‘R/O path’ – a quality assurance study of emergency department medical imaging requests during on-call hours
摘要
To assess the appropriateness of emergency department (ED) on-call medical imaging requests and the corresponding radiology reports.
MethodsTwo researchers reviewed all CT, MRI, and ultrasound scans ordered from January 1 to December 31, 2022 by ED physicians during on-call hours at a major Canadian hospital network (estimated 2.5 million catchment population). Scan indications were evaluated for: (1) containing any clinical history and (2) providing a clinical question and/or specific differential diagnosis. Radiology report impressions were categorized as either: (1) normal (including chronic and/or stable findings); (2) positive for the clinical question and/or differential diagnosis provided; (3) positive for a condition that could reasonably explain the provided clinical history/question; or (4) only containing incidental findings unrelated to the indication. Inter-rater reliability for the scoring was assessed with the κ statistic.
ResultsAmong 85,066 scans ordered during ED encounters in 2022, 18,201 met inclusion criteria including 248 (1.4%) MRI, 3587 (19.7%) US, and 14,366 (78.9%) CT scans. Most requests (74.2%) provided clinical history and a differential diagnosis; a minority provided only clinical history (15.0%) or only a clinical question/differential diagnosis (10.7%). Most imaging studies contained only normal or chronic findings (62.0%). Approximately 12.5% of studies were positive for the specified differential diagnosis, 4.1% were positive for a condition that could reasonably explain the provided clinical history/question, and 21.3% identified unrelated incidental findings.
ConclusionThe high percentage of negative scans may be an unavoidable consequence of modern medicine’s growing dependence on imaging. However, ensuring that imaging tests are appropriately ordered and automating routine steps of the image acquisition workflow may mitigate strains on radiology departments.
Graphical Abstract