Algorithm for Ordering Abdominal CT Scans in Patients After Gastric Bypass: Is It Possible?

Luise I Pernar, MD1, Ryan Lockridge, BS, BSN, RN1, Colleen MccCormack1, Minghua Chen, MD, PhD1, Judy Chen, MD2, Scott A Shikora1, David Spector3, Ali Tavakkoli3, Malcom K Robinson, MD1, Ashley H Vernon3. 1Brigham and Women’s Hospital, 2Swedish Medical Center, 3Brigham and Women’s Faulkner Hospital

Introduction: Gastric bypass patients frequently present to the emergency department (ED) with complaints for which abdominal CT (abdCT) scans are ordered. This is done to rule out potentially catastrophic complications such as an internal hernia or anastomotic leak. However, anecdotally, the majority of such scans do not reveal intra-abdominal pathology. Given the health care costs of potentially unnecessary CT scans and risks of radiation exposure, this study was undertaken to determine if clinical and laboratory parameters could be used to develop an algorithm for more rational ordering of abdCTs.

Methods: The study is a retrospective review of patients who underwent primary open or laparoscopic Roux-en-Y gastric bypass (RYGBP) who presented to the ED with complaints potentially related to their RYGBPs regardless of whether an abdCT was ordered.  Data were collected on a variety of symptoms, signs, and laboratory data. All abdCT scan results were reviewed and classified as normal or abnormal. We searched for parameters that could reliably predict a normal or abnormal CT scan.

Results: 1643 primary RYGBs were performed at our institution between 2005 and 2015. 355 patients (22%) had a total of 675 ED visits; during 390 visits (58%) abdCT scans were performed. Of the total number of abdCT scans performed 244 (63%) were normal and 46 (12%) demonstrated intra-abdominal pathology that required surgical intervention, related to prior bariatric surgery in 39 (10%). The remaining 25% required either endoscopy, IR intervention, or other therapy. A history of abdominal pain was sensitive (0.98) but had poor specificity (0.03) for predicting whether an abdCT scan would be normal or abnormal. Abdominal tenderness had poor sensitivity (0.71) and specificity (0.24) for predicting the results of an abdCT scan. Other factors including any level of temperature, heart rate, blood pressure, WBC, lactic acid level, and lipase were highly insensitive and non-specific and could not be used to reliably predict abdCT scan results.

Conclusions: RYGB patients frequently have abdCT scans when they present to the ED but are infrequently found to have intra-abdominal pathology upon scanning. Despite examining a wide variety of sign, symptoms and lab data we could not identify a parameter which could reliably delineate those for whom it is safe to forgo an abdCT. ED evaluation and imaging decisions must be based on astute clinical judgment and do not appear to be amenable to a simple algorithm. 

« Return to SAGES 2016 abstract archive