Abdominal 64-MDCT for Suspected Appendicitis: The Use of Oral and IV Contrast Material Versus IV Contrast Material Only

Transcription

1 Gastrointestinal Imaging Original Research Anderson et al. Use of Contrast Material for MDCT of Suspected Appendicitis Gastrointestinal Imaging Original Research Stephan W. Anderson 1 Jorge A. Soto 1 Brian C. Lucey 1 Al Ozonoff 1,2 Jacqueline D. Jordan 1 Jirair Ratevosian 1 Andrew S. Ulrich 3 Niels K. Rathlev 3 Patricia M. Mitchell 3 Casey Rebholz 3 James A. Feldman 3 James T. Rhea 1 Anderson SW, Soto JA, Lucey BC, et al. Keywords: appendicitis, contrast media, CT, diverticulitis, emergency medicine, MDCT DOI: /AJR Received January 16, 2009; accepted after revision May 11, Please see the commentary on this article and the article by Keyzer et al., which appears on the preceding pages. 1 Department of Radiology, Boston University Medical Center, 88 E Newton St., 2nd Fl., Boston, MA Address correspondence to S. W. Anderson (stephan.anderson@bmc.org). 2 Department of Biostatistics, Boston University, Boston, MA. 3 Department of Emergency Medicine, Boston University Medical Center, Boston, MA. AJR 2009; 193: X/09/ American Roentgen Ray Society Abdominal 64-MDCT for Suspected Appendicitis: The Use of Oral and IV Contrast Material Versus IV Contrast Material Only OBJECTIVE. The objective of our study was to compare the diagnostic accuracy of IV contrast-enhanced 64-MDCT with and without the use of oral contrast material in diagnosing appendicitis in patients with abdominal pain. MATERIALS AND METHODS. We conducted a randomized trial of a convenience sample of adult patients presenting to an urban academic emergency department with acute nontraumatic abdominal pain and clinical suspicion of appendicitis, diverticulitis, or smallbowel obstruction. Patients were enrolled between 8 am and 11 pm when research assistants were present. Consenting subjects were randomized into one of two groups: Group 1 subjects underwent 64-MDCT performed with oral and IV contrast media and group 2 subjects underwent 64-MDCT performed solely with IV contrast material. Three expert radiologists independently reviewed the CT examinations, evaluating for the presence of appendicitis. Each radiologist interpreted 202 examinations, ensuring that each examination was interpreted by two radiologists. Individual reader performance and a combined interpretation performance of the two readers assigned to each case were calculated. In cases of disagreement, the third reader was asked to deliver a tiebreaker interpretation to be used to calculate the combined reader performance. Final outcome was based on operative, clinical, and follow-up data. We compared radiologic diagnoses with clinical outcomes to calculate the diagnostic accuracy of CT in both groups. RESULTS. Of the 303 patients enrolled, 151 patients (50%) were randomized to group 1 and the remaining 152 (50%) were randomized to group 2. The combined reader performance for the diagnosis of appendicitis in group 1 was a sensitivity of 100% (95% CI, %) and specificity of 97.1% (95% CI, %). The performance in group 2 was a sensitivity of 100% ( %) and specificity of 97.1% ( %). CONCLUSION. Patients presenting with nontraumatic abdominal pain imaged using 64- MDCT with isotropic reformations had similar characteristics for the diagnosis of appendicitis when IV contrast material alone was used and when oral and IV contrast media were used. C T protocols for evaluating patients presenting with acute nontraumatic abdominal pain, including appendicitis, commonly include the use of oral and IV contrast agents [1 4]. Recently, several studies have suggested that oral contrast material may not be necessary in these patients [5, 6]. The clinical implications of examining a patient with nontraumatic abdominal pain without oral contrast material are multiple and are important. A published comparison of two groups of patients presenting emergently with abdominal pain one of whom received oral contrast material and the other who did not revealed a significant reduction in the time to eventual disposition after admission to the emergency department (ED) [7]. For suspected appendicitis, a method of reducing the bowel preparation time using rectal administration of barium also yielded predictable decreases in the duration of patient stay in the ED [8]. Other clinical implications of oral contrast material use include, first, the risk of aspiration; second, the need for nasogastric tubes in patients with severe emesis or who are unable to tolerate orally administered contrast material; and, third, the increased examination cost and potential adverse effects of the oral contrast material. The 64-MDCT technology affords a considerable reduction in scanning time and an improvement in image quality, derived from the isotropic voxel data sets and 1282 AJR:193, November 2009

