We did not find the logarithmicchanges in mortality seen in the studyof Kumar et al, and we hypothesize thatthe reason for this is threefold: 1) in thestudy of Kumar et al, the median timeto appropriate antibiotics was 6 hrs (IQR _ 2–15 hrs), whereas in our study,it was just over 2 hrs (127 mins [IQR,79 –224 mins]);