Radiology
Nicolas Kessler, MD Catherine Cyteval, MD, PhD Benoıˆt Gallix, MD, PhD Alvian Lesnik, MD Paul-Marie Blayac, MD Joseph Pujol, MD Jean-Michel Bruel, MD Patrice Taourel, MD, PhD
Index terms: Appendicitis, 751.291 Appendix, US, 751.12981, 751.12983 Ultrasound (US), comparative studies Ultrasound (US), Doppler studies Published online before print 10.1148/radiol.2302021520 Radiology 2004; 230:472– 478 Abbreviations: CRP ⫽ C-reactive protein NPV ⫽ negative predictive value PPV ⫽ positive predictive value WBC ⫽ white blood cell 1
From the Department of Radiology, Lapeyronie Hospital, 371 Avenue du Doyen Gaston Giraud, 34295 Montpellier cedex 5, France (N.K., C.C., A.L., P.M.B, J.P., P.T.) and the Department of Radiology, Saint-Eloi Hospital, Montpellier, France (B.G., J.M.B.). Received November 15, 2002; revision requested January 21, 2003; final revision received May 16; accepted June 25. Address correspondence to P.T. (e-mail:
[email protected]).
Appendicitis: Evaluation of Sensitivity, Specificity, and Predictive Values of US, Doppler US, and Laboratory Findings1 PURPOSE: To evaluate the sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) of ultrasonography (US), Doppler US, and laboratory findings in the diagnosis of appendicitis. MATERIALS AND METHODS: A total of 125 consecutive patients suspected of having appendicitis were prospectively included for US appendiceal (diameter enlarged to 6 mm or greater, intraluminal fluid, lack of compressibility) and periappendiceal (periileal inflammatory changes, cecal wall thickening, periileal lymph nodes, peritoneal fluid) evaluation, Doppler US evaluation (appendiceal wall signal), and laboratory assessment (leukocytosis, C-reactive protein [CRP]). Definite diagnoses were established at surgery in 61 patients, at endoscopy with biopsy in two patients, and at clinical follow-up in 62 patients. RESULTS: The prevalence of appendicitis was 46%. The appendix was identified with US in 86% of the patients, which included 96% of patients with and 72% of patients without appendicitis. The most accurate appendiceal finding for appendicitis was a diameter of 6 mm or larger, with a sensitivity, specificity, NPV, and PPV of 98%. The lack of visualization of the appendix with US had an NPV of 90%. The most accurate periappendiceal finding of appendicitis was the presence of inflammatory fat changes, with an NPV of 91% and a PPV of 76%, whereas other findings had both NPV and PPV less than 65%. An increase in both white blood cell (WBC) count and CRP level had a PPV of 71%, whereas combined normal WBC count and CRP value had an NPV of 84%. CONCLUSION: A threshold 6-mm diameter of the appendix under compression is the most accurate US finding for appendicitis and has high NPV and PPV. ©
Author contributions: Guarantors of integrity of entire study, N.K., P.T.; study concepts, J.M.B., P.T.; study design, N.K., B.G., P.T.; literature research, N.K., J.P.; clinical studies, N.K., J.M.B., P.T.; data acquisition, N.K., A.L., P.M.B., P.T.; data analysis/ interpretation, N.K., P.T.; statistical analysis, J.P., P.T., C.C.; manuscript preparation, N.K., P.T., C.C.; manuscript definition of intellectual content, C.C.; manuscript editing, revision/review, and final version approval, P.T. ©
RSNA, 2003
472
RSNA, 2003
Appendicitis is the most common cause of acute abdominal pain that necessitates surgical intervention in the Western world (1). Clinical diagnosis of acute appendicitis is based primarily on symptoms and physical findings. However, this diagnosis is often difficult, and up to 50% of patients hospitalized for possible appendicitis do not actually have this disorder. Authors of large prospective studies report a 22%–30% removal rate of normal appendices at surgery (2–5). To reduce the frequency of unnecessary appendectomy, the importance of laboratory findings that include both white blood cell (WBC) counts and C-reactive protein (CRP) values has been stressed (6 – 8), and the use of ultrasonography (US) as a diagnostic tool for appendicitis has been widely evaluated (9 –13). Reported US signs of appendicitis can be grouped into the two categories of (a) appendiceal findings and (b) periappendiceal findings, which mainly include inflammatory changes in the right lower abdominal quadrant. Many laboratory findings or US signs are present in most suspected cases of appendicitis (6 –14). Some of these signs, however, are also present in alternative conditions that can clinically mimic appendicitis. To our knowledge, the frequencies of laboratory and US
Radiology
findings in both appendicitis and alternative conditions have not been compared. The purpose of this investigation was to evaluate the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the individual US, Doppler US, and laboratory findings in the diagnosis of appendicitis.