2 Use of Contrast Material for MDCT of Suspected Appendicitis very short scanning times, that result in minimal degradation from motion artifact. Multiplanar reformations particularly benefit from isotropic data sets. Recent work in patients presenting with an acute abdomen has shown 64-MDCT to be of benefit by increasing reader confidence and interobserver agreement in diagnosis [9]. Similar benefits have previously been described using MDCT technology in patients with abdominal pain [10, 11]. In light of the enhanced image quality afforded by 64-MDCT, we sought to critically evaluate the need for oral contrast material given the known improvement in ED duration of stay without its use. Therefore, the purpose of our study was to conduct a randomized trial to compare the diagnostic performance of IV contrast-enhanced 64-MDCT performed with oral contrast material versus without oral contrast material in patients presenting to the ED with acute nontraumatic abdominal pain clinically suspected to be secondary to appendicitis, diverticulitis, or small-bowel obstruction. For the purposes of this article, the effects of these two CT protocols on the diagnosis of appendicitis are evaluated. Materials and Methods Patient Population This randomized clinical trial was approved by our institutional review board and was conducted in a HIPAA-compliant fashion. We enrolled a convenience sample of adult subjects who presented to the ED with acute abdominal pain. Included were adults 21 years old or older with nontraumatic abdominal pain and differential considerations of abdominal pain including appendicitis, diverticulitis, and small-bowel obstruction. The age of 21 years old or older was chosen because patient recruitment occurred in the adults ED; patients younger than 21 years are seen in the pediatric ED in our institution. We excluded patients with an IV contrast allergy, pregnant patients, patients with a traumatic cause of abdominal pain, patients with positive β-hcg findings, patients unable to provide a telephone number for a 24-hour followup call, and patients unable to provide written informed consent. A convenience sample of patients was recruited for this study between April 2006 and February 2007; the duration of the study was based on the time required to recruit the number of patients determined by power calculations, which we discuss later. Research assistants who had been trained by ED physicians in screening for and recruiting potential patients screened all ED patients Monday through Friday from 8 am to 11 pm. A convenience sample is a sample of patients recruited at the convenience of the researcher, given the defined hours of screening in the ED by the research assistants in this protocol. Three hundred three patients (140 men, 163 women; mean age of men, 42 years; mean age of women, 44 years) were enrolled in this study. After providing written informed consent, subjects were randomized into one of two groups: abdominal CT performed with both oral and IV contrast material (group 1) or abdominal CT performed solely with IV contrast material (group 2). Group 1 included 75 men and 76 women with a mean age of 43 years (range, years), and group 2 included 65 men and 87 women with a mean age of 43 years (range, years). After discharge from the ED, all patients were followed up 24 hours and 6 weeks later via a telephone questionnaire administered by a research assistant. Additionally, electronic medical records at our institution were reviewed 72 hours after patient discharge to ensure that patients had not returned to the ED with continued symptoms. If patients could not be contacted by telephone at the 6-week interval, electronic medical records were reviewed and evaluated for any additional ED or outpatient visits with ongoing abdominal symptoms. Adult patient presents with clinical suspicion of acute appendicitis to emergency department Group 1: