An audit of local discrepancy rates in acute abdominal CT: does subspecialist reporting reduce discrepancy rates?

Abstract

To assess the discrepancy rates (DR) for patients undergoing abdominopelvic computed tomography (CT) for acute non-traumatic abdominal pain who have a subsequent emergency laparotomy in a large university teaching hospital, in particular identifying the differences between subgroups of reporters, to assess factors that may influence the discrepancy rates, to examine the pathologies with the highest discrepancy rate, to identify learning points, and give recommendations on current practice. The surgical data and CT reports of 1,176 patients who underwent urgent laparotomy after CT from 2014-2018 in a large university hospital were analysed retrospectively. A major discrepancy was defined as an error of fact in the radiology report, which led to incorrect management or patient harm. Registrars have higher DR than consultants (6.86% versus 2.77%). The major DR for consultants met national standards (<5%). The major DRs for registrars met the national audit standard (<10%), but not the National Emergency Laparotomy Audit (NELA) standard (<5%). When comparing between reporter subgroups, gastrointestinal (GI) radiologists have a lower major DR than general radiologists (1.22% versus 3.44%). GI radiologists were also found to correct more registrar provisional reports. The existence of a documented preoperative discussion between radiologists and surgeons was associated with a lower DR. DR for registrars and consultants are below the national audit standard. Several factors associated with a lower DR in acute abdominopelvic CT were also identified, including reporting by consultants, reporting by GI radiologists and preoperative discussions between the radiologist and surgeon.