MATERIALS AND METHODS Study Population From September 2000 to June 2001, 125 consecutive patients suspected at clinical evaluation of having appendicitis underwent prospective evaluations, which included laboratory tests (WBC counts and CRP assays) and US of the right side of the abdomen. All patients were selected by the senior physician in the emergency department, and in all enrolled patients, appendicitis was diagnosed in the first three of the differential diagnoses. Our institutional review board (Lapeyronie Hospital) approved the study and waived patient informed consent because blood studies and US are part of the routine assessment in patients with right-side abdominal pain at our institution. The study group included patients aged 15– 83 years (mean age, 29.5 years) and consisted of 67 female patients (mean age, 29.1 years; age range, 15– 83 years) and 58 male patients (mean age, 30.4 years; age range, 15– 80 years).
US Technique All US examinations included in this study were performed by four radiologists who had 2 years (N.K., P.M.B.), 5 years (A.L.), or 10 years (P.T.) of experience in gastrointestinal US examination. In each patient, the abdomen was initially examined at US (Elegra; Siemens, Erlangen, Germany) by using a broadband 4 –7-MHz convex-array transducer. This evaluation was supplemented with US assessment of the appendix and the surrounding region by using a broadband 5–10-MHz linear-array transducer and the graded-compression technique described by Puylaert (14). Color Doppler US was performed at the end of the grayscale US examination by using a low-velocity scale (pulse repetition frequency, 1,500 Hz) and a low wall filter (100 Hz) to detect slow blood flow.
Interpretation Our prospective real-time US assessment of the appendix and of inflammaVolume 230
䡠
Number 2
tory changes in the right lower abdominal quadrant was based on a set of eight criteria derived from reports in the literature (9,15,16): enlarged appendix, fluid in the appendiceal lumen, lack of compressibility of the appendix, color in the appendiceal wall on color Doppler US images, inflammatory changes in perienteric fat in the right lower quadrant, cecal wall thickening, right lower quadrant lymph nodes, and peritoneal fluid. The appendix was considered enlarged when its outer anteroposterior diameter under compression, measured in the transverse plane, was 6 mm or larger. Flow in the appendiceal wall was investigated as follows: Color gain was increased until clutter was observed and then was reduced just enough to remove clutter from the image of the appendix. Inflammatory changes were defined as the presence of an area of regionally increased echogenicity (hyperechoic halo) adjacent to or surrounding the distal ileum wall, cecum, or appendix that possibly contained ill-defined hypoechoic zones. A lymph node in the right lower quadrant was considered clinically important when it measured 5 mm or larger at its smallest diameter (17). Cecal wall thickness from outer wall to luminal surface was measured on transverse sections under compression (18), and thickening was defined as when the cecal wall measured 5 mm or larger. The laboratory assessment included evaluation of WBC counts, which were considered positive for appendicitis when greater than 1010/L, and a CRP assay, which was considered positive when levels were greater than 10 mg/L. The surgeon was aware of the results of both US and laboratory evaluations before the decision to operate was made.
Follow-up Procedures The laboratory and US findings were compared with surgical and pathologic results when laparotomy was performed. For all patients who did not undergo surgery, follow-up was conducted by reviewing hospital notes and telephone calls to the patients; the latest call was made at least 6 months after the patient’s inclusion in the study. No patients were lost to follow-up.
