Table 2:
ROC analysis: dichotomized favorable outcome (mRS ≤2)a
| Variable | AUC | P Value | 95% CI | OOP | Sensitivity | Specificity |
|---|---|---|---|---|---|---|
| Final infarct (mL) | 0.96 | <.001 | 0.91–1.00 | 29.1 | 91% | 88% |
| RAPID rCBV core | 0.86 | .001 | 0.74–0.96 | 10.4 | 85% | 78% |
| RAPID rCBF core | 0.81 | <.001 | 0.68–0.93 | 5.6 | 73% | 72% |
| RAPID Tmax >4 sec | 0.80 | <.01 | 0.65–0.93 | 146.6 | 73% | 67% |
| RAPID Tmax >6 sec | 0.77 | <.01 | 0.63–0.92 | 81.9 | 77% | 72% |
| RAPID Tmax >8 sec | 0.76 | <.01 | 0.60–0.91 | 53.1 | 77% | 72% |
| RAPID Tmax >10 sec | 0.74 | <.01 | 0.58–0.89 | 26.6 | 73% | 72% |
| NCCT-ASPECTS | 0.72 | .01 | 0.57–0.87 | 9.5 | 68% | 68% |
| CBV-ASPECTS | 0.75 | .01 | 0.61–0.89 | 5.0 | 76% | 62% |
| CBF-ASPECTS | 0.72 | .02 | 0.57–0.87 | 5.0 | 58% | 66% |
| Clot burden score | 0.74 | .01 | 0.59–0.89 | 6.5 | 58% | 77% |
| Collateral scoreb | .72 | .20 | 0.56–0.89 | 1.5 | 70% | 70% |
Note:—OOP indicates optimal operating point, optimal cut-off from ROC analysis for the variable of interest.
a
Results of ROC analysis relative to dichotomized favorable clinical outcome (90-day mRS ≤2). Performance indicated by ROC AUC in descending-order performance.
b
Significant relationships were observed for all tested variables, with the exception of the CTA collateral score.