Search Annals: Advanced search

Rapid Responses to:

Articles: Kanaka D. Shetty and Jayanta Bhattacharya Changes in Hospital Mortality Associated with Residency Work-Hour Regulations Ann Intern Med 2007; 147: 73-80 [Abstract] [Full text] [PDF]

Send comment/rapid response letter

Electronic letters published:

In Response

Kanaka D Shetty, Jayanta Bhattacharya, Stanford University   (16 October 2007)

What about the years from 1998 to 2002?

Milo A Puhan   (7 June 2007)

In Response 16 October 2007
Kanaka D Shetty, MD, MS Stanford University & Veterans Affairs Palo Alto, Jayanta Bhattacharya, Stanford University Send rapid response to journal: Re: In Response kshetty{at}stanford.edu Kanaka D Shetty, et al.	Dr. Puhan would like us to clarify why we presented results for years 1998-2002 in the lower panel of Figure 1. As Dr. Puhan noted, in our primary analysis, we compared outcomes in July 2003-December 2004 to those from January 2001-June 2003 and found that mortality improved in 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 non-teaching hospitals before the regulations took effect. We found that teaching hospitals did not improve in 2000-2001 compared to 1998-1999 (relative to non-teaching hospitals), which suggests that outcomes in teaching hospitals were stable prior to the regulations and improved after their implementation. (See lower panel of Figure 1). This makes us more confident about our main results. Conflict of Interest: None declared
What about the years from 1998 to 2002? 7 June 2007
Milo A Puhan, MD PhD University of Zurich Send rapid response to journal: Re: What about the years from 1998 to 2002? milo.puhan{at}usz.ch Milo A Puhan	Doctors Shetty and Bhattacharya present a careful analysis of the impact of the new regulations on work hours for residents in the United States. While reporting is excellent in general there is a major inconsistency in the reporting of the years. As I understand, the period before July 2003 was compared with the period after July 2003. In figure 1 and in the description of the sensitivity analyses, however, there is a description of the years from 1998 to 2002. Did the authors use these years to test assumptions about the time when regulatory changes took place (although not described in the methods section) or is it simply a typing error? I would appreciate a short clarification. Yours sincerely, Milo Puhan, MD, PhD Horten Centre for patient-oriented research University Hospital of Zurich Postfach Nord CH-8091 Zurich Switzerland Conflict of Interest: None declared