Determination and communication of critical findings in neuroradiology

Purpose: The aims of this study were to analyze reporting of critical findings among neuroradiologists in a university setting and to revise a list of critical findings reflecting an academic clinical practice as part of a practice quality improvement project. Materials and Methods: Neuroradiologic studies performed between January 1 and February 28, 2011, containing "critical finding" notations were searched. Reports were matched with an institutionally approved list of critical findings. These findings and unlisted items that were labeled critical were analyzed for frequency, clinical severity, and diagnosis category. The list was revised on the basis of frequency and severity results. Results: A total of 12,607 reports contained 871 critical findings, 608 of which (69.8%) matched the preexisting list. One-third of the findings (263 of 871) labeled critical were not found on the list. Facial, spinal, and calvarial fractures (76 of 263 [28.9%]) and neurovascular injuries (38 of 263 [14.4%]) were the most frequent unlisted findings. A revised list encompassed 86.7% of all communicated neuroradiologic critical findings. Conclusions: Clinician-approved and neuroradiologist-approved standardized sets of critical findings can facilitate the communication of important results without "overcalling" and decreasing efficiency. Physician judgment of what constitutes a critical finding supersedes any such list, as clinical scenarios are highly variable from patient to patient. Critical findings lists require intermittent revision to reflect practice patterns and changing incidence of disease. Such a review can constitute a practice quality improvement initiative.