A novel predictor of severe dengue: The aspartate aminotransferase/platelet count ratio index (APRI)

To evaluate risk factors for the development of dengue into severe dengue in Guangdong. A retrospective analysis of clinical data from 212 dengue patients between June and October 2014. A total of 174 (82.1%) patients in our study had classic dengue, of which 38 (17.9%) had severe diseases. The frequencies of jaundice, pleural effusion, ascites, and vaginal bleeding were significantly different between the two groups (P < 0.05). The routine laboratory test results for alanine aminotransferase, aspertate aminotransferase, albumin, leukocyte count, platelet count, activated partial prothrombin time, prothrombin time, and aspartate aminotransferase/platelet count ratio index showed a significant association with severe dengue (P < 0.01). The areas under the receiver operating characteristic curves (AUC) were 0.727 (95% CI 0.662-0.78), 0.699 (95%CI 0.632-0.760), 0.634 (95%CI 0.565-0.698), 0.757 (95%CI 0.694-0.813), 0.775 (95%CI 0.713-0.829), 0.713 (95%CI 0.647-0.773), 0.719 (95%CI 0.730-0.843), and 0.785 (95%CI 0.724-0.893), respectively. The logistic regression analysis identified three factors, including high WBC (OR 1.52), prolonged PT (OR 1.745). and high APRI (OR 1.05) may be associated with the discrimination criteria to identify patients with and without severe diseases. The combination of the three factors (WBC, PT, and APRI) showed better AUC (0.877) and OR (1.52) scores. Our study indicates that laboratory tests such as WBC, PT, and APRI, helped identify patients at risk of developing severe dengue. The APRI was identified as a valuable predictor of patients with severe dengue. Combining the WBC, PT, and APRI scores allowed a better prediction of severe dengue.