A Bayesian Model to Predict Right Ventricular Failure Following Left Ventricular Assist Device Therapy.
A Bayesian Model to Predict Right Ventricular Failure Following Left Ventricular Assist Device Therapy.
复制标题
DOI：
10.1016/j.jchf.2016.04.004
复制
发表时间：
2016-09
影响因子：
13
通讯作者：
Antaki, James F.
中科院分区：
文献类型：
--
作者：
Loghmanpour, Natasha A.;Kormos, Robert L.;Kanwar, Manreet K.;Teuteberg, Jeffrey J.;Murali, Srinivas;Antaki, James F.
关键词：
右心室衰竭（RVF）仍然是左心室辅助装置（LVAD）植入后的主要不良事件。本研究探讨了贝叶斯统计模型的使用，以解决现有的多变量分析得出的风险评分的有限预测能力。这是基于一个假设，即有必要考虑自变量之间的相互关系和条件概率，以达到足够的统计精度。本研究使用的数据来自2006年12月至2014年3月INTERMACS的10,909名原发性LVAD成年患者。考虑了176个植入前变量的初始集。裂谷热植入后分为急性（<48小时）、早期（48小时- 14天）和晚期（bb0 - 14天）发病。对于每个端点，使用最具预测性的变量，使用开源贝叶斯推理引擎（SMILE）构建了一个单独的树增强Naïve贝叶斯模型。急性裂谷热模型由33个变量组成，包括：肺动脉收缩压（PAP）、白细胞计数、左心室射血分数、心脏指数、钠水平和淋巴细胞百分比。早期裂谷热模型包括34个变量，包括收缩期PAP、前白蛋白、LDH、INTERMACS谱、右心室射血分数、前b型利钠肽、年龄、心率、三尖瓣反流和BMI。裂谷热晚期模型包括33个变量，主要由外周血管阻力、MELD评分、白蛋白、淋巴细胞百分比、平均PAP和舒张期PAP预测。所有贝叶斯模型的准确率在91-97%之间，AUC在0.83-0.90之间，灵敏度为90%，特异性为98-99%，显著优于先前发表的风险评分。基于大型多中心INTERMACS登记的裂谷热贝叶斯预后模型基于术前变量提供了急性、早期和晚期裂谷热的高度准确预测。这些模型可以促进临床决策，同时筛选LVAD治疗的候选人。
Right ventricular failure (RVF) continues to be a major adverse event following left ventricular assist device (LVAD) implantation. This study investigates the use of a Bayesian statistical model to address the limited predictive capacity of existing risk scores derived from multivariate analyses. This is based on the hypothesis that it is necessary to consider the inter-relationships and conditional probabilities amongst independent variables to achieve sufficient statistical accuracy. The data used for this study was derived from 10,909 adult patients from INTERMACS who had a primary LVAD from December 2006 – March 2014. An initial set of 176 pre-implant variables were considered. RVF post-implant was categorized as acute (<48 hours), early (48 hours–14 days) and late (>14 days) in onset. For each of these endpoints, a separate tree-augmented Naïve Bayes model was constructed using the most predictive variables using an open source Bayesian inference engine (SMILE.) The acute RVF model consisted of 33 variables, including: systolic pulmonary artery pressure (PAP), white blood cell count, left ventricular ejection fraction, cardiac index, sodium levels, and lymphocyte percentage. The early RVF model consisted of 34 variables, including systolic PAP, pre-albumin, LDH, INTERMACS profile, right ventricular ejection fraction, pro-B-type natriuretic peptide, age, heart rate, tricuspid regurgitation and BMI. The late RVF model included 33 variables and was mostly predicted by peripheral vascular resistance, MELD score, albumin, lymphocyte percentage, mean PAP and diastolic PAP. The accuracies of all the Bayesian models were between 91–97%, AUC between 0.83–0.90 sensitivity of 90% and specificity between 98–99%, significantly outperforming previously published risk scores. A Bayesian prognostic model of RVF, based on the large, multi-center INTERMACS registry provided highly accurate predictions of acute, early, and late RVF based on preoperative variables. These models may facilitate clinical decision-making while screening candidates for LVAD therapy.
登录
查看更多