Figure 3.

A novel severity scoring system for dogs with heatstroke upon admission to the hospital; In the left side, 4 receiver operator (ROC) analyses curves for 4 models done evaluate the performance of the proposed scoring system in 30 dogs with naturally occurring heatstroke upon admission to the hospital. The area under the ROC curve (AUC) for the final predictive score using model A as a predictor of the outcome (i.e., survival vs. death) was 0.92 (95% confidence interval (0.86-0.99), with an optimal cutoff point of 35.0, corresponding to sensitivity and specificity of 93% and 86%, respectively. In the right side, the graph depicts the probability of survival against the predictive score upon admission to the hospital. The probability of survival is 50% with a predictive score of 35.0. * Receiver Operator Characteristic- implies for the best sensitivity and specificity at various threshold points. (From Segev et al. JVECCS 2016).³⁸ © Wiley-Blackwell. Reproduced by permission of Wiley-Blackwell. Permission to reuse must be obtained from the rightsholder.