Negative predictive value of intravenous contrast-enhanced CT of the abdomen for patients presenting to the emergency department with undifferentiated upper abdominal pain

• Published in: Emergency radiology Vol. 19; no. 1; pp. 19 - 26
• Main Authors: Ham, Hyungjoo, McInnes, Matthew D. F., Woo, Michael, Lemonde, Sylvie
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer-Verlag 2012; Springer Nature B.V
• Subjects: Abdominal Pain - diagnostic imaging; Chi-Square Distribution; Contrast Media; Diagnosis, Differential; Emergency Medicine; Emergency Service, Hospital; Female; Humans; Imaging; Male; Medicine; Medicine & Public Health; Middle Aged; Original Article; Predictive Value of Tests; Radiology; Retrospective Studies; Statistics, Nonparametric; Surveys and Questionnaires; Tomography, X-Ray Computed - methods; Abdominal pain; Emergency; Computed tomography; Negative predictive value
• ISSN: 1070-3004; 1438-1435
• DOI: 10.1007/s10140-011-0996-x

The purpose of this study is to calculate the negative predictive value (NPV) CT of the abdomen in patients presenting to the emergency department (ED) with undifferentiated upper abdominal pain. Approved by the hospital research ethics board, this retrospective study examined consecutive patients presenting to the ED with undifferentiated upper abdominal pain whose intravenous contrast-enhanced CT of the abdomen was reported as “normal” from June 2006–August 2010. Exclusion criteria included active malignancy, trauma, and known inflammatory bowel disease. True-negative (TN) vs. false-negative (FN) cases were categorized by consensus opinion of radiologist and emergency physician using a composite reference standard including clinical, laboratory, imaging, surgery, pathology, and patient self-reporting via phone questionnaire. The NPV was calculated with confidence intervals of 95%. The TN and FN groups were compared based on gender, age, site of pain, oral contrast use, and laboratory values. One hundred twenty-seven patients were included for analysis. The NPV was 64% (95% CI 55–72). The FN group had a higher proportion of patients with epigastric pain ( p  = 0.02) and a lower proportion of patients with left upper quadrant pain ( p  = 0.02). The WBC, lipase, and ALT were all higher in the FN group compared with the TN group. The most commonly missed pathologies were inflammatory conditions of the biliary tract and upper gastrointestinal systems. The NPV of CT for evaluation of undifferentiated upper abdominal pain in the ED was low at 64%. Physicians should consider this limitation and the commonly missed pathology when discharging patients with a “normal” CT report.