Correspondence

Computed Tomography of the Appendix

N Engl J Med 1998; 338:1847-1848June 18, 1998

Article

To the Editor:

Rao et al. (Jan. 15 issue)1 found that computed tomography (CT) was very accurate for the diagnosis of appendicitis and that clinical diagnosis was very inaccurate. The clinical diagnoses were inaccurate because they were the surgeon's first estimates of the probability of appendicitis in patients in the emergency department. It is well known that the diagnostic accuracy increases on reexamination after a short period of observation. A planned exploration is not necessarily carried out on admission, as assumed by Rao et al. These points alter the calculated cost savings due to the use of CT.

The histologic criteria that they used for the diagnosis of appendicitis need to be reported, especially if mucosal inflammation alone was one of them.2 The possibility of bias during the workup is evident if mucosal inflammation was a criterion for appendicitis.

Does a positive finding on CT scanning mandate an operation, or can the method detect cases of minor appendicitis that would otherwise resolve on their own? The latter have been identified with ultrasonography,3 and we have presented epidemiologic evidence that minor appendicitis is common.4 Its importance can be appreciated from the report that the rate of detection of appendicitis was 50 percent higher in centers with a liberal attitude toward exploration than in those with a more conservative approach to management.5 Information about the distribution of the patients according to the severity of disease would therefore be of interest.

Cost savings in the care of patients suspected of having appendicitis are certain if unnecessary appendectomies are avoided. These savings may be augmented or lost, depending on the number of operations performed in patients with spontaneously resolving appendicitis. Clinical aptitude and in-hospital observation will still be needed until it is shown that CT can differentiate patients who need an operation from patients with resolving appendicitis.

Roland Andersson, M.D.
Ryhov Hospital, S-551 85 Jönköping, Sweden

Per Olof Nyström, M.D., Ph.D.
Gunnar Olaison, M.D., Ph.D.
University Hospital, 581 85 Linkoping, Sweden

5 References

1
Rao PM, Rhea JT, Novelline RA, Mostafavi AA, McCabe CJ. Effect of computed tomography of the appendix on treatment of patients and use of hospital resources. N Engl J Med 1998;338:141-146
Full Text | Web of Science | Medline

2
Pieper R, Kager L, Nasman P. Clinical significance of mucosal inflammation of the vermiform appendix. Ann Surg 1983;197:368-374
CrossRef | Web of Science | Medline

3
Heller MB, Skolnick ML. Ultrasound documentation of spontaneously resolving appendicitis. Am J Emerg Med 1993;11:51-53
CrossRef | Web of Science | Medline

4
Andersson R, Hugander A, Thulin A, Nystrom PO, Olaison G. Indications for operation in suspected appendicitis and incidence of perforation. BMJ 1994;308:107-110
CrossRef | Web of Science | Medline

5
Howie JGR. Too few appendectomies? Lancet 1964;1:1240-1242
CrossRef | Web of Science | Medline

To the Editor:

We believe that Rao et al. might have overstated the accuracy of CT in the diagnosis of appendicitis, which they reported as 98 percent. According to Table 1 of their article, 20 percent of the patients enrolled in the study had symptoms for five or more days. Pain caused by appendicitis that lasts longer than 36 hours often results in perforation1; in one series, a delay of 72 hours or more from the onset of symptoms to surgery resulted in a 90 percent incidence of perforation.2 It seems clear that in patients who have had symptoms for five days or more, the results of abdominal CT will be either impressively positive for appendicitis and its complications or negative. Findings of this nature may thus have led to an increase in the number of either true positive results or true negative results. We would have preferred to see this study performed in patients who had had symptoms for less than 24 hours. These are the patients in whom the likelihood of missing the diagnosis is highest.

