Table_1_A Simple Clinical Pre-procedure Risk Model for Predicting Thrombocytopenia Associated With Periprocedural Use of Tirofiban in Patients Undergoing Percutaneous Coronary Intervention.docx

Background: No risk model for predicting thrombocytopenia associated with periprocedural tirofiban exposure is available. The purpose of this study was to develop a simple clinical pre-procedure risk model based on pre-procedural characteristics for early prediction of thrombocytopenia before patients were exposed to tirofiban.
Methods: The series included 1862 patients who underwent percutaneous coronary intervention with tirofiban exposure. Baseline demographic and clinical characteristics were collected from the hospital information system on admission. The earliest pro-procedural platelets within 72 h were used to evaluate the thrombocytopenia incidence. Risk factors associated with thrombocytopenia in patients with tirofiban exposure were investigated by univariable and multivariable analyses. Locally weighted scatterplot smoothing procedure was used to identify the cut points for the numeric variables. The discriminatory power of the scoring system was assessed with the receiver operating characteristic (ROC) curve analysis.
Results: The occurrence of thrombocytopenia was 4.02% (75 of 1862), 4.01% (56 of 1396), and 4.08% (19 of 466) in the overall, developmental, and validation data sets, respectively. The risk score was developed based on five independent predictors: age ≥65y, white blood cell ≥12 × 109/L, diabetes mellitus, congestive heart failure, and chronic kidney disease. This tool was well calibrated (Hosmer Lemeshow χ2 = 6.914; P = 0.546) and good discrimination was well obtained in validation data set (C-statistic, 0.82).
Conclusion: The clinical pre-procedure risk model is a simple and accurate tool for early identification of high-risk patients of thrombocytopenia before tirofiban exposure, allowing for timely and appropriate intervention.

【背景】目前尚无用于预测围手术期暴露于替罗非班相关血小板减少症的风险模型。本研究旨在基于患者术前临床特征，构建一款简便的术前风险模型，以在患者接受替罗非班给药前，早期预测其血小板减少症发生风险。
【方法】本研究队列共纳入1862例行经皮冠状动脉介入治疗且暴露于替罗非班的患者。于患者入院时通过医院信息系统采集其基线人口学与临床特征数据。以入院前72小时内的首次术前血小板计数作为评估血小板减少症发生率的依据。通过单变量及多变量分析，探究替罗非班暴露患者发生血小板减少症的相关危险因素。采用局部加权散点图平滑法（Locally weighted scatterplot smoothing）确定数值型变量的截断值。采用受试者工作特征（Receiver Operating Characteristic, ROC）曲线分析评估该评分系统的区分效能。
【结果】整体队列、建模队列及验证队列的血小板减少症发生率分别为4.02%（1862例中75例）、4.01%（1396例中56例）及4.08%（466例中19例）。该风险评分基于5项独立预测因子构建：年龄≥65岁、白细胞计数≥12×10^9/L、糖尿病、充血性心力衰竭及慢性肾脏病。该模型校准度良好（霍斯默-莱梅肖检验χ²=6.914，P=0.546），且在验证队列中展现出优异的区分效能（C统计量为0.82）。
【结论】这款术前临床风险模型是一款简便且精准的工具，可在患者接受替罗非班给药前，早期识别血小板减少症高风险人群，从而为及时开展恰当的干预措施提供依据。