Statistical Analysis The sensitivity, specificity, accuracy, PPV, and NPV of each US, Doppler US, and laboratory finding in the diagnosis of acute appendicitis were calculated. For each test characteristic, the 95% CI was
TABLE 1 Final Diagnosis in Patients without Appendicitis Final Diagnosis
No. of Patients (n ⫽ 68)
Nonspecific abdominal pain Mesenteric adenitis Ileitis and ileocolitis Gynecologic disease Gastroenteritis Pain colic Psoas hematoma Cystitis Mesenteric ischemia Prostatitis Sigmoid diverticulitis Gastric ulcer
26 13 9 8 5 1 1 1 1 1 1 1
determined by using the standard normal approximation of the binomial distribution table. In addition, we attempted to identify combinations of biologic findings and combinations of US findings that provided the best NPV or PPV. The McNemar test for paired data was used to investigate differences between the NPVs and PPVs of US and laboratory tests in the diagnosis of appendicitis. Differences were considered to be statistically significant when P values were less than .05.
RESULTS US Findings Of the 125 patients clinically suspected of having appendicitis, 57 had acute appendicitis and 68 did not (prevalence, 46%). In all 57 patients with acute appendicitis, the diagnosis was confirmed with surgery and histologic evaluation. In the appendicitis group, the surgical and histologic findings showed perforation of the appendix in 15 patients. Among the 68 patients without acute appendicitis, the diagnosis was confirmed at surgery in four patients, at endoscopy with biopsy in two patients, and at clinical follow-up in 62 patients. Table 1 lists the final diagnoses established in the nonappendicitis group. According to the final diagnosis, all patients who did not undergo surgery had resolution of symptoms in a period ranging from 2 hours to 1 month after inclusion in the study. The appendix was identified at US assessment in 55 (96%) of 57 patients with appendicitis (Fig 1), whereas it was identified in 49 (72%) of 68 patients without appendicitis (Fig 2). Thus, our entire study group included 104 (86%) patients in whom the appendix was identified by Diagnosis of Appendicitis with US
䡠
473
Radiology
levels (95% vs 47%) and had a significantly higher PPV (95% vs 71%). The USaided identification of a normal appendix was a significantly more common finding for the exclusion of appendicitis than was the normality of both WBC and CRP levels (72% vs 47%) and had a significantly higher NPV (98% vs 84%).
DISCUSSION
Figure 1. (a) Transverse and (b) longitudinal US images obtained in a 27-year-old man with appendicitis (arrows). The appendix has an anteroposterior diameter of 9.2 mm.
using US and 21 patients in whom it was not identified. Of these latter 21 patients, two had a final diagnosis of appendicitis, thus giving the lack of visualization of the appendix at US an NPV of 90%; each of these two patients was found to have an appendiceal perforation at surgery. The 104 patients in whom the appendix was visualized constituted a group in which appendiceal US findings could be tested. Table 2 shows (a) the frequency with which each appendiceal finding was interpreted as positive or negative, (b) the number of true-positive, true-negative, false-positive, and false-negative results, and (c) the sensitivity, specificity, accuracy, NPV, and PPV of each appendiceal finding. The two most accurate appendiceal findings for appendicitis were a diameter of 6 mm or larger and a lack of compressibility. In the patient with appendicitis and an outer appendiceal diameter less than 6 mm, surgical and pathologic examination revealed distal appendicitis, but, although the distal appendix was dilated in comparison with the proximal one, both the proximal and distal appendix measured less than 6 mm at US (Fig 3). In the nonappendicitis group, the appendiceal diameter was 6 mm or larger in one patient who had a final diagnosis of pyelonephritis. Fluid in the appendiceal lumen (Fig 4) and flow in the appendiceal wall were also specific findings for appendicitis, with a specificity of 92% and 96%, respectively; however, these two findings were not sensitive since they were encountered in only half of the patients with appendicitis. The periappendiceal US findings were tested in the entire group, as listed in Table 3. Fat changes in the right lower abdominal quadrant were present in 91% of the appendicitis group but also in 24% of the nonappendicitis group, which in474
䡠
Radiology
䡠
February 2004
cluded eight patients with ileocolitis (Fig 5), four patients with mesenteric adenitis, one patient with mesenteric ischemia, and one patient with pyelonephritis. Cecal wall thickening (Fig 5) was detected at US in only 25% of patients with appendicitis and in more than 10% of patients without appendicitis. Peritoneal fluid was noted in 51% of patients with appendicitis but also in almost 30% of patients without appendicitis. Right lower abdominal quadrant adenopathy was present in 32% of patients with appendicitis and in 38% of patients without appendicitis. Combining the nonappendiceal findings with appendiceal findings did not increase the NPV or PPV of individual appendiceal findings, such as an appendix 6 mm or larger in diameter or noncompressibility of the appendix.
