Nine studies (n=619) were included in the review, and were used to estimate the size-specific sensitivity of CTA.
The size distributions for ruptured aneurysms (2,787 aneurysms) and for unruptured aneurysms (901 aneurysms) were each estimated from 6 studies.
The crude sensitivity ranged from 66 to 98%. The predicted change point in aneurysm size was 7.7 mm (95% confidence interval, CI: 7.0, 8.9), above which false negatives did not occur. The sensitivities stratified by aneurysm size ranged from 53% (95% CI: 44, 62) for 2-mm aneurysms to 95% (95% CI: 92, 97) for 7-mm aneurysms. The crude specificity ranged from 77 to 100%. The overall estimated specificity for the studies was 98.9% (95% CI: 91.5, 99.99), but there was between-study heterogeneity. The rate of aneurysms per case, sensitivity of the study, and study size were not statistically associated with between-study heterogeneity. The medium sensitivity for ruptured aneurysms, as derived from computer simulations, was 92% (95% CI: 90, 94); the corresponding medium sensitivities for unruptured aneurysms and for aneurysms greater than 5 mm were 82% (95% CI: 78, 86) and 92% (95% CI: 89, 95), respectively.
The LR for negative tests was 0.18 for unruptured aneurysms, 0.012 for unruptured aneurysms of at least 6 mm, and 0.081 for ruptured aneurysms. The LR for positive tests for unruptured aneurysms with a low probability ranged from 15 for 2-mm aneurysms, to 61 for 5-mm aneurysms, to 99 for 6-mm aneurysms. The LR for positive results for ruptured aneurysms with a 50% pre-test probability ranged from 3.9 (2-mm aneurysm) to 56 (5-mm aneurysm). The LR for positive results for a solitary ruptured aneurysm with a high pre-test probability (85%) ranged from 3.4 (2-mm aneurysm) to 40 (5-mm aneurysm).
Interpretation of CTA results derived from computer simulations.
Negative CTA with a suspected unruptured aneurysm.
For a low pre-test probability of 2%, a negative CTA resulted in a post-test probability of 0.37%. For a high pre-test probability of 10%, the post-test probability would be 2.0%.
Positive CTA with a suspected unruptured aneurysm. For a low pre-test probability of 2%, the post-test probability of an identified 2-mm aneurysm was 23%, while that of an identified 5-mm aneurysm was 56%. With a high pre-test probability of 10%, the post-test probabilities for a 2-mm and 5-mm aneurysm were 62% and 87%, respectively.
Negative CTA with a suspected ruptured aneurysm. For a pre-test probability of 5%, the post-test probability after a negative CTA was 0.42%. For a pre-test probability of 50%, the post-test probability was 7.5%, while for a high pre-test probability of 90%, the post-test probability was 42%.
Positive CTA with a suspected ruptured aneurysm.
For a low pre-test probability of 5%, the post-test probability of a 2-mm identified aneurysm being ruptured was 18%. A 5-mm identified aneurysm has a probability of 78% of being a ruptured aneurysm. For a pre-test probability of 50%, the post-test probabilities for a 2-mm and 5-mm aneurysm being ruptured were 69% and 92%, respectively. With a pre-test probability of 90%, the post-test probabilities for a 2-mm and 5-mm aneurysm being ruptured were 82% and 99%, respectively. The likelihood of a false-positive aneurysm is very small for aneurysms of greater than 5 mm, identified on CTA.
When screening for an unruptured aneurysm of at least 6 mm, with a low pre-test probability of 2%, a negative CTA resulted in a post-test probability of 0.024%. For a high pre-test probability of 10%, the post-test probability after a negative CTA was 0.13%. For a pre-test probability of 2% the post-test probability of a true-positive was 68%, and for a pre-test probability of 10%, the post-test probability was 92%. False-positives for aneurysms greater than 6 mm are unlikely.