Oral and IV contrast-enhanced CT Agreement Study recruitment Randomization Images interpreted by two of three emergency radiologists Comparison to reference standard CT Technique CT examinations were performed using a 64- MDCT scanner (LightSpeed VCT, GE Healthcare). Portal venous phase CT images were acquired from the diaphragm to the greater trochanters with the following parameters: slice thickness, 1.25 mm; reconstruction interval, 1.25 mm; pitch, 1:0.987; noise factor, 19 (automatic attenuation-based dose modulation); and rotation time, 0.5 second. The direct multiplanar reformation function was used to generate coronal and sagittal reformations with a slice thickness of 2.5 mm and a reconstruction interval of 2.5 mm. All patients received 100 ml of ioversol (Optiray 350 mg I/mL, Mallinckrodt Imaging) administered via power injection through an IV cannula located in an antecubital or hand vein at a rate of 4 ml/s. We routinely use a dual-syringe power injector (Stellant CT Injection System, Medrad) to administer a 30-mL normal saline chasing bolus immediately after the injection of IV contrast material. The saline bolus was also injected at 4 ml/s. The acquisition of portal venous phase images started 70 seconds after the initiation of the injection of IV contrast material. Patients randomized to group 1 received our department s standard positive oral contrast preparation (900 ml of 2.2% barium sulfate suspension [Medescan barium sulfate, Lafayette Pharmaceuticals]). According to department protocol, a 2-hour preparatory time was used to ensure adequate bowel opacification. Image Analysis Three expert radiologists with 8 20 years of experience interpreting emergent abdominal CT scans independently interpreted the CT images at PACS workstations (Centricity, GE Healthcare). At the conclusion of subject enrollment, each radiologist was given an electronic sequential list of 202 patients (101 from group 1 and 101 from group 2) by an independent investigator, thus ensuring that each CT examination was interpreted by two radiologists. The scans were interpreted so that for the three possible pairings of the observers, there were 101 cases of overlap interpreted by two radiologists independently (Fig. 1). The radiologists were asked to assess for the presence of Group 2: IV contrastenhanced CT Adjudication by third emergency radiologist Fig. 1 Patient flowchart details time course from emergency department presentation to combined image interpretation paradigm involved in this study. AJR:193, November

3 Anderson et al. appendicitis. Additionally, they were instructed to determine their confidence in diagnosis using the following scale of 1 5 for each case: 1, definitely not present; 2, probably not present; 3, indeterminate; 4, probably present; or 5, definitely present. Appendicitis was considered present on CT images when the appendiceal diameter was enlarged with evidence of periappendiceal stranding (Figs. 2 and 3). Other signs consistent with a diagnosis of appendicitis included a fluid-filled lumen and hyperenhancement of the appendiceal mucosa as well as the absence of filling with oral contrast material when oral contrast material was seen within the cecum. The radiologists interpreted the studies using axial images and sagittal and coronal reformations. They completed a data collection form for each patient, documenting the presence or absence of appendicitis as well as their confidence level using the scale mentioned. Once the combined interpretations were compared with the standard of reference, two of the radiologists responsible for the combined interpretation retrospectively evaluated all false-negative and false-positive interpretations by consensus to provide a possible explanation for each misinterpretation. To assess the adequacy of our routine oral contrast preparation, a single radiologist evaluated the examinations of all patients who received the oral contrast preparation and noted the distal most point of oral contrast opacification. Fig. 2 True-positive interpretation of acute appendicitis in 47-year-old man. Axial IV contrastenhanced CT image obtained without oral contrast material reveals dilated appendix with significant periappendiceal fat stranding (arrow). The radiologist recorded the distal most point of opacification using the following categories: stomach, proximal small bowel, mid small bowel, distal