Laboratory Findings The number of true-positive, true-negative, false-positive, and false-negative results and the sensitivity, specificity, accuracy, NPV, and PPV of the two laboratory findings (ie, CRP and WBC levels) are shown in Table 4. A WBC level above 1010/L had a sensitivity of 77% and a specificity of 63%, whereas a CRP level above 10 mg/L had a sensitivity of 60% and a specificity of 68%. Both WBC and CRP values were elevated in 27 patients with appendicitis and 11 patients without it, with a PPV of 71% for the increase in both biologic values; both values were normal in 32 patients without appendicitis and in six patients with it, with an NPV of 84% for combined normal CRP and WBC values. The identification of an enlarged appendix at US was a significantly more common finding for appendicitis than was the increase of both WBC and CRP
The inability to visualize the normal appendix is classically considered a major weakness of using US in the assessment of patients suspected of having appendicitis, because it represents a serious limitation to confidently excluding the diagnosis of appendicitis (5). In their state-ofthe-art article, Birnbaum and Wilson (9) claimed that in their experience and in that of others (11,15), a normal appendix is visualized in only 0%– 4% of cases in the adult population, regardless of the US technique used, and they stated that the results of Rioux (12), who visualized a normal appendix in 82% of patients without acute appendicitis, were “amazing.” In contrast with this classic viewpoint, we visualized a normal appendix in 72% of the patients without appendicitis, which is a rate close to the 64% appendix visualization rate obtained by Rettenbacher et al (19) in a population of healthy subjects. The notion that the normal appendix is seldom visualized at US is based on reports published more than 10 years ago (10,14) or on data obtained by sonographers who are not radiologists with experience in US assessment of the gastrointestinal tract (20). Technologic advances combined with indepth radiologic experience have dramatically improved the use of US in the visualization of a normal appendix. The same improvement has been reported with use of computed tomography (CT); the normal appendix was visualized in more than 95% of cases published in 1997 (21), which is in contrast with the visualization in 51% of cases reported in 1991 (22). We found that identification of an appendix measuring less than 6 mm in diameter was a very accurate indication to exclude appendicitis, with an NPV of 98%. Rettenbacher et al (19) obtained an NPV of 100% with this sign, whereas Rioux (12) reported an NPV of 98% with it. Concerning the nonvisualization of the appendix, we obtained a high NPV for appendicitis (90%); the same results were reported in the studies with a high Kessler et al
Radiology Figure 2.
(a– c) Longitudinal US images show a normal appendix (arrows) in three different patients.
