Skip to main content
. 2012 Dec 20;2(Suppl 1):S21. doi: 10.1186/2110-5820-2-S1-S21

Figure 7.

Figure 7

ROC curve. Mortality predictive values of IAP. The curve describes the association between sensitivity and specificity at different thresholds of IAP in predicting mortality. Cut-off values shown in the figure are 15.13 mmHg (true positive rate 65%, false positive rate 68.3%), 17.09 mmHg (true positive rate 76.9%, false positive rate 50%) and 18.87 mmHg (true positive rate 82.9%, false positive rate 25%). ROC curves that approach the upper leftmost corner represent highly accurate studies - in this case AUC is 0.703. Accuracy of IAP in predicting mortality is fair, and ROC curve analyses is statistically significant - p value 0.021.