A growing body of research flies in the face of this common radiology practice
The routine use of contrast (both oral and IV, and certainly rectal) is unnecessary for the majority of abdominal CT scans performed in the ED. At least that is what the literature says over and over.
Unfortunately, many radiologists disagree. Is their objection based on a sound analysis of the literature? Hardly. In most cases it is a matter of personal preference. They have been using contrast since their residency, or at least since CTs came on the scene, and just feel more comfortable with it. Have they made an honest effort to compare results with and without contrast ? Probably not. Do they care that oral contrast will add about two hours to an ED stay and, even when given, frequently doesn’t get to the cecum? Probably not.
But when the use of contrast is subject to the intense searchlight of scientific inquiry, the answer seems to be pretty clear. It is the atypical patient with nontraumatic abdominal pain who needs contrast.
Before we take on the “contrast, no contrast” arm wrestle, we need to take a major step back and ask, “Why are we doing so many abdominal CTs in the first place?” I won’t spend this column making the case for why clinical exam can make many CTs unnecessary. That could be an entire column. What I want to talk about is the use of ultrasonography as a first test to visualize the abdominal contents and then, and only then, considering CT if results are equivocal and imaging is still felt to be necessary. This would be a dramatic departure from our current testing culture, where CTs are the new CBCs (don’t call the surgeon without one!).
Even the American College of Radiology has taken the radical position that, at least in children with suspected appendicitis, an ultrasound should be considered as an option. Why not in all ages and why not use some stronger language than “should be considered”? We’ll need to be satisfied with baby steps since it would likely be the uncommon radiologist in the U.S. who would actively promote following this advice. But here’s exactly what the ACR says. (As an aside, this is one of the ACR’s five “Choosing Wisely” recommendations.)
Don’t do computed tomography (CT) for the evaluation of suspected appendicitis in children until after ultrasound has been considered as an option.
Although CT is accurate in the evaluation of suspected appendicitis in the pediatric population, ultrasound is nearly as good in experienced hands. Since ultrasound will reduce radiation exposure, ultrasound is the preferred initial consideration for imaging examination in children. If the results of the ultrasound exam are equivocal, it may be followed by CT. This approach is cost-effective, reduces potential radiation risks and has excellent accuracy, with reported sensitivity and specificity of 94 percent.
Seems American radiologists don’t have nearly the experience (read: skill + confidence) of their European brethren in embracing this “ultrasound first” approach, but a two week course in Cancun would likely go a long way towards solving the problem. EPs can help them get this experience by ordering the correct test in these cases – an ultrasound and not a CT. EPs need to stop aiding and abetting this process. And we need to cite the support of the ACR so we don’t look like a bunch of no-nothing radicals.
But we’ll stick with the issue at hand – contrast in abdominal CTs. So, how about some papers indicating that oral contrast is a waste of time in the setting of suspected appendicitis – the most common setting in which oral contrast is used.
Instead of tediously going through study after study on this topic, here are two analyses, one published in 2010 involving seven high-quality trials and another in 2005 involving 23 trials. And yes, if you’re wondering, the answer to this dilemma was clearly known as far back as 2005! And catch the concluding sentence in the 2005 paper: the diagnostic accuracy of CT without contrast “was at least as comparable” as with contrast in assessing appendicitis. In fact, the data shows it was a little better.
DIAGNOSTIC ACCURACY OF NONCONTRAST COMPUTED TOMOGRAPHY FOR APPENDICITIS IN ADULTS: A SYSTEMATIC REVIEW
Hlibczuk, V., et al, Ann Emerg Med 55(1):51, January 2010
BACKGROUND: Non-contrast abdominal computed tomography (CT) is often used in the evaluation of suspected appendicitis, but its diagnostic accuracy has been questioned.
METHODS: The authors, coordinated at Columbia University in New York, performed a systematic literature review, examining all relevant electronic databases and the bibliographies of pertinent studies and previous systematic reviews, to evaluate the accuracy of non-contrast helical CT in adult emergency department (ED) patients with suspected acute appendicitis. There were specific inclusion and exclusion criteria, and two authors independently screened and chose articles, extracted data, and evaluated study quality. Patients must have been scanned in a multi-slice helical scanner. The diagnostic reference standard was final diagnosis at surgery or attempted clinical follow-up at a minimum of two weeks.