normal appendix visualization rate (12). As a result, nonvisualization of the appendix can only be valid as an accurate finding to exclude appendicitis for sonographers who can usually identify a normal appendix. US evaluation of the appendix ideally includes evaluation of the appendiceal wall and appendiceal content. We decided to measure the outer appendiceal diameter rather than appendiceal wall thickness for two reasons. First, as shown by Rioux (12), inflammation of the appendiceal wall may be indistinguishable from hypoechoic intraluminal pus, thus making measurement of the appendiceal wall inaccurate. Second, the mucosal surface may be difficult to identify within the appendix. The threshold diameter of 6 mm, which is the most commonly reported threshold, had both high NPV and PPV (98%) in our study. The high PPV is out of line with the data obtained by Rettenbacher et al (19), who reported an appendiceal diameter of 6 mm or larger in 32% of symptomatic patients without appendicitis in whom the appendix was identified. Rioux (12) obtained intermediate results with an appendiceal diameter of more than 6 mm in 6% of patients without appendicitis. We assume that this discrepancy is due to differences in measurement of appendiceal diameter. We thus performed measurements under maximal compression to standardize the measurement because the relevant anteroposterior diameter of a compressible appendix may vary according to the graded compression applied to the abdominal wall, and we hypothesize that we may have compressed some loose feces or air out of the normal patent lumen. We found almost identical PPV and NPV for an appendiceal diameter of 6 mm or larger under compression and for the lack of compression of the Volume 230
䡠
Number 2
TABLE 2 Appendiceal US and Doppler US Signs for the Diagnosis of Appendicitis Finding and Value Finding at US Positive Negative Finding at final diagnosis True-positive True-negative False-positive False-negative Value* Sensitivity Specificity PPV NPV Accuracy
Diameter ⱖ 6 mm
Lack of Compressibility
Intraluminal Fluid
Flow in Wall
55 50
55 49
33 71
30 74
54 48 1 1
53 47 2 2
29 45 4 26
28 47 2 26
98 (95, 100) 98 (95, 100) 98 (95, 100) 98 (95, 100) 97 (94, 100)
96 (94, 100) 96 (94, 100) 96 (94, 100) 96 (94, 100) 96 (94, 100)
53 (43, 63) 92 (87, 97) 88 (82, 94) 63 (64, 72) 71 (62, 80)
52 (43, 61) 96 (94, 100) 93 (88, 98) 64 (55, 73) 73 (64, 81)
Note.—Data obtained in 104 patients in whom the appendix could be visualized at US. Unless otherwise noted, data are numbers of patients. * Data are percentages, and data in parentheses are 95% CIs.
appendix, but these two findings are very close since, if compressible, the appendiceal diameter will likely be less than 6 mm under compression. We found that hyperemia in the appendiceal wall shown on color Doppler images was a specific finding for appendicitis that was encountered in only two of the patients without appendicitis. The same high specificity was already reported in previously published studies in which flow was never identified in the normal appendiceal wall (16,23). Our evaluation of the content of the appendiceal lumen focused on the presence of intraluminal fluid as a sign of appendicitis, whereas Rettenbacher et al (24) considered the absence of gas in the appendiceal lumen as a criterion for appendicitis. The same mechanism might explain both the presence of fluid and the absence of gas in an inflamed appen-
dix. Obstruction, which is the most common cause of appendicitis, could lead to retention of pus or appendiceal secretion with resorption of intraluminal gas (24). We did not evaluate the presence or absence of gas in the appendix because we considered that the US appearance of a tiny appendicolith or a small amount of feces could resemble gas. The clinical importance of appendiceal air is under debate (25), but we believe that its evaluation is easier at CT than at US. By contrast, appendiceal fluid, which is a finding that has never been evaluated to our knowledge, is easier to identify, and its presence could be a useful ancillary sign. Among right lower abdominal quadrant changes, inflamed fat has been concluded to be 100% sensitive but not a specific sign of appendicitis at CT (21). It is well known that inflamed fat in the Diagnosis of Appendicitis with US
䡠
475
Radiology Figure 3. Longitudinal US image shows appendicitis in a 43-year-old woman. This was a false-negative diagnosis at US. The appendiceal diameter measured less than 6 mm, but after a retrospective review of this longitudinal view, the distal appendix (arrowheads) was dilated compared to the proximal appendix (arrows).
Figure 4. US image shows appendicitis in a 26-yearold man. Note the hypoechoic fluid in the appendiceal lumen (ⴱ) and the presence of a hyperechoic appendicolith (arrows).
TABLE 3 Periappendiceal US Signs for the Diagnosis of Appendicitis
Finding and Value Finding at US Positive Negative Finding at final diagnosis True-positive True-negative False-positive False-negative Value* Sensitivity Specificity PPV NPV Accuracy
Inflammatory Fat Changes
Cecal Wall Thickening
Periileal Lymph Nodes
Peritoneal Fluid
68 57
22 103
44 81
49 76
52 52 16 5
14 60 8 43
18 42 26 39
29 48 20 28
91 (86, 96) 76 (69, 83) 76 (69, 83) 91 (86, 96) 83 (76, 90)
25 (17, 33) 88 (82, 94) 64 (56, 72) 58 (49, 67) 59 (50, 68)
32 (24, 40) 62 (53, 71) 41 (32, 50) 52 (43, 61) 48 (39, 57)
51 (42, 60) 71 (63, 79) 59 (50, 68) 63 (55, 71) 62 (53, 71)
Note.—Data obtained in 125 patients. Unless otherwise noted, data are numbers of patients. * Data are percentages, and data in parentheses are 95% CIs.
