The article by Petersen and colleagues [1] contains several flaws that raise serious doubts about its conclusion. First, internal medicine residents in a competitive academic center were asked to report adverse events that occurred during hospitalization of patients with whom they were familiar. It is only natural that they would tend to report events that occurred during another team's supervision and to minimize any errors committed either by themselves or the team they were supervising. To do otherwise would put themselves at risk for censure, or at least unwelcome attention.

Second, working conditions experienced by housestaff during times of cross-coverage and non-cross-coverage are not the same. Cross-coverage tends to occur at night, when housestaff have already been working for many hours. The number of patients covered is at least four times greater than the number covered during the day, given that one team covers for the other three. In addition, fewer supervising residents per intern are on duty at night than during the day. To state that discontinuity of care is the only factor contributing to adverse events is to ignore these facts, which are very apparent to the on-call intern.

The authors, who are at an academic institution, benefit indirectly by having housestaff work long hours for wages that are a fraction of their free market value. Academic medical centers would never survive financially if they had to pay interns and residents a fair wage. Thus, the authors' motives must be questioned. Very little has been written about the costs of major depression, suicide, physical illness, and divorce incurred because of requirements that housestaff work 36-hour shifts. I doubt that academicians will study these events, simply because it is not in their best interest to do so.