In reply
I agree with Drs Ravakhah and Kumar's criticism regarding my study.1 As I pointed out, this was an observational study, not a prospective, randomized study; therefore, undetected differences in the severity of illness could have been introduced between the 2 groups. In choosing to focus on differences in antibiotic use between the 2 groups rather than in the training backgrounds of the hospitalists (ID vs IM), I wanted to avoid, as I phrased it, an "us vs them" posture. It might have been, however, that the differences in training were what led to the differences in antibiotic use, which in turn affected outcomes of the groups.