The effectiveness results were as follows:
In 68% of patients (n=36), the initial and reinterpretation reports agreed.
Of the 16 patients considered to have resectable tumours by both reports, 81% (n=13) underwent successful Whipple's procedures.
32% of patients (n=17) had discrepancies between the initial and reinterpretation reports.
The Kappa value was 0.42, indicating moderate overall agreement between reports.
All discrepancies involved the initial report's indicating resectability and the reinterpretation report's being consistent with unresectable disease.
The discrepancies were resolved by the findings at surgery (n=9), percutaneous biopsy (n=3), dedicated pancreatic CT (n=3), dedicated liver CT (n=1), and follow-up abdominal CT (n=1).
The reinterpretation reports were correct in 94% of the 17 patients (n=16) with discrepancies. The reinterpretation was found to be wrong in one case by dedicated liver CT, leading to a successful Whipple?s procedure. Surgery was therefore averted in 13% of patients (n=7).