Methods:

This was a national survey of adult hospitalists in the United States. The authors developed and administered a questionnaire that included clinical vignettes for two conditions: preoperative evaluation and management of syncope. Vignettes contained the same clinical presentation, but differed in the following two circumstances: whether a family member requested additional care and whether the family member’s occupation was provided (son identified as a physician or wife identified as an attorney). Treatment choice within each clinical vignette was the primary outcome, which was analyzed as a dichotomous variable (overuse vs. appropriate use). Overuse was defined as more testing than required by current guidelines. For preoperative testing, any response other than “no further testing” was considered overuse; for syncope, any response that included “hospitalization” constituted overuse. Respondents were also asked to provide “the most likely primary driver of the hospitalist’s decision” (in particular, whether the practice choices resulted from a desire to follow scientific guidance or some other motivation).