VACEP's EBM Series: Impact of IV contrast on the diagnostic accuracy of abdominal CT
VACEP Evidence-Based Medicine for General Emergency Physicians Series
Authors: Alex Schell, DO & Martin D. Klinkhammer MD, MPH | Eastern Virginia Medical School
Reviewer: Joshua Easter, MD, MSc | University of Virginia School of Medicine, UVA Faculty
The VACEP Evidence-Based Medicine review series highlights a recent peer-reviewed clinical study. You can also read the full article, Diagnostic Accuracy of Unenhanced Computed Tomography for Evaluation of Acute Abdominal Pain in the Emergency Department.
CASE
A 56-year-old woman with a history of hypertension, chronic kidney disease, Crohn’s, and a surgical history of cesarean section comes into your emergency department for evaluation of abdominal pain. She reports a worsening sharp pain in the left lower quadrant of her abdomen for the past five days.
She has had a fever at home as well as some nausea without vomiting and multiple episodes of non-bloody diarrhea. Vital signs on arrival are BP, 130/90; HR, 110; RR, 20; T, 100.2℉; and O2 saturation, 98% on room air.
Abdominal exam shows tenderness to deep palpation in the suprapubic area and left lower quadrant without rebound or guarding. Labs are ordered in addition to an abdominal computerized tomography (CT) scan with IV contrast. The CT technologist calls you, as the patient’s GFR is 25 ml/mn/1.73 m2. Hospital protocol calls for a CT without IV contrast.
Should you change your order and scan without the contrast?
STUDY SUMMARY
It is unclear how diagnostic accuracy is affected for most conditions when obtaining a CT without contrast.
With the debate over contrast induced nephropathy (McDonald, 2013; Davenport, 2020), increased lengths of stay to pretreat contrast allergies, and supply line shortages in contrast material, the push for forgoing contrast when obtaining a CT has increased.
Most prior studies have focused on single disease entities, such as diverticulitis or appendicitis, and the effect of excluding IV contrast (Tack, 2005; K H In't Hof, 2004). Others have attempted to compare contrast-enhanced to non-contrast abdominal CT but lacked a clear reference standard (Hill, 2010; Lamoureux, 2019).
This recent study sought to determine the diagnostic accuracy for both the principal cause of ED patients’ acute abdominal pain (primary finding) as well as any actionable incidental findings between contrast enhanced vs. non-contrast studies (Shaish, 2023).
About the EBM Review Series
This is a literature review series started by the University of Virginia’s Josh Easter, MD, MSc, a VACEP board member working to connect the academic community in Virginia. We invite each residency in Virginia (and D.C.) to create a faculty/resident team to submit and review articles. Sign up to submit one.
Goals
Provide a brief monthly synopsis of a high yield article germane to the practice of emergency medicine for distribution to all VACEP members
Provide an opportunity for a peer reviewed publication and invited presentation for faculty and trainees
Foster an academic community focused on evidenced based medicine for emergency medicine residency programs in the region
This study was a multicenter retrospective diagnostic accuracy study using CT data from a single emergency department in April 2017.
It included 202 consecutive patients over the age of 18 who underwent oral and intravenous contrast enhanced CT imaging for evaluation of acute abdominal pain at a single ED site.
The CT performed was then virtually unenhanced to remove the IV and oral contrast to acquire a new set of unenhanced images that was then reviewed by 3 physicians from a pool.
They were blinded to the results of the CT with contrast. Primary and actionable incidental findings were agreed upon as the reference standard, if 2 out of 3 reviewers agreed on the findings in the CT with contrast images.
This became the reference standard for which to compare the CT without contrast images for both primary and actionable incidental findings. The images without contrast were then reviewed in the same way by another 3 attending and 3 resident radiologists.
The images without contrast interpreted by radiologists were 30% less accurate than the contrast enhanced images. The decreased accuracy arose from false positive and false negative findings. The false negative frequencies were approximately 20% and the false positive rates around 15%. Missed primary diagnoses (false negative readings) on unenhanced CTs included appendicitis, splenic infarct, vascular thrombus or dissection, disseminated malignancy, hemoperitoneum, and urinary tract obstruction (Shaish, 2023).
False positive readings included pancreatitis, bowel perforation, diverticulitis, pyelonephritis, and neoplasm. On further analysis the findings were found to be worse for elderly patients and not affected by BMI.
STRENGTHS
Overall, the study had many strengths.
The study had images reviewed by multiple blinded radiologists at different levels of training, with overall moderate interrater agreement (Gwet AC, 0.58).
The study created a good reference group (diagnosis based on contrast study), and had a unique way to obtain unenhanced images without placing patients at risk of receiving a worse study or forcing them to undergo radiation from the CT twice.
The results showed a significant decrease in diagnostic accuracy for CT without contrast.
LIMITATIONS
The study had some significant limitations.
It was a retrospective review, and therefore multiple unmeasured confounders may impact the results.
Additionally, the study used the contrast-enhanced study as the reference standard, and this may have differed from the final clinical diagnosis.
Furthermore, it is difficult to determine if the process employed to remove the contrast from the images led to any degradation of the images that would not have existed in a de novo non-contrast scan.
Lastly, the study used both oral and IV contrast as the reference standard for all patients. In most emergency departments in which we practice, the typical patient only receives IV contrast for undifferentiated abdominal pain. While most studies would suggest that oral contrast only adds minimally to the diagnostic yield of abdominal CT for undifferentiated abdominal pain (Triche, 2022; Farrell, 2018; Wadhwani, 2016 ), it is still possible that the addition of oral contrast may have improved the accuracy of the original CTs. It is also unclear about possible selection bias given patients included all received oral contrast for an unclear reason.
These results suggest that we need to have risk versus benefit discussions with our patients in regards to potential harm from IV contrast compared to the decrease in diagnostic accuracy. Those patients with GFRs >30 ml/min/1.73 m2 have almost no risk of contrast induced acute kidney (Davenport, 2020). The risk is more uncertain for those with GFRs below 30 ml/min/1.73 m2, ranging from 0 to 17%. It is also difficult to estimate the exact risk a contrast bolus confers to those patients with a history of a previous moderate to severe hypersensitivity reaction.
Conclusion
Returning to our patient with a GFR of 25 ml/min/1.73 m2 and a concerning history and abdominal exam that could arise from multiple emergent etiologies, after a discussion with the patient about the risks and benefits of contrast, you decide together to scan with IV contrast.
The CT shows the development of a small abscess. The patient does not develop any significant nephropathy.
