Hemostatic phenotypes in dengue: integrating platelet kinetics and coagulation dysfunction across the spectrum of disseminated intravascular coagulation

Dengue-associated disseminated intravascular coagulation (DIC) represents a severe coagulation complication characterized by coagulation activation and thrombocytopenia. However, the relationship between platelet kinetic phenotypes and concurrent ISTH-defined coagulation status remains incompletely characterized. This study evaluated the interaction between platelet indices, thrombocytopenia phenotypes, and coagulation dysfunction to explore an integrated platelet–coagulation phenotyping framework for dengue-associated DIC. This retrospective cross-sectional study included 334 patients with laboratory-confirmed dengue. Concurrent laboratory findings were classified by the ISTH 2001 overt DIC score (< 5 versus ≥ 5). Four exploratory platelet kinetic phenotypes and three composite hemostatic indices, including the Thrombo-Platelet Consumption Ratio (TPCR), Platelet-Coagulation Imbalance Index (PCI-I), and Platelet-Coagulation Dissociation Index (PCDI), were evaluated using group comparisons, receiver operating characteristic analyses, and exploratory univariable regression. Of 334 patients, 309 (92.5%) had ISTH scores < 5, whereas 25 (7.5%) met the overt DIC threshold. Phenotypes were inadequate response (73, 21.9%), hyperdestructive (167, 50.0%), hypoproductive (37, 11.1%), and lazy-suppressed (57, 17.1%). Overt DIC occurred in 3/73 (4.1%), 21/167 (12.6%), 0/37 (0.0%), and 1/57 (1.8%), respectively (Fisher–Freeman–Halton exact p = 0.0043). The apparent AUCs were 0.941 for D-dimer, 0.953 for TPCR, 0.960 for PCI-I, and 0.945 for PCDI. However, PCI-I demonstrated the highest AUC (0.960); its discriminatory performance was not statistically superior to the other biomarkers after adjustment for multiple comparisons. Exploratory platelet kinetic phenotypes were associated with concurrent ISTH-defined overt DIC status. Composite indices showed high apparent discrimination but share components with the reference score, creating incorporation bias. These findings do not establish prediction of clinical outcomes and require independent validation.