small bowel, cecum, ascending colon, hepatic flexure, transverse colon, splenic flexure, descending colon, sigmoid colon, or rectum. Standard of Reference A review of the patients electronic medical records was undertaken by a single investigator who was blinded to CT results to determine clinical outcomes, including operative management. The determination of clinical outcomes also was based on follow-up information obtained 24 hours and 6 weeks after discharge from the ED using telephone questionnaires, which included any additional clinical information from ED visits or hospital admission in the interim. Also, clinical information from review of electronic medical record reviews 72 hours after ED discharge in all patients and 6 weeks after discharge in patients not contacted by telephone was factored into the A Fig. 3 True-positive interpretation of acute appendicitis in 26-year-old man. A, Sagittal IV contrast-enhanced CT image obtained with oral contrast material reveals contrast material opacification of proximal appendiceal lumen (arrow). B, Sagittal image of distal appendix shows lack of contrast opacification, distention, and periappendiceal inflammation (arrow), which are consistent with acute appendicitis. B 1284 AJR:193, November 2009

4 Use of Contrast Material for MDCT of Suspected Appendicitis clinical outcome determination. Appendicitis was considered present only when patients underwent operative intervention and pathologic analysis yielded a diagnosis of acute appendicitis. Statistical Analysis To determine the necessary sample size, we assumed 40% prevalence of a positive diagnosis of appendicitis, diverticulitis, or small-bowel obstruction using a one-sided comparison of proportions across the two arms with a significance level of The effect size of interest was a difference in either sensitivity or specificity of 90% versus 70%, respectively. We calculated an effective sample size of 303 subjects that would be required to provide 80% power. Once all interpretations were completed by the three radiologists, an independent investigator reviewed the data collection forms to determine interobserver agreement. Interpretations scored 1 and 2 were considered negative interpretations, and those scored 4 and 5 were considered positive interpretations. Interpretations scored 3 were considered false-positive or false-negative depending on whether the standard of reference showed the diagnosis to be absent or present, respectively. If both observers agreed on the presence or absence of the diagnosis, that combined interpretation was considered final and was compared with the standard of reference to calculate the following parameters of diagnostic performance: sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). In cases in which there was disagreement between the two readers (scored as negative by one reader and positive by the other), a blinded third reader was asked to interpret the examination. In these cases of disagreement, the third reader was the third expert radiologist who was not originally assigned to interpret that particular examination. For cases in which either one or both of the expert interpretations was indeterminate (scored 3), the third reader also rendered an interpretation; this third read acted as the tiebreaker to yield the final (i.e., combined) interpretation of the study (Fig. 1). Receiver operating characteristic (ROC) curves for each expert were created, and the areas under the ROC curves (AUCs) were calculated. These calculations and comparisons between the AUC values for group 1 versus group 2 were carried out for each radiologist. AUCs were calculated using a software package (ROCR package, version 1.0.2, 2007) for the R statistical system (version 2.0.1, R Foundation for Statistical Computing, Vienna University of Technology, Vienna, Austria) [12]. The ROC calculations used the aforementioned 5-point confidence scale. The interpretations of the individual observers were compared with the clinical outcomes to determine the sensitivity, specificity, PPV, and NPV for the diagnosis of appendicitis using the R statistical system. The performance of observers 1, 2, and 3 was calculated based on the 202 examinations interpreted by each. Using the Fisher s exact test, we compared the sensitivity and specificity of the combined interpretations of the two groups for the diagnosis of