Background   Many specialty societies have found that neuroimaging in headache is a   low-value intervention for benign presentations. This study describes factors   that influence Emergency Room (ER) physicians' adherence to Choosing Wisely   (CW) recommendations for low-risk headache patients presenting to Calgary's   Emergency Departments (EDs). Emergency medicine has yet to address imaging in   headache as a CW topic; however, this study may inform that decision. 

Methods   Data were retrospectively collected for all patients presenting to Calgary   EDs with headaches from April 1, 2014 to March 31, 2016. Patients were deemed   low-risk by virtue of discharge home from the ED, age < 50, and no lumbar   puncture (LP), trauma, neurology, or neurosurgery consult or red flags on   history. The primary outcome was computed tomography (CT) ordering rates with   an eye to medical doctor (MD) practice variation. Patient, physician, and   environmental factors were analyzed to compare patients who did and did not   receive a CT. 

Results Two thousand seven hundred and thirty-four headache   patients met the eligibility criteria. A total of 117 Calgary ER physicians   were included, all of whom had seen 10 or more headache patients over the   study period. Physician practice variation was vast, with a mean ordering   rate of 38.0% and a range of 0% to 95% (M = 39.0%, IQR = 21.0%). CTs were   ordered more often in males than females (39.9%; 34.1%; p = 0.002) and in   patients presenting during the day and evening (38.1%; 39.0%) compared to the   night (29.7%; p < 0.001). Patients were divided into quartiles by age,   with the oldest group (41.6 - 50 years) receiving significantly more head CTs   (45.1%) than the other quartiles (34.9%; 34.9%; 27.5%; p < 0.001). Longer   triage-to-discharge times were associated with an increase in CT ordering   rates (12% for < 2.95 hours; 35% for > 4 hour wait; p < 0.001).   Lastly, patients who did not have a CT were more likely to revisit the ED   within seven days compared to those who did (6.9% vs 4.0%; p = 0.003), but   their seven-day admission rate was unaffected (0.6% in the group that got CTs   and 0.3% in the group that did not get a CT). Time to assessment, the day of   the week, physician gender, years of experience, and training program did not   influence CT ordering practices. 

Conclusion To our knowledge, this is the   first study to assess how patient, physician, and environmental factors   relate to the use of CT scans in low-risk headaches presenting to the ED. CW   guidelines are not optimally adhered to, and the findings in this study   findings may inspire new ideas for maximizing the judicious use of healthcare   resources.

Authors: Meaghan J Mackenzie, Rashi Hiranandani, Dongmei Wang, Tak Fung, Eddy Lang