The Utility of CT Imaging in Abdominal Pain Patients Presenting to the ED: A Tertiary-Care Center Experience

Abstract

Introduction: Recent studies highlighted the overuse of diagnostic imaging, especially computerized tomography (CT), with its negative impact on cost, radiation exposure and potentially unnecessary secondary studies due to incidental findings. The goal of this study was to identify factors contributing to decisions on CT scanning in the emergency room (ER). Methods: We used an independent broker system to obtain consecutive records of patients seen in the ER of a tertiary care hospital, with abdominal pain as the primary complaint. Records included presenting complaints, nature of comorbidities, number of ER visits, hospitalizations and CT scans during the 6 months before and after the index visit, laboratory testing, and radiographic imaging performed during the ER encounter, CT findings and patient disposition. Results: A total of 169 encounters were retrieved from March 2014. There were 45.6% (n=77) males; age (46.2 ± 18.1 years). In 66 cases, an abdominal CT was obtained, which showed potentially relevant findings in 43 patients. In a bivariate analysis, undergoing a CT scan correlated with frequent ER visits, chronic pain syndromes, cardiac comorbidities, antidepressant use, fever, acute rather than chronic pain, opioid administration in the ER and leukocytosis. Using logistic regression analysis, only the use of opioids at the time of presentation to the ER functioned as an independent predictor (OR 2.53 [95% CI: 1.09-5.84]). Consistent with prior studies, the opioid use in the ER was not predicted by pain intensity or acuity, but by the use of antidepressants (OR 3.04; 95% CI: 1.06-8.76]) or chronically prescribed opioid analgesics (OR 2.93 [95% CI: 1.25-6.85]). Patients undergoing CT scans in the ER were more likely to be admitted rather than discharged to home. Conclusion: About one third of patients presenting with abdominal pain to a large urban ER underwent CT scanning as part of the evaluation. While the overall diagnostic yield exceeded 50%, availability of CT scans obtained within the last 6 months prior to the incident visit did not correlate with testing. Only the decision to use opioids in the ER functioned as independent predictor of CT scanning. As opioid administration was most closely associated with chronic use of chronic prescription opioids and antidepressants, the results suggest that algorithms with close review of recent imaging data and integration of psychiatric and prescription opioid use may further reduce the number of CT scans obtained during ER visits.