right lower quadrant may be present in a broad spectrum of alternative diagnoses to appendicitis (26), and we found inflamed fat in 24% of the patients without appendicitis. We did not detect inflamed fat in every patient with appendicitis, which is contrary to known data from CT studies. However, we hypothesize that some subtle inflammatory changes may have been missed in our US assessment, especially because we did not prospectively analyze the noncompressibility of the fat, which could be an interesting finding for diagnosing inflamed fat. Right lower abdominal quadrant adenopathy is a common reaction to ileal, cecal, or appendiceal inflammatory disease that 476
䡠
Radiology
䡠
February 2004
is encountered in both patients with appendicitis and those without it. We agree with previous conclusions (17,21,27) that the only definitive way to differentiate an appendicitis adenopathy from mesenteric adenitis is to identify either an enlarged inflamed appendix or a normal appendix. Cecal wall changes have been extensively analyzed by using CT, which enables identification of focal cecal apical thickening and arrowhead and cecal bar signs suggestive of appendicitis, whereas circumferential diffuse wall thickening is present in colitis (28,29). However, such evaluations require adequate cecal distention as is obtained by using the CT technique described by Rao
et al (21). Therefore, we limited our evaluation with US to the identification of cecal wall thickening but did not obtain sufficient predictive values to differentiate appendicitis from nonappendicitis. The value of laboratory assessments for appendicitis diagnosis has already been studied. An increase in WBC count was shown to be an early marker of appendiceal inflammation, whereas CRP values were generally elevated when symptoms were present for more than 12 hours (1,30) or after appendiceal perforation or abscess formation (7). The main issue discussed in the literature is the ability of laboratory evaluations to exclude appendicitis when both WBC and CRP levels are normal (7,29,31). In a group of patients that included both children and adults, Gro ¨ nroos and Gro ¨ nroos (7) found a 100% predictive value to exclude appendicitis when both WBC and CRP values were normal, but the same authors did not confirm this result in a population composed exclusively of children (32). In our population, which included patients 15 years and older, the association of a normal WBC count and CRP value had an NPV of 84%. To our knowledge, this is the first study in which the value of both laboratory and imaging findings was evaluated in the same patients. US examinations were superior to laboratory tests for affirming appendicitis, and, more strikingly, US examinations were also superior to laboratory tests for excluding appendicitis. Therefore, we do not recommend using laboratory tests as part of an algorithm to Kessler et al
Radiology
restrict imaging indications for patients with increased WBC count and/or CRP value, as previously suggested (8). On the other hand, in patients whose appendix is not visualized at US, laboratory tests could be performed to strengthen the NPV of nonvisualization of the appendix. However, further studies that include more patients in whom the appendix is not identified at US are necessary to confirm this potential role of laboratory tests. A number of limitations of our study must be considered. First, there was no reliable way to confirm that all patients without appendicitis would have had a normal appendix at histopathologic analysis if surgery had been performed. Some of these patients might represent unrecognized cases of self-limiting appendicitis. The concept of spontaneously resolving appendicitis is now supported by evidence (33,34), and for each sign evaluated in our study, the number of false-negative results may have been underestimated and the number of falsepositive results overestimated. However, the same limitations apply to the results of other studies published on the same topic because not all patients suspected of having appendicitis undergo surgery. Second, our study design was based on a prospective evaluation of several US and biologic criteria. We are aware that some additional criteria not included in our protocol may have been interesting to evaluate, especially the thickness of the appendiceal wall, the presence of air in the appendiceal lumen, and the noncompressibility of the periappendiceal fat. Third, we compared US data obtained by experienced radiologists involved in US examination of the gastrointestinal tract with laboratory data obtained from standard laboratory tests. It would be interesting to compare laboratory tests and US examinations performed by medical professionals with less experience in gastrointestinal US, such as residents or US technologists, thus more closely representing the most common clinical setting. In conclusion, an appendix with a threshold anteroposterior diameter of 6 mm under compression is the most indicative US finding for appendicitis, with high NPV and PPV. When the appendix is identified, the evaluation of periappendiceal findings does not improve the usefulness of US. Finally, the use of laboratory tests does not exclude the need for US examinations in patients with normal laboratory values. Volume 230
䡠
Number 2
Figure 5. Transverse US image shows hyperechoic fat infiltration (ⴱ) in a 38-year-old woman with an ileocolitis. Note the thickening of the cecum (arrowheads).
