OBJECTIVETo explore the clinical value of immature platelet fraction (IPF), absolute immature platelet fraction (A- IPF) and thrombelastograph (TEG) on assessment of bleeding risk of immune thrombocytopenia (ITP).

METHODStwo hundred and seventy- one patients with ITP were assessed based on ITP-BAT bleeding grading system. IPF, A-IPF were determined in 271 patients ,TEG in 125 patients. The correlations between bleeding grades and IPF, A-IPF, variables of TEG in subgroups were analyzed by statistical method. The predictive value of IPF, A-IPF, and variables of TEG on bleeding risk of ITP patients was evaluated.

RESULTSThere were no significant differences in bleeding degree in all patients with different gender and disease stage (P>0.05). Mild bleeding rate in children was higher than that in adult (P<0.05). PLT inversely correlated with bleeding grade for the entire cohort (P<0.001). In all subjects, PLT< 30 × 10⁹/L and pediatric cohorts with PLT< 30 × 10⁹/L, PLT were negatively correlated with IPF (P<0.05), positive correlated with A-IPF (P<0.001) and the maximum amplitude (MA (P<0.05). Bleeding grades were significantly correlated with IPF, A-IPF, MA in all subjects and patients with PLT< 30 × 10⁹/L (P<0.001). IPF, A-IPF and MA did not correlate with bleeding grades in children with PLT< 30 × 10⁹/L (P>0.05). ROC curve analysis revealed IPF, A-IPF and MA had better predictive value (AUC 0.745, 0.744, 0.813, P<0.001). Multivariate analysis showed that IPF and MA were independence factors for predicting bleeding risk in ITP patients and comprehensive predictive value was higher (AUC 0.846, P<0.001) than single variable.

CONCLUSIONIPF, A-IPF and MA could accurately evaluate bleeding risk in ITP patients. It may be considered as reference index of the treatment and observation index of curative effect.