appendicitis. One-sided tests were used only for the purposes of sample size calculations. All other comparisons and reported p values are two-tailed, nondirectional tests. Kappa values for interobserver agreement were calculated for each reader pair (101 studies per pair) using the method of Fleiss kappa [13]. The levels of agreement were interpreted as follows: 0.20, poor agreement; , fair agreement; 0.41 A 0.60, moderate agreement; , good agreement; and , very good agreement. The kappa values for each pair of observers were calculated based on the 101 examinations of overlap that existed for each pair of the three observers. For kappa value calculations, interpretations of 1 and 2 were considered negative; 4 and 5, positive; and 3, equivalent only to another interpretation of 3. Results Based on the results of the standards of reference, 27 (9%) of the 303 patients were diagnosed with appendicitis. Of the 303 patients, 121 patients (40%) were successfully contacted 24 hours after discharge from the ED and 155 patients (51%) were successfully contacted at 6 weeks. Additionally, nine patients (3%) returned to our institution for care within the 6-week follow-up interval: Two patients were admitted, six patients were seen in the ED and discharged, and one patient returned for a follow-up visit. Appendicitis Fourteen (9%) of the 151 patients in group 1 and 13 (9%) of the 152 patients in group 2 had a final diagnosis of acute appendicitis based on operative intervention and histopathologic correlation. All patients who underwent operative intervention were found to have appendicitis. There were 14 true-positive, 133 true-negative, four false-positive, and 0 false-negative interpretations for the CT examinations based on the combined interpretations for group 1. There were 12 true-positive, 136 true-negative, four false-positive, and 0 false-negative interpretations for the CT examinations based B Fig. 4 False-positive interpretation for acute appendicitis in case of surgically resected appendiceal mucocele in 43-year-old man. A and B, Axial (A) and coronal (B) oral and IV contrastenhanced images show marked enlargement of appendix with mild periappendiceal fat stranding (arrows). This case represents false-positive interpretation for acute appendicitis. AJR:193, November

5 Anderson et al. on the combined interpretations for group 2. On retrospective consensus review of the four false-positive misinterpretations of appendicitis in group 1, the radiologists thought that three of the cases were equivocal in that the caliber of the appendix was slightly enlarged with questionable mild hyperenhancement of the appendiceal mucosa but there was a lack of periappendiceal inflammatory changes. The final false-positive in this group was a markedly abnormal appendix, which on pathology was shown to be an appendiceal mucocele, that had been misinterpreted as an abscess by the radiologists. In retrospect, the radiologists thought that the absence of significant stranding surrounding the mucocele would be atypical for an abscess (Fig. 4). On consensus review of the four false-positive misinterpretations of appendicitis in group 2, these cases were thought to be equivocal with apparent hyperenhancing wall and appendiceal caliber at the upper limits of normal but without surrounding appendiceal stranding. Statistical Analysis In constructing ROC curves for the three radiologists, AUCs were calculated. A comparison of the AUC values between the two groups for each radiologist is shown in Table 1. Routinely used Monte Carlo permutations were A Fig. 5 True-positive interpretation of acute appendicitis with lack of oral contrast opacification of colon in 39-year-old man. A and B, Axial (A) and coronal (B) oral and IV contrast-enhanced images show distended appendix (arrows), consistent with acute appendicitis, and oral contrast opacification terminating in proximal small bowel. used to calculate p values, and no statistically significant differences in the AUC values were found between the two groups [14, 15]. Using Fisher s exact test, comparisons of sensitivity and specificity between the combined interpretations for the diagnosis of appendicitis were carried out (Table 2). Although there was no significant difference in sensitivity or specificity for the diagnosis of appendicitis, our sample size limits the conclusion we can draw from this finding. In 