TABLE 4 Laboratory Signs for the Diagnosis of Appendicitis
Finding and Value Finding at laboratory evaluation Positive Negative Finding at final diagnosis True-positive True-negative False-positive False-negative Value‡ Sensitivity Specificity PPV NPV Accuracy
Elevated WBC and/ or CRP Levels
Elevated WBC Level*
Elevated CRP Level†
Elevated WBC and CRP Levels
69 56
56 69
38 87
87 38
44 43 25 13
34 46 22 23
27 57 11 30
51 32 36 6
77 (70, 84) 63 (55, 71) 64 (56, 72) 77 (70, 84) 70 (62, 78)
60 (51, 69) 68 (60, 76) 61 (52, 70) 67 (59, 75) 64 (56, 72)
47 (38, 56) 84 (78, 90) 71 (63, 79) 66 (58, 74) 67 (59, 75)
88 (82, 94) 53 (44, 62) 59 (50, 68) 84 (78, 90) 66 (58, 84)
Note.—Data obtained in 125 patients. Unless otherwise noted, data are numbers of patients. * ⱖ1010/L. † ⱖ10 mg/L. ‡ Data are percentages, and data in parentheses are 95% CIs.
References 1. Graffeo CS, Counselman FL. Appendicitis. Emerg Med Clin North Am 1996; 14: 653–71. 2. de Dombal FT, Leaper DJ, Staniland JR, McCann AP, Horrocks JC. Computeraided diagnosis of acute abdominal pain. BMJ 1972; 2:9 –13. 3. Deutsch A, Shani N, Reiss R. Are some appendectomies unnecessary: an analysis of 319 white appendices. J R Coll Surg Edinb 1983; 28:35– 40. 4. Simmen HP, Decurtins M, Rotzer A, Duff C, Brutsch HP, Largiader F. Emergency room patients with abdominal pain unrelated to trauma: prospective analysis in a surgical university hospital. Hepatogastroenterology 1991; 38:279 –282. 5. Rao PM, Rhea JT, Novelline RA. Helical CT of appendicitis and diverticulitis. Radiol Clin North Am 1999; 37:895–910.
6. 7.
8.
9. 10. 11.
Gro ¨ nroos JM, Gro ¨ nroos P. Diagnosis of acute appendicitis. Radiology 2001; 219: 297–298. Gro ¨ nroos JM, Gro ¨ nroos P. Leucocyte count and C reactive protein in the diagnosis of acute appendicitis. Br J Surg 1999; 86:501–504. Gro ¨ nroos JM, Gro ¨ nroos P. A fertile-aged woman with right lower abdominal pain but unelevated leukocyte count and C-reactive protein: acute appendicitis is very unlikely. Langenbecks Arch Surg 1999; 384:437– 440. Birnbaum BA, Wilson SR. Appendicitis at the millennium. Radiology 2000; 215:337– 348. Jeffrey RB Jr, Laing FC, Lewis FR. Acute appendicitis: high-resolution real-time US findings. Radiology 1987; 163:11–14. Puylaert JBCM, Rutgers PH, Lalisang RI, et al. A prospective study of ultrasonograDiagnosis of Appendicitis with US
䡠
477
20.