RESULTS: Seven high-quality studies, from an initial group of 1,258 publications, met inclusion criteria, yielding a sample of 1,060 patients. Pooled estimates for sensitivity and specificity were 92.7% (95% CI 89.5%-95.0%) and 96.1% (94.2%-97.5%), respectively. The positive likelihood ratio was 24 and the negative likelihood ratio was 0.08. The 7.3% false-negative rate in this study was in the same range as that found for CT scanning using contrast (3%-17%).
Copyright 2010 by Emergency Medical Abstracts - All Rights Reserved 7/10 - #18
A SYSTEMATIC REVIEW OF WHETHER ORAL CONTRAST IS NECESSARY FOR THE COMPUTED TOMOGRAPHY DIAGNOSIS OF APPENDICITIS IN ADULTS
Anderson, B.A., et al, Am J Surg 190(3):474, September 2005
BACKGROUND: Although CT scanning facilitates the identification of appendicitis while reducing the risk of negative laparotomy, low rates of CT imaging have been reported. Administration of oral contrast, which can be problematic and prolongs the diagnostic process, might be at least partially responsible for reluctance to request CT scanning for these patients.
METHODS: The authors, from the University of Washington in Seattle, reviewed the results of 23 studies (19 prospective, 4 retrospective) involving 3,474 patients over the age of 16 having abdominal CT scanning for suspected appendicitis.
RESULTS: The overall sensitivity, specificity and diagnostic accuracy were 97%, 97% and 97%, respectively, for five studies in which scanning was performed with rectal contrast, and 83%, 95% and 92%, respectively, for two studies in which it was performed with oral contrast. Sensitivity, specificity and diagnostic accuracy were 95%, 96% and 96%, respectively, in two studies of scanning with both rectal and oral contrast, and 93%, 93% and 92%, respectively, in seven studies of scanning with oral plus IV contrast. In eight studies in which no contrast was utilized, overall sensitivity and specificity were 93% and 98%, respectively, and diagnostic accuracy was 96%. When all studies in which oral contrast was used (alone or with rectal or IV contrast) are compared with studies in which no oral contrast was used, corresponding sensitivities were 92% vs. 95%, specificities were 94% vs. 97%, and diagnostic accuracy was 92% vs. 96%.
Copyright 2006 by Emergency Medical Abstracts - All Rights Reserved 1/06 - #19
So what is the real problem? It appears it is the radiologists. Here are two papers indicating the substantial variability in the interpretations of the same studies by different radiologists.
PROSPECTIVE COMPARISON OF HELICAL CT OF THE ABDOMEN AND PELVIS WITHOUT AND WITH ORAL CONTRAST IN ASSESSING ACUTE ABDOMINAL PAIN IN ADULT EMERGENCY
Lee, S.Y., et al, Emerg Radiol 12(4):150, May 2006
BACKGROUND: The need for oral contrast enhancement with newer abdominal CT technologies is uncertain.
METHODS: In this study, from Baystate Medical Center in Springfield, MA, 100 adults with acute nontraumatic abdominal pain underwent abdominopelvic helical CT scanning without and then with administration of oral contrast. The scans were interpreted independently by different radiologists, and a significant difference in interpretation was defined as one that resulted in a change in patient management.
RESULTS: The most common indications for scanning were clinically suspected appendicitis or diverticulitis. There was significant disagreement in interpretation of the contrast-enhanced and unenhanced CT scans for 21 patients, representing an agreement level of 79%. Eleven of these 21 patients had a negative unenhanced CT and an abnormality on the contrast-enhanced scan. The opposite was true for six patients, and in the remaining four patients both scans were felt to be abnormal but there were significant differences in the reported abnormalities. On unblinded review of the discordant CTs by two radiologists, two pairs were felt to be discordant on the basis of the oral contrast protocol, and the remaining discrepancies were felt to be primarily related to interobserver variability.
Copyright 2006 by Emergency Medical Abstracts - All Rights Reserved 8/06 - #1
The following study demonstrates it doesn’t matter what contrast, or combination thereof, is used, but that the major variable is the radiologist.
MDCT FOR SUSPECTED ACUTE APPENDICITIS IN ADULTS: IMPACT OF ORAL AND IV CONTRAST MEDIA AT STANDARD-DOSE AND SIMULATED LOW-DOSE TECHNIQUES
Keyzer, C., et al, Am J Roent 193:1272, November 2009
BACKGROUND: CT protocols for the evaluation of patients with suspected appendicitis vary from one institution to another, but often involve use of both IV and oral contrast.