11 cases, there was disagreement between the two interpreting radiologists and the third radiologist interpreted the case, serving as a tiebreaker. Of these 11 cases necessitating a third interpretation, 10 were in group 2. Interobserver agreement for the diagnoses of appendicitis was also calculated (Table 3). Interobserver agreement was variable for the three diagnoses and ranged from poor to very good. Position of Oral Contrast Material The position of the distal most extent of the orally administered contrast material was recorded in all patients in group 1. All patients ingested the entire amount of oral contrast material; however, vomiting during administration was neither controlled for nor recorded. The distal most position of the contrast material was as follows: the stomach, n = 6 (4%); proximal small bowel, n = 8 (5%); mid small bowel, n = 21 (14%); distal small bowel, n = 13 (9%); cecum, n = 8 (5%); ascending colon, n = 8 (5%); hepatic flexure, n = 5 (3%); transverse colon, n = 19 (13%); splenic flexure, n = 8 (5%); descending colon, n = 22 (15%); sigmoid colon, n = 11 (7%); and rectum, n = 22 (15%). Of TABLE 1: Receiver Operating Characteristic (ROC) Curve Analysis Reader Diagnosis AUC Group 1 a Group 2 b p Lower Limit Upper Limit AUC Lower Limit Upper Limit 1 Appendicitis Appendicitis Appendicitis Note AUC = area under ROC curve. a Group 1 subjects underwent 64-MDCT performed with oral and IV contrast material. b Group 2 subjects underwent 64-MDCT performed solely with IV contrast material. B 1286 AJR:193, November 2009

6 Use of Contrast Material for MDCT of Suspected Appendicitis TABLE 2: Reader Performance in Diagnosing Appendicitis Using 64-MDCT With Oral and IV Contrast Material or IV Contrast Material Alone Reader 1 Reader Sensitivity (%) Specificity (%) Positive Predictive Value (%) Negative Predictive Value (%) Group 1 a 100 [ %] 97.8 [ %] 80.0 [ %] 100 [ %] Group 2 b 88.9 [ %] 99.0 [ %] 88.9 [ %] 99.0 [ %] Reader 2 Group 1 a 100 [ %] 96.8 [ %] 80.0 [ %] 100 [ %] Group 2 b 90.0 [ %] 93.9 [ %] 60.0 [ %] 98.9 [ %] Reader 3 Group 1 a 100 [ %] 100 [ %] 100 [ %] 100 [ %] Group 2 b 77.8 [ %] 98.8 [ %] 87.5 [ %] 97.6 [ %] Readers combined Group 1 a 100 [ %] 97.1 [ %] 77.8 [ %] 100 [ %] Group 2 b 100 [ %] 97.1 [ %] 75.0 [ %] 100 [ %] p Note The data in brackets are the 95% CIs. a Group 1 subjects underwent 64-MDCT performed with oral and IV contrast material. b Group 2 subjects underwent 64-MDCT performed solely with IV contrast material. the 14 patients with appendicitis, five patients (36%) had oral contrast material opacification only to the point of the mid or distal small bowel (Fig. 5). The remaining nine patients (64%) were noted to have oral contrast material that extended to the cecum or beyond. Discussion To date, relatively few trials comparing various CT protocols for evaluating nontraumatic abdominal pain in patients presenting to the ED have been published [5]. In light of advancing MDCT technology and the increasing significance of ED length of stay as well as patient discomfort related to ingesting oral contrast material in those presenting with acute abdominal symptoms, we sought to critically evaluate whether oral contrast material continues to be needed in this patient population of ED patients presenting with abdominal pain, which represents a large proportion of patients imaged with CT [7]. With regard to the need for IV contrast material in the evaluation of nontraumatic abdominal pain, studies have shown increased sensitivity in diagnosing appendicitis with the administration of IV contrast material [16]. IV contrast material has also been proven to be particularly useful in diagnosing the complications of common causes of abdominal pain [4, 17]. Therefore, we believe that the benefits afforded by the use of IV contrast material outweigh the risks in patients with nontraumatic abdominal pain. Appendicitis The benefits of oral contrast material for simplifying the diagnosis of appendicitis have been questioned [18]. Investigators have shown that oral contrast material does not reliably fill the appendiceal lumen of healthy patients and, therefore, a lack of filling is not always an indicator of acute appendicitis [19]. Although filling of the appendix may help to exclude appendicitis, this is shown to occur in only 61% of normal appendices [19]. The most useful CT findings in diagnosing