Twelve studies (n=1,172) of CT were included, of which three were RCTs (n=192) and the rest were prospective cohort studies (n=980). Fourteen studies of ultrasonography (all prospective cohort studies; n=1,516) were included. Four of the included studies (including one RCT) directly compared CT and ultrasonography.
All of the included studies were at risk of differential reference standard bias (differing reference standards for patients with positive or negative test results). Most studies did not provide a full description of the reference standard and whether its interpretation was conducted blind to the imaging test results and other clinical information. Only a few studies reported the disease spectrum.
There was no evidence of statistical heterogeneity amongst the positive or negative LRs (P>0.2, I-squared 0%) for the studies of CT. However, there was significant evidence of statistical heterogeneity for the studies of ultrasonography studies (positive LR, P<0.001, I-squared 85%; negative LR, P<0.001, I-squared 75%).
For the detection of appendicitis, CT had pooled estimates of 0.94 (95% confidence interval, CI: 0.91, 0.95) for sensitivity, 0.95 (95% CI: 0.93, 0.96) for specificity, 13.3 (95% CI: 9.9, 17.9) for the positive LR and 0.09 (95% CI: 0.07, 0.12) for the negative LR. Ultrasonography had pooled estimates of 0.86 (95% CI: 0.83, 0.88) for sensitivity, 0.81 (95% CI: 0.78, 0.84) for specificity, 5.8 (95% CI: 3.5, 9.5) for the positive LR and 0.19 (95% CI: 0.13, 0.27) for the negative LR.
For the four studies that directly compared CT against ultrasonography, CT demonstrated better diagnostic performance. The results of the other subgroup analyses were similar to the main analyses. The sensitivity analyses of the treatment of uninterpretable results demonstrated pooled sensitivities and specificities of, respectively, 0.90 (95% CI: 0.86, 0.92) and 0.92 (95% CI: 0.89, 0.93) for CT, and 0.84 (95% CI: 0.80, 0.86) and 0.79 (95% CI: 0.76, 0.82) for ultrasonography.