Suspected appendicitis in children: diagnosis with contrast-enhanced versus nonenhanced Helical CT.

PURPOSE

To compare the diagnostic accuracy of limited-area (lower abdominal) nonenhanced helical computed tomography (CT), intravenous contrast material-enhanced helical CT of the entire abdomen, and the combination of both.

MATERIALS AND METHODS

Three hundred six children suspected of having appendicitis, who were previously included in a prospective study, underwent limited-area nonenhanced helical CT of the lower abdomen and contrast-enhanced CT of the entire abdomen. No oral or rectal contrast material was administered. The CT scans were retrospectively reviewed by three independent readers both separately and together. The readers were blinded to all clinical information and to the results of previous ultrasonographic and CT examinations. Final diagnoses were established on the basis of surgical, histopathologic, or clinical follow-up findings. The Pearson chi(2) test was performed to compare values between groups. The Student two-sample t test was performed to determine statistically significant differences in age and sex.

RESULTS

One hundred twenty-nine patients (42%) had appendicitis. Readers diagnosed appendicitis with 66% pooled sensitivity and 96% pooled specificity with limited-area nonenhanced CT. With contrast-enhanced CT of the entire abdomen, appendicitis was diagnosed with 90% pooled sensitivity and 94% pooled specificity. With both sequences together, readers diagnosed appendicitis with 90% pooled sensitivity and 94% pooled specificity. The difference between the sensitivity of limited-area nonenhanced CT and that of contrast-enhanced CT was statistically significant (P <.001).

CONCLUSION

Sensitivity of helical CT for suspected appendicitis in children improved significantly with abdominal contrast-enhanced CT compared with limited-area nonenhanced CT. No further improvement in sensitivity was achieved with the combination of both sequences in comparison to that with contrast-enhanced CT alone.