METHODS: In this Belgian study, CT scanning (4-MDCT) with oral, IV or both oral and IV contrast was performed in 131 adults aged 18-87 (mean, 37) being evaluated for suspected acute appendicitis. Scans were interpreted independently by two radiologists using standard-dose and simulated low-dose settings.
RESULTS: Based on surgical pathology or clinical follow-up and other diagnostic procedures, a final diagnosis of appendicitis was made in 25% of the patients. In the group not receiving oral contrast, the sensitivity achieved by the two readers in interpreting scans with or without IV contrast using standard-dose or simulated low-dose CT protocols ranged between 70% and 85%; specificity ranged between 91-100% and diagnostic accuracy ranged between 88-94%. In the group receiving oral contrast, sensitivity ranged between 85-100%, specificity ranged between 88-100% and accuracy ranged between 91-100%. The accuracy of CT interpretation was reader-dependent but was not dependent upon contrast type or the CT dose protocol. In general, both readers provided a correct diagnosis of acute appendicitis or an alternative disease for 61% of the patients, regardless of the radiation dose or the use of IV, oral, or IV plus oral contrast.
CONCLUSIONS: In these adults being evaluated for suspected acute appendicitis, the accuracy of CT interpretation was dependent upon the reader more than on the contrast protocol or CT dosing protocol that was used. The authors acknowledge the need for validation of their findings. 35 references (
- no reprints)
Copyright 2010 by Emergency Medical Abstracts – All Rights Reserved 3/10 - #17
And how about if CTs were done without contrast and in those cases where the results were equivocal the studies were repeated with contrast (if the use of contrast makes interpretation easier then this study should help prove it). Didn’t seem to be the case – there was no change in overall accuracy.
ACUTE APPENDICITIS: DIAGNOSTIC VALUE OF NONENHANCED CT WITH SELECTIVE USE OF CONTRAST IN
ROUTINE CLINICAL SETTINGS
Tamburrini, S., et al, Eur Radiol 17(8):2055, August 2007
BACKGROUND: There is disagreement regarding the need for contrast enhancement in patients undergoing CT scanning for possible appendicitis.
METHODS: This study, from Naples, Italy, and UC San Diego, reviewed CT readings and clinical charts in 536 patients with suspected appendicitis who received CT scanning according to a protocol which called for initial unenhanced scanning followed by repeat scanning with contrast at the discretion of the radiologist if the initial scan was felt to be inconclusive. The accuracy of CT interpretation was based on findings at laparotomy or on clinical follow-up in patients with a reportedly negative scan.
RESULTS: The initial unenhanced CT was judged to be conclusive in 75% of the patients. Repeat scanning with contrast was performed in all but six of the remaining 132 patients, with contrast choices including IV administration in 118 cases, and oral and rectal administration in 33 and 12 cases respectively (37 patients received contrast by more than one route). The sensitivity and specificity of the initial scans were 90% and 96%, respectively, and for the entire group (including all those who had a second, contrast-enhanced, CT) was 91% and 95%, respectively.
CONCLUSIONS: CT scanning without contrast was reasonably accurate in the three-quarters of cases in which it was felt be “conclusive.” Selective repeat scanning with contrast enhancement maintained essentially the same overall accuracy, even including those patients whose initial scan was inconclusive, but would lead to the costs and radiation exposure incurred by a second scan. 32 references (
Copyright 2008 by Emergency Medical Abstracts - All Rights Reserved 1/08 - #22
Finally, we know what a pain in the butt the use of oral contrast can be in the ED setting. Although this isn’t the only paper on the subject, it nicely demonstrates a variety of the problems.
THE LACK OF EFFICACY FOR ORAL CONTRAST IN THE DIAGNOSIS OF
APPENDICITIS BY COMPUTED
Laituri, C.A., et al, J Surg Res 170(1):100, September 2011
BACKGROUND: Some protocols for CT scanning for possible appendicitis call for administration of oral contrast. This practice can be difficult for patients and prolongs the time to performance of the scan. The necessity for oral contrast has been questioned, particularly with the use of helical and multidetector scanning.
METHODS: This study, from Children’s Mercy Hospital in Kansas City, MO, retrospectively reviewed findings in 1,561 patients (mean age, 10 years) undergoing CT scanning with IV and oral contrast for possible appendicitis.