M.H. Moustafa, M.D.
John Kare, M.D.
Martin Luther King–Drew Medical Center, Los Angeles, CA 90059

2 References

1
Pieper R, Kager L, Nasman P. Acute appendicitis: a clinical study of 1018 cases of emergency appendectomy. Acta Chir Scand 1982;148:51-62
Medline

2
Von Titte SN, McCabe CJ, Ottinger LW. Delayed appendectomy for appendicitis: causes and consequences. Am J Emerg Med 1996;14:620-622
CrossRef | Web of Science | Medline

To the Editor:

In the series of patients evaluated by Rao et al., CT had excellent specificity and sensitivity, as shown by the nearly perfect 90-degree receiver-operating-characteristic curve shown in Figure 3 of the article. However, the curve representing the accuracy of clinical diagnosis makes it look worse than it actually was. According to Table 3, 18 of the 53 confirmed cases of appendicitis (34 percent) and 5 of the 47 cases in which it was ruled out (11 percent) were rated as “definitely appendicitis” by the clinicians. Thus, the first point on the curve should have been plotted at the intersection of 0.11 and 0.34, considerably to the left of the clinical-likelihood curve in Figure 3.

Furthermore, the curve for the radiologic likelihood includes patients categorized as definitely not having appendicitis, whereas patients categorized clinically as definitely not having appendicitis were not included in the study. Including these patients, who presumably were correctly classified clinically as not having appendicitis, would shift the curve for clinical likelihood farther to the left. Although the data do suggest that the use of helical CT is promising as an aid to the diagnosis of appendicitis, the study design does not permit direct comparison between CT and clinical examination.

Edward P. Frothingham, M.D.
Thomas B. Newman, M.D., M.P.H.
University of California, San Francisco, San Francisco, CA 94143

Author/Editor Response

Dr. Rao replies:

To the Editor: Andersson et al. state that clinical accuracy would have increased after a short period of observation. This may be true, but a delay in diagnosis can be detrimental and costly for patients, whether they occupy emergency department bays or hospital beds.1 On the basis of clinical findings, patients were divided into an observation group (55 patients) and an urgent-appendectomy group (45 patients); CT was performed only after this determination had been made on the basis of all available clinical and laboratory data. The histologic diagnosis of appendicitis was based on the presence of transmural inflammation; the presence of mucosal inflammation alone was insufficient. We are unaware of any clinical or imaging criteria that allow specific identification of spontaneously resolving appendicitis at the time of initial diagnosis.

Moustafa and Kare indicate that the accuracy of appendiceal CT might have been overstated because 20 patients had had symptoms for five or more days. Among these patients, eight had simple appendicitis, two had perforated appendicitis, seven had other conditions, and three had nonspecific abdominal pain; these outcomes were very similar to those in the patients who had symptoms for less than five days. It is well recognized that both chronic appendicitis (continuous symptoms for longer than two weeks) and recurrent appendicitis (one or more episodes of very similar or identical symptoms) occur. There is no significant difference between the appearance of these conditions on CT and that of acute appendicitis.2

Frothingham and Newman suggest that the receiver-operating-characteristic curve for the clinical likelihood of appendicitis understates clinical accuracy. There was a graphing error that slightly understated clinical accuracy in the lower half of the curve and slightly overstated clinical accuracy in the upper half of the curve; however, the area under the curve is unchanged, as is overall clinical accuracy. There was no clinical category entitled “definitely not appendicitis,” because only patients with some clinical likelihood of appendicitis would logically be referred for appendiceal CT.

We should note that the article in the Journal described the value of appendiceal CT in terms of the care of patients suspected of having appendicitis and the use of hospital resources. The detailed CT findings in these patients were reported separately in the American Journal of Roentgenology.3

Patrick M. Rao, M.D.
Massachusetts General Hospital, Boston, MA 02114

3 References

1
Velanovich V, Satava R. Balancing the normal appendectomy rate with the perforated appendicitis rate: implications for quality assurance. Am Surg 1992;58:264-269
Web of Science | Medline

2
Rao PM, Rhea JT, Novelline RA, McCabe CJ. The computed tomography appearance of recurrent and chronic appendicitis. Am J Emerg Med 1998;16:26-33
CrossRef | Web of Science | Medline

3
Rao PM, Rhea JT, Novelline RA, Mostafavi AA, Lawrason JN, McCabe CJ. Helical CT combined with contrast material administered only through the colon for imaging of suspected appendicitis. AJR Am J Roentgenol 1997;169:1275-1280
Web of Science | Medline

Citing Articles (1)