phy in the diagnosis of appendicitis. N Engl J Med 1987; 317:666 – 669. Rioux M. Sonographic detection of the normal and abnormal appendix. AJR Am J Roentgenol 1992; 158:773–778. Lim HK, Lee WJ, Lee SJ, Namgung S, Lim JH. Focal appendicitis confined to the tip: diagnosis at US. Radiology 1996; 200: 799 – 801. Puylaert JB. Acute appendicitis: US evaluation using graded compression. Radiology 1986; 158:355–360. Jeffrey RB Jr, Laing FC, Townsend RR. Acute appendicitis: sonographic criteria based on 250 cases. Radiology 1988; 167: 327–329. Lim HK, Lee WJ, Kim TH, Namgung S, Lee SJ, Lim JH. Appendicitis: usefulness of color doppler US. Radiology 1996; 201: 221–225. Sivit CJ, Newman KD, Chandra RS. Visualization of enlarged mesenteric lymph nodes at US examination: clinical significance. Pediatr Radiol 1993; 23:471– 475. Siegel MJ, Friedland JA, Hidebolt CF. Bowel wall thickening in children: differentiation with US. Radiology 1997; 203: 631– 635. Rettenbacher T, Hollerweger A, Macheiner P, et al. Outer diameter of the vermiform appendix as a sign of acute appendicitis: evaluation at US. Radiology 2001; 218:757–762. Wise SW, Labuski MR, Kasales CJ, et al.
478
䡠
12.
Radiology
13.
14. 15.
16.
17.
18.
19.
Radiology
䡠
February 2004
21.
22.
23. 24.
25.
26.
27.
Comparative assessment of CT and sonographic techniques for appendiceal imaging. AJR Am J Roentgenol 2001; 176:933– 941. Rao PM, Rhea JT, Novelline RA. Sensitivity and specificity of the individual CT signs of appendicitis: experience with 200 helical appendiceal CT examinations. J Comput Assist Tomogr 1997; 21:686 – 692. Grosskreutz S, Goff WB, Balsara Z, Burkhard TK. CT of the normal appendix. J Comput Assist Tomogr 1991; 15:575– 577. Quillin SP, Siegel MJ. Appendicitis: efficacy of color Doppler sonography. Radiology 1994; 191:557–560. Rettenbacher T, Hollerweger A, Macheiner P, et al. Presence or absence of gas in the appendix: additional criteria to rule out or confirm acute appendicitis— evaluation with US. Radiology 2000; 214: 183–187. Rao PM. Presence or absence of gas in the appendix: additional criteria to rule out or confirm acute appendicitis— evaluation with US (letter). Radiology 2000; 217:599 – 600. Jain KA, Quam JP, Ablin DS, Gerscovich EO, Shelton DK. Imaging findings in patients with right lower quadrant pain: alter diagnoses to appendicitis. J Comput Assist Tomogr 1997; 21:693– 698. Rao PM, Rhea J, Novelline RA. CT diag-
28.
29.
30.
31.
32.
33.
34.
nosis of mesenteric adenitis. Radiology 1997; 202:145–149. Rao PM, Rhea JT, Novelline RA, et al. Helical CT technique for diagnosing appendicitis: prospective evaluation of a focused appendix CT examination. Radiology 1997; 202:139 –144. Rao PM, Wittenberg J, McDowell RK, Rhea JT, Novelline RA. Appendicitis: use of arrowhead sign for diagnosis at CT. Radiology 1997; 202:363–366. Thimsen DA, Tong GK, Gruenberg JC. Prospective evaluation of C-reactive protein in patients suspected of having acute appendicitis. Am Surg 1989; 55:466 – 468. Asfar S, Safar H, Khoursheed M, Dashti H, al-Bader A. Would measurement of C-reactive protein reduce the rate of negative exploration for acute appendicitis? J R Coll Surg Edinb 2000; 45:21–24. Gro ¨ nroos JM. Do normal leucocyte count and C-reative protrein value exclude acute appendicitis in children? Acta Paedriatr 2001; 90:649 – 651. Migraine S, Atri M, Bret PM, Lough JO, Hinchey JE. Spontaneously resolving acute appendicitis: clinical and sonographic documentation. Radiology 1997; 205:55–58. Cobben LPJ, de van Otterloo AM, Puylaert JBCM. Spontaneously resolving acute appendicitis: frequency and natural history in 60 patients. Radiology 2000; 215:349 –352.
Kessler et al
