Do routine narrative style CT reports provide sufficient information to adequately stage pancreatic cancer.

Background: Pancreatic ductal adenocarcinoma (PDAC) is staged using CT imaging generally reported in a narrative prose style, which can omit key details critical for an operating surgeon. This study assessed sixty-three potential radiological descriptors derived from the International Consensus Guidelines (ICG) to classify the resectability status of PDAC to determine whether reports adequately staged PDAC.

Methods: 150 CT reports from patients with confirmed PDAC were collected from 19 hospitals across Australia (July-December 2023), as part of the SCANPatient trial. Reports were evaluated based on how many of the 63 relevant features were addressed.

Results: Less than half all relevant fields were appropriately addressed in prose reports (41.1 %, 10.7 of an average 26 relevant fields, as not all fields were relevant for each case). Less than 35 % of reports adequately addressed the tumour-vessel relationship with the superior mesenteric artery and vein. The coeliac artery was addressed in just 30 reports (20 %), and non-standard and vague terminology was common.

Conclusion: Prose CT reporting for PDAC resectability in Australia does not specifically address many important features determining resectability. Prose reporting is associated with high rates of omission of relevant negatives as well as some errors of commission and considerable language variability.