Fig. 2

ROC curves for combination of clinical characteristics for the diagnosis of encephalitis in an encephalopathic population; Legend: Figure showing ROC curves of the model with the highest diagnostic accuracy of predicting encephalitis from regression analyses. AUC (95% CI) of predicted probabilities; fever and nausea: 0.76 (95% CI, 0.65–0.87); fever, nausea and ESR <17 mm/hr: 0.85 (95% CI, 0.76–0.94). When including the finding of CSF WBC of 5- 100 ×10⁶/L to the final model, an AUC of 0.9 (95% CI, 0.0.81–0.98) was reached