Dr. Puhan would like us to clarify why we presented results for years 1998 to 2002 in the lower panel of Figure 1. In our primary analysis, we compared outcomes from July 2003 to December 2004 with those from January 2001 to June 2003 and found an improvement in mortality rates of medical teaching patients. However, we were concerned that our findings reflected underlying trends unrelated to the regulations. We tested this by examining whether teaching hospitals improved relative to nonteaching hospitals before the regulations took effect. We found that teaching hospitals did not improve in 2000 to 2001 compared with 1998 to 1999 (relative to nonteaching hospitals), which suggests that outcomes in teaching hospitals were stable before the regulations and had improved after (see lower panel of Figure 1). This makes us more confident about our main results.