RESULTS: Appendicitis was diagnosed in 41.8% of the patients. An average of two doses of oral contrast were administered, for a mean total volume of 616.9ml. Contrast was observed in the terminal ileum in only 72.4% of the patients (67.0% of those with appendicitis and 76.2% of patients without appendicitis). About 19% of the patients with appendicitis and 13% of those without appendicitis vomited the contrast material, and nasogastric tubes were placed in 5.8% and 5.1% of the patients with and without appendicitis, respectively. There was no difference between the groups with and without contrast in the terminal ileum with regard to the percentage of patients with operative or pathologic confirmation of appendicitis, in CT concordance with operative findings, or in the percentage of patients with equivocal CT readings (7.5% in the group with contrast in the terminal ileum vs. 5.3% in the comparison group).
CONCLUSIONS: Oral contrast administered for CT scanning for possible appendicitis is associated with substantial disadvantages. In this series, it often failed to reach the terminal ileum and did not appear to affect diagnostic accuracy. 23 references (
Copyright 2012 by Emergency Medical Abstracts - All Rights Reserved 3/12 - #12
And for those who are fans of IV contrast in appendicitis, not so fast – hopefully the following study will rattle their confidence. It assessed the value of IV contrast in identifying a normal appendix in asymptomatic cancer patients comparing the use of IV or no contrast. The use of IV contrast did not significantly improve visualization of a normal appendix. Again, intra- and inter-reader differences were substantial.
NORMAL APPENDIX IN ADULTS:
REPRODUCIBILITY OF DETECTION WITH UNENHANCED AND CONTRAST-ENHANCED MDCT
Keyzer, C., et al, Am J Roent 191:507, August 2008
BACKGROUND: Visualization of a normal appendix on CT scanning is believed to reliably exclude appendicitis.
METHODS: This prospective Belgian study examined the effect of IV contrast enhancement and other variables on visualization of a normal appendix in 102 adult cancer patients aged 36-94 (mean 63, 74% male) without intraabdominal digestive tumors or suspected appendicitis who were referred for CT evaluation of the abdomen. Each patient underwent 64-multidetector CT scanning with and without IV contrast. CT studies were read independently by two experienced radiologists and one first-year radiology resident. Interpretation of unenhanced and contrast-enhanced CT studies was separated by a two-week period, and repeated after one month. The gold standard was interpretation by two independent experts.
RESULTS: The experts identified a normal appendix in 96% of the patients. For the three readers, rates of identification of a normal appendix with certainty ranged between 70- 91% with unenhanced CT studies, and between 77-92% with contrast- enhanced CT studies. Differences between unenhanced and contrast- enhanced CT studies were not statistically significant. There was perfect intra- and interreader agreement for 71% of the patients, and agreement in categorizing confidence in identification of the appendix ranged between fair and good. The level of agreement was influenced by the patients’ body mass index and intraabdominal fat volume.
CONCLUSIONS: The level of agreement between radiologists in the visualization of a normal appendix on CT scanning was fair to good. IV contrast enhancement did not significantly improve the visualization of a normal appendix or confidence in CT interpretation. 29 references
Copyright 2008 by Emergency Medical Abstracts - All Rights Reserved 12/08 - #21
This cluster of abstracts is by no means the totality of evidence indicating there is no advantage to the routine use of contrast in the patient with suspected appendicitis. The literature regarding this topic is extensive and is largely beyond the scope of this column. Hopefully the few papers that are presented make it clear that by no means is the use of contrast, especially oral contrast, evidence-based in suspected appendicitis. The same is true of renal or ureteral calculi, so looking for this diagnosis won’t serve to justify the use of oral or IV contrast either.
As a reflection of just how clear the totality of the evidence is, below is the official policy of the Department of Radiology of the Los Angeles County / Keck School of Medicine (USC) as of February 26, 2010:
“All patients from the DEM at LACUSC who have a CT scan of the Abdomen and pelvis requested will be scanned without oral contrast. The only exception being patients in who a fistula/ leak from bowel is the primary concern or is to be excluded. Please contact the radiology attending/ resident on duty if there any questions” [SIC]
As can be seen, the policy is not just for the senior radiologists who specialize in reading abdominal CTs, but applies to all the radiologists, even the lowly radiology residents.
Hopefully this column will serve to further the dialogue between EPs and their radiology colleagues, and begin to change behavior towards more evidence-based practices.