appendicitis that have been described include an enlarged appendix, appendiceal wall thickening, periappendiceal fat stranding, and appendiceal wall enhancement, none of which is particularly aided by the administration of oral contrast material [20]. Interestingly, the combined interpretation results for the comparison of the sensitivity and specificity of the two groups in diagnosing appendicitis were similar; however, the sensitivity of each individual reader is, in all cases, lower, but the sample size limits the conclusions that can be drawn from this finding. Nevertheless, this is an interesting finding that deserves further research and that has previously been shown based on a systematic review of the literature [18]. In 32% of the patients in group 1, the oral contrast material did not extend to the level of the colon. Although this represents a confounder in comparing the two groups of patients, the oral contrast group did receive a controlled, 2-hour preparatory time, which is routine in many institutions. Therefore, the position of the oral contrast material likely reflects clinical practice. Interobserver Agreement Interobserver agreement ranged significantly, from poor to very good. This finding may partly reflect the inherent subjectivity involved in the diagnosis of appendicitis on CT. The lack of pertinent clinical information, which is often particularly useful in the diagnosis of appendicitis, was a possible contributing factor in the wide range of the kappa values. All three radiologists had experience interpreting abdominopelvic CT examinations with IV contrast material in the absence of oral contrast material (3 7 years TABLE 3: Kappa Values for Interobserver Agreement Between Observer Pairs for the Diagnosis of Appendicitis Observer Pair Group 1 a κ Group 2 b 1 and and and a Group 1 subjects underwent 64-MDCT performed with oral and IV contrast material. b Group 2 subjects underwent 64-MDCT performed solely with IV contrast material. AJR:193, November

7 Anderson et al. of experience) because our blunt trauma CT protocols do not use oral contrast material. Limitations Limitations of this study include the fact that the expert emergency radiologists were asked to evaluate for the diagnosis of appendicitis, thereby priming them to scrutinize the CT scans for this finding. Although possibly increasing sensitivity over what may be expected in practice, such a study design may also lower specificity because of overcalling this diagnosis. The fact that this study enrolled a convenience sample of patients, given the limited recruitment times in the ED, imposes a limitation: Namely, the sample does not represent consecutive enrollment and may inaccurately represent our larger patient population. For example, it may be easier for employed patients to visit the ED before work or on the weekends, yielding a somewhat skewed patient sample. Finally, the study was powered assuming a prevalence of 40% of any one of three diagnoses (appendicitis, diverticulitis, or smallbowel obstruction). For the purposes of this article, the diagnosis of appendicitis was evaluated. For an evaluation of a single diagnosis, the study is slightly overpowered to detect a difference in specificity and, conversely, is slightly underpowered to detect a difference in sensitivity. We have greater than 80% power to detect between-group differences in specificity and less than 80% power to detect differences in sensitivity. Comparisons of diagnostic sensitivity for individual diagnoses or individual readers are also underpowered; thus, a failure to reject the null hypothesis of no difference cannot necessarily be interpreted to mean there is no true difference between groups. Instead, it may reflect the relatively low power for those particular comparisons. The sample size limitations are also reflected in the wide 95% CIs, another significant limitation in drawing conclusions from comparisons between the two groups. Previously published investigations suggest that the difference in sensitivity or specificity for using or not using enteric contrast material in the diagnosis of appendicitis is less than 10% [18]. Our study was not powered to detect a difference of this magnitude. However, ours is the largest study, to our knowledge, that has directly compared the use of oral and IV contrast material with the use of IV contrast material alone for abdominal