Emergency physician perceptions of medically unnecessary advanced diagnostic imaging

The objective was to determine emergency physician (EP) perceptions regarding 1) the extent to which they order medically unnecessary advanced diagnostic imaging, 2) factors that contribute to this behavior, and 3) proposed solutions for curbing this practice.

As part of a larger study to engage physicians in the delivery of high-value health care, two multispecialty focus groups were conducted to explore the topic of decision-making around resource utilization, after which qualitative analysis was used to generate survey questions. The survey was extensively pilot-tested and refined for emergency medicine (EM) to focus on advanced diagnostic imaging (i.e., computed tomography [CT] or magnetic resonance imaging [MRI]). The survey was then administered to a national, purposive sample of EPs and EM trainees. Simple descriptive statistics to summarize physician responses are presented.

In this study, 478 EPs were approached, of whom 435 (91%) completed the survey; 68% of respondents were board-certified, and roughly half worked in academic emergency departments (EDs). Over 85% of respondents believe too many diagnostic tests are ordered in their own EDs, and 97% said at least some (mean = 22%) of the advanced imaging studies they personally order are medically unnecessary. The main perceived contributors were fear of missing a low-probability diagnosis and fear of litigation. Solutions most commonly felt to be "extremely" or "very" helpful for reducing unnecessary imaging included malpractice reform (79%), increased patient involvement through education (70%) and shared decision-making (56%), feedback to physicians on test-ordering metrics (55%), and improved education of physicians on diagnostic testing (50%).

Overordering of advanced imaging may be a systemic problem, as many EPs believe a substantial proportion of such studies, including some they personally order, are medically unnecessary. Respondents cited multiple complex factors with several potential high-yield solutions that must be addressed simultaneously to curb overimaging.