Computed Tomography and Ultrasonography To Detect Appendicitis

TO THE EDITOR:

In their meta-analysis, Terasawa and colleagues (1) suggested that overall, computed tomography (CT) is better than ultrasonography (1) for diagnosing appendicitis in patients with atypical presentations. Their findings will clarify clinical decision making for many but will complicate it for those practicing in low-resource settings. In many low-income countries, CT often costs many times a family's monthly salary. It is important to carefully answer the following questions before we recommend CT as a standard of care.

First, what about cost and cost-effectiveness? In my practice setting, a focal appendiceal CT, which would probably tell the clinician whether the patient has appendicitis, costs about 6 times more than ultrasonography. When confronted with the broader question of what else could be causing the abdominal pain, the cost differences are even greater. A complete abdominal and pelvic CT scan with oral and intravenous contrasts costs about 18 times more than ultrasonography. It may therefore be less expensive to admit or observe a patient than to order a CT. The only patients who would benefit from immediate CT use are those who would receive surgery earlier with CT than with ultrasonography, a relatively small group. Those with negative CT results would still need admission or close follow-up, since no test has 100% specificity.

Second, is CT a better diagnostic test than ultrasonography for all patients, or is CT better in some while ultrasonography is better in others? Current literature needs to better define sensitivities and specificities for subgroups of patients. It would help to know whether CT is as good for a 30-year-old thin man as it is for a 30-year-old woman or a 60-year-old diabetic patient with nonspecific abdominal pain. Another issue is more specifically describing the degree of atypicality or the level of pretest probability. The literature does not determine how decisions should be made when a clinician thinks that a patient may have appendicitis but is not sure versus when he does not think a patient has appendicitis and is not sure.

Third, since most of the literature examined by Terasawa and colleagues originated in high-income countries, would their conclusions hold true for relatively leaner populations in low- and middle-income countries?