Development of a model for predicting gastrointestinal bleeding in patients with ischemic stroke after dual antiplatelet therapy

Objective: To establish a model for predicting gastrointestinal bleeding in patients with ischemic stroke after dual antiplatelet therapy (DAPT). Methods: A model for predicting gastrointestinal bleeding in patients with ischemic stroke after DAPT was established based on a retrospective study that involved 1,217 patients diagnosed with ischemic stroke in the Neurology Department of Nanchang University Affiliated Ganzhou Hospital from January 2019 to June 2021. A receiver operating characteristic curve was constructed to evaluate the model's power. Data from patients with ischemic stroke between July and December 2021 were used to validate the power of the model. Results: A total of 1,217 patients with ischemic stroke between January 2019 and June 2021 were included in the model. The cohort comprised 1,164 patients in the non-gastrointestinal bleeding group and 53 in the gastrointestinal bleeding group. Multivariate logistic regression analysis revealed that age, fibrinogen level, neutrophil-to-lymphocyte ratio, and National Institute of Health Stroke Scale score were independent risk factors for gastrointestinal bleeding. A model for predicting gastrointestinal bleeding in patients with ischemic stroke after DAPT was established, Logit(P) = -7.269 + 0.074 ×1 + 0.071 ×2 + 0.361 ×3 + 0.082 ×4 (X1, National Institute of Health Stroke Scale score; X2, neutrophil-to-lymphocyte ratio; X3, fibrinogen; X4, activated partial thromboplastin time). Receiver operating characteristic analysis showed that the area under the curve for the model was 0.733. Data from the validation group showed that the area under the curve for the model was 0.665. Conclusion: A model for predicting gastrointestinal bleeding in patients with ischemic stroke after DAPT was established and demonstrated its predictive ability. Although the predictive ability of the model was not perfect, this was an important attempt. Further studies are needed to establish better models to predict gastrointestinal bleeding.