Abstract: 【Abstract】 Objective To establish a nomogram to predict prolonged drainage time of postoperative cardiac, and get better clinical management. Methods Data on 152 consecutive patients who underwent open cardiac surgery in Sun Yat-sen memorial hospital of Sun Yat-sen University between January 2014, and January 2016. Data collect include the patient's general information, past medical history, perioperative related information and postoperative drainage time etc. Logistic regression analysis was used to identify the independent risk factors associated with postoperative drainage time that then were incorporated into the nomogram. Results Univariate logistic regression found that gender, CPB time （cardiopulmonary bypass time）, smoking, white blood cell count, creatinine preoperatively, left atrial diameter, Left ventricular ejection fraction （EF） significant influence cardiac postoperative drainage time；Multivariate logistic regression showed that smoking history, preoperative creatinine, white blood cell count, CPB time, and PT-INR（Prothrombin time-International Normalized Ratio） were an independent prognostic factor, and were incorporated into nomogram. Nomogram initial concordance index （C-idex） 0.78. After 1 000 times of internal model validation, C-idex was 0.76. The sensitivity of nomogram prediction was 80.0% （95% CI, 95%-69.2%）, and the area under the curve of the ROC analysis was 0.78（95% CI，0.74-0.82）. Positive odds ratio（PLR）was 2.43，negative odds ratio（NLR）of 0.30. Conclusion Cardiac postoperative drainage time related to many factors, the prediction modelbased on the related factors can predict postoperative drainage time accurately.

Key words: Drainage time, Nomogram, Cardiac surgery

摘要： 【摘要】目的建立可预测心脏术后患者引流时间延长的列线图，便于进行更好的临床管理。方法对中山大学附属孙逸仙纪念医院2014年1月至2016年1月期间152例行开胸心脏手术的病人进行回顾性分析，收集患者的一般资料、既往病史、围手术期相关情况和术后引流时间等资料。通过Logistic回归法分析并筛选术后引流时间的显著影响因素，建立预测术后引流时间延长的列线图。结果单因素分析显示性别、体外循环转机时间、升主动脉阻断时间、吸烟、疾病类型、凝血酶原国际标准比值（PT-INR）、白细胞计数、谷草转氨酶（AST）、谷丙转氨酶（ALT）、术前肌酐、N端脑那肽前体（N-proBNP）、左房内径、左室收缩功能与心脏术后引流时间具有相关性。多因素Logistic回归分析中吸烟史、术前肌酐、白细胞计数、体外循环转机时间、凝血酶原国际比值是独立预后因素，并用于绘制了便于临床使用的列线图。列线图初始的一致性指数（C-idex）为0.78。经过1000次的模型内部验证，并进行矫正，C-idex为0.76。列线图模型的敏感度为80.0%（95%置信区间69.2%~88.4%），其ROC分析的曲线下面积为0.78（95%置信区间0.74~0.82）。阳性比值比（PLR）为2.43，阴性比值比（NLR）为0.30。结论心脏术后引流时间与多种因素相关，基于相关影响因素建立的预测模型能较为准确预测术后引流时间延长的风险。

关键词: 引流时间, 列线图, 心脏手术