A new score for predicting intracranial hemorrhage in patients using antiplatelet drugs

Fuxin Ma1,2, Zhiwei Zeng2, Jiana Chen2

  • 1School of Pharmacy, Fujian Medical University, Fuzhou, China.

Annals of Hematology
|April 17, 2024
PubMed

Insights

A new predictive model identifies nine key factors to assess the risk of intracranial hemorrhage (ICH) in patients taking antiplatelet drugs. This tool aims to help clinicians reduce ICH occurrence and improve patient outcomes.

Area of Science:

  • Neurology
  • Pharmacology
  • Medical Informatics

Background:

  • Antiplatelet drugs are crucial for preventing thrombotic events but increase the risk of intracranial hemorrhage (ICH).
  • Currently, no specific clinical score exists to predict ICH risk in patients using antiplatelet therapy.
  • ICH poses a significant threat to patient quality of life and survival.

Purpose of the Study:

  • To identify independent risk factors associated with ICH in patients treated with antiplatelet drugs.
  • To develop and validate a predictive model for ICH in this patient population.
  • To provide a validated clinical tool for early risk assessment and prevention of ICH.

Main Methods:

  • A retrospective, single-center study involving 753 patients on antiplatelet drugs.
  • Logistic regression was used to build the predictive model.
  • Internal validation employed AUC, calibration curves, and the Hosmer-Lemeshow test.

Main Results:

  • Nine factors were identified as independent risk factors for ICH: male sex, headache/vomiting, hypertension, cerebrovascular disease, white matter hypodensity, abnormal GCS, elevated fibrinogen, and D-dimer.
  • Lipid-lowering drugs emerged as a protective factor.
  • The predictive model demonstrated strong discriminatory power (AUC=0.949 in development, 0.943 in validation) and good calibration.

Conclusions:

  • A validated predictive model incorporating nine factors has been developed for assessing ICH risk in patients on antiplatelet drugs.
  • This model shows excellent predictive value and may serve as an effective clinical tool for reducing ICH incidence.
  • Further application of this model can aid in personalized risk management and prevention strategies for ICH.