Background

  • Computed tomography of the abdomen and pelvis (CT AP) is the most common diagnostic imaging test to evaluate abdominal pain in the ED. Intravenous (IV) and oral contrast improve diagnostic accuracy of CT AP.
  • There are currently no randomized clinical trials comparing contrast-enhanced versus unenhanced CT AP given ethical concerns. 
  • The most common reasons for withholding contrast from a patient include hypersensitivity reactions to contrast medium, kidney disease, or contrast shortages.

Clinical Question

  • What is the diagnostic accuracy of using CT abdomen pelvis (CT AP) without contrast compared to contrast-enhanced CT AP in ED patients with acute abdominal pain?

Design

Design

  • Multi-center, retrospective study of 201 consecutive ED patients spanning from April 1, 2017, to April 22, 2017.
  • Three quaternary, residency-affiliated academic centers. One institution was the location where all imaging took place, and radiologists were from all three institutions.  

Intervention

  • Patients underwent dual-energy contrast-enhanced CT in the portal venous phase at a single institution (Institution A). Two retrospective reviewers and the prospective radiologist involved in the clinical care interpreted these studies. This was the reference standard. 
  • Virtual unenhanced CT data was derived from the contrast-enhanced data. This virtual data did not have IV or oral contrast media. Six blinded radiologists interpreted the unenhanced CT scans (3 were senior radiology residents and three subspecialist abdominal radiologists).

Population

  • 201 patients
    • 108 female, 93 male 
  • Mean age = 50.1 years old
  • Mean BMI = 25.5
  • Inclusion criteria: 
    • Adults > 18 years.
    • Received contrast-enhanced dual-energy CT at institution A.
    • Imaging was done to evaluate for acute abdominal pain.
    • Virtual unenhanced CT (without oral or IV contrast) derived from the contrast-enhanced data.

Excluded

  • Patients who were missing imaging or clinical data.
    • One patient was excluded due to the prospective clinical CT report not being available.

Primary Results

  • Unenhanced CT was 30% less accurate for all primary and secondary diagnoses. 
  • BMI was not predictive of diagnostic error. There was an increase in diagnostic errors in older patients.

Strengths

  • Concurrent contrast-enhanced and unenhanced CT scans deter radiologist bias that would affect the study interpretations if patients were imaged at two separate times and control for changes in clinical situations that may occur in between imaging.
  • Radiologists reading the unenhanced studies were blind to the reference standard.
  • The patient population was vast and had a large range of diagnoses.

Limitations

  • CT imaging occurred amongst patients from one center (Institution A), potentially affecting the study population’s generalizability. 
  • Fellowship-trained abdominal radiologists and senior diagnostic radiology residents read the unenhanced studies. 
    • This may affect the generalizability in institutions with more junior residents and non-subspecialty radiologists available for reading. Additionally, there were no such criteria for radiologists to review the reference standard.
    • Only six radiologists were involved in reading the study.
  • 49/201 patients had a suggested diagnosis stated on the CT order and 98/201 patients said a specific location of abdominal pain, which may have biased the reviewers.
  • There may be limitations to external validity given the patient population steps from a single institution. Despite the population having a large range of diagnoses, the median BMI is 25 and age 50.1, which may not be readily generalizable to other populations.

Author's Conclusions

  • Contrast-unenhanced CT was 30% less accurate than contrast-enhanced CT.
  • Accuracy should be balanced with risks for AKI and hypersensitivity reactions. 
    • CIN rates are nearly 0% with GFR >30 mL/min. 
    • Mild prior hypersensitivity reactions have a <1% risk of life-threatening reaction with repeat contrast administration.

Our Conclusions

  • 70% accuracy is higher than expected for the unenhanced studies, but remember that the radiologists interpreting the studies were also subspecialty trained.
  • The decision to use contrast for a CT AP assessing for acute abdominal pain should be balanced with the risks CIN for patients with renal disease, and the risks of life-threatening hypersensitivity reactions for patients with allergies.
  • Non contrast-enhanced CT AP is a reasonable test to pursue in patients requiring abdominal imaging who have a strong contra-indication to contrast-enhanced CT AP; however, limitations in accuracy of non contrast-enhanced CT AP compared to CT AP should be understood and utilized in clinical decision making.

Read More

  • Shaish H, Ream J, Huang C, et al. Diagnostic Accuracy of Unenhanced Computed Tomography for Evaluation of Acute Abdominal Pain in the Emergency Department. JAMA Surg. 2023;158(7):e231112. doi:10.1001/jamasurg.2023.1112