A Predictive Nomogram and Risk Stratification for Repeat Platelet Transfusion in Thrombocytopenia: A Retrospective Study of 290 Cases
Description
The dataset contains de-identified clinical and laboratory data from a retrospective study of 290 hospitalized thrombocytopenic patients who received at least one platelet (PLT) transfusion between April 2020 and December 2025 at our institution. Among them, 145 patients required repeat PLT transfusions (≥2 transfusions), while 145 patients received only a single transfusion. The dataset includes patient demographics (age, sex, BMI, smoking and alcohol history), comorbidities (e.g., hypertension, diabetes, cardiovascular and renal diseases), primary disease types (hematologic malignancies, bone marrow failure syndromes, solid tumors, liver disease, immune thrombocytopenia, and infectious diseases), clinical parameters (PLT count, hemoglobin, white blood cell count, coagulation profile, albumin), bleeding severity, splenomegaly, fever and infection status, prior use of thrombopoietic agents, type and timing of transfusion, and outcome of repeat transfusion. This dataset was used to identify independent risk factors for repeat PLT transfusion, develop a predictive nomogram, and construct a simplified scoring system for risk stratification. All data have been de-identified according to the Safe Harbor Method to protect patient privacy.