jama.2011.158
UL http://dx.doi.org/10.1001/jama.2011.158
AB 
In Reply: Drs Cohen and Benenson highlight challenges in assessing central line–associated BSI rates, particularly with respect to denominator measurement. Central line–associated BSI rates are calculated with a denominator of central-line days, which provides important risk adjustment among hospitals that differ in patient severity of illness.1 The denominator reflects “at-risk” days, and theoretically patients with infected central lines should be removed from the denominator.2 However, many patients have more than 1 central line, and these patients remain eligible for additional central line–associated BSIs beyond the first event. Since enumerating central-line days is already labor intensive at most hospitals and the days are often aggregated to the unit level, it is unlikely that achieving a marginally more accurate denominator would meaningfully impact central line–associated BSI rates. For example, since on average less than 1% of central-line days incur an infection, the adjustment would be minimal.