Optimizing detection of postoperative leaks on upper gastrointestinal fluoroscopy: a step-by-step guide.

Postoperative leaks after gastrointestinal surgery are important to identify to decrease patient morbidity and mortality. Fluoroscopic studies are commonly employed to detect postoperative leak. While the literature addresses the sensitivity and specificity of these examinations, there is generally a lack of description of the fluoroscopic technique itself and there may be variability between radiologists in how these studies are performed. It is important to balance a standardized fluoroscopy protocol while tailoring the exam for each surgical and patient situation. Here we will briefly review common postoperative anatomy in the upper gastrointestinal tract, propose fluoroscopic techniques to improve postoperative leak detection, and illustrate teaching points with clinical cases.

Quantifying Language Before and After Instituting Structured CT Reports.

AIM: Our aim was to quantify the language in abdominopelvic CT reports before and after implementing structured reporting templates, specifically evaluating the terms normal and unremarkable. METHODS: We performed a search of our database for all CT abdomen and pelvis examinations before and after full template implementation. There were 1,753 studies included-878 "pretemplate" reports, 875 "posttemplate" reports. Pre- and posttemplate reports were analyzed for word count in preliminary report, word count in final report, number of changed words between preliminary and final report, and frequency of normal and unremarkable. The averages of each data point for both pre- and posttemplate reports were calculated and compared. RESULTS: Between all pre- and posttemplate reports, the total number of words and number of changes between preliminary and final reports was not significantly changed. The frequency of the word normal increased from an average of 5.29 (pretemplate) to 8.92 (posttemplate) per report (P < .0001). The frequency of the word unremarkable increased from an average of 0.11 (pretemplate) to 0.22 (posttemplate) per report (P < .0001). CONCLUSIONS: When comparing language before and after implementing structured reports, we found an increase in the frequency of normal, without any significant change in number of words per report. The increase in normal in posttemplate reports reflects a trend toward more definitive reporting, which may increase clarity and satisfaction among referring physicians. The increased use of the less definitive term unremarkable identifies an area in need of further training and improvement in our department.