imaging in the ED setting. In conclusion, in patients presenting with acute nontraumatic abdominal pain, there is no statistically significant difference in specificity for diagnosing appendicitis identified in our study between CT scans obtained after the administration of oral and IV contrast media versus those obtained with the use of only IV contrast material. Given the statistical limitations conferred by the power of the study and sample size, we cannot make definitive conclusions regarding the comparative sensitivity of the two contrast preparations. References 1. Johnson PT, Horton KM, Mahesh M, Fishman EK. Multidetector computed tomography for suspected appendicitis: multi-institutional survey of 16-MDCT data acquisition protocols and review of pertinent literature. J Comput Assist Tomogr 2006; 30: Kaewlai R, Nazinitsky KJ. Acute colonic diverticulitis in a community-based hospital: CT evaluation in 138 patients. Emerg Radiol 2007; 13: Thoeni RF, Cello JP. CT imaging of colitis. Radiology 2006; 240: Sheedy SP, Earnest F 4th, Fletcher JG, Fidler JL, Hoskin TL. CT of small-bowel ischemia associated with obstruction in emergency department patients: diagnostic performance evaluation. Radiology 2006; 241: Lee SY, Coughlin B, Wolfe JM, Polino J, Blank FS, Smithline HA. Prospective comparison of helical CT of the abdomen and pelvis without and with oral contrast in assessing acute abdominal pain in adult emergency department patients. Emerg Radiol 2006; 12: Mun S, Ernst RD, Chen K, Oto A, Shah S, Mileski WJ. Rapid CT diagnosis of acute appendicitis with IV contrast material. Emerg Radiol 2006; 12: Huynh LN, Coughlin BF, Wolfe J, Blank F, Lee SY, Smithline HA. Patient encounter time intervals in the evaluation of emergency department patients requiring abdominopelvic CT: oral contrast versus no contrast. Emerg Radiol 2004; 10: Berg ER, Mehta SD, Mitchell P, Soto J, Oyama L, Ulrich A. Length of stay by route of contrast administration for diagnosis of appendicitis by computed-tomography scan. Acad Emerg Med 2006; 13: Zangos S, Steenburg SD, Philips KD, et al. Acute abdomen: added diagnostic value of coronal reformations with 64-slice multidetector row computed tomography. Acad Radiol 2007; 14: Paulson EK, Harris JP, Jaffe TA, Haugan PA, Nelson RC. Acute appendicitis: added diagnostic value of coronal reformations from isotropic voxels at multi-detector row CT. Radiology 2005; 235: Jaffe TA, Martin LC, Thomas J, Adamson AR, DeLong DM, Paulson EK. Small-bowel obstruction: coronal reformations from isotropic voxels at 16-section multi-detector row CT. Radiology 2006; 238: Fleiss JL. Measuring nominal scale agreement among many raters. Psychol Bull 1971; 76: Sing T, Sander O, Beerenwinkel N, Lengauer T. ROCR: visualizing classifier performance in R. Bioinformatics 2005; 21: Metropolis N, Ulam S. The Monte Carlo method. J Am Stat Assoc 1949; 44: Robert CP, Casella G. Monte Carlo statistical methods, 2nd ed. Berlin, Germany: Springer-Verlag, Jacobs JE, Birnbaum BA, Macari M, et al. Acute appendicitis: comparison of helical CT diagnosis focused technique with oral contrast material versus nonfocused technique with oral and intravenous contrast material. Radiology 2001; 220: Chou CK, Wu RH, Mak CW, Lin MP. Clinical significance of poor CT enhancement of the thickened small-bowel wall in patients with acute abdominal pain. AJR 2006; 186: Anderson BA, Salem L, Flum DR. A systematic review of whether oral contrast is necessary for the computed tomography diagnosis of appendicitis in adults. Am J Surg 2005; 190: Rao PM, Rhea JT, Novelline RA, et al. Helical CT technique for the diagnosis of appendicitis: prospective evaluation of a focused appendix CT examination. Radiology 1997; 202: Choi D, Park H, Lee YR, et al. The most useful findings for diagnosing acute appendicitis on contrast-enhanced helical CT. Acta Radiol 2003; 44: Kircher MF, Rhea JT, Kihiczak D, Novelline RA. Frequency, sensitivity, and specificity of individual signs of diverticulitis on thin-section helical CT with colonic contrast material: experience with 312 cases. AJR 2002; 178: Cui L, Hung HMJ, Wang SJ. Modification of sample size in group sequential clinical trials. Biometrics 1999; 55: FOR YOUR INFORMATION The reader s attention is directed to the commentary on this article and the article by Keyzer et al., which appear on the preceding pages AJR:193, November 2009