To the Editor.
—Dr Rosenthal and colleagues1 should be congratulated for undertaking a scientific study of an important real-life issue—quality of care at teaching hospitals. Although, as an educator, I would be thrilled if their conclusion that "hospital performance may be higher in major teaching hospitals" is correct, as a researcher, I have serious concerns about their methods.My major concern is with the trichotomization of hospitals into nonteaching, major teaching, and minor teaching based on membership in the Council of Teaching Hospitals (COTH). It seems that this is not a valid measure of the role of teaching in an institution, since it depends on paid membership. The cost for a hospital to join COTH for the 1997-1998 academic year is $12 500. Hospitals, especially in these tight economic times, may decide not to join COTH because of this expense. Although it is true that the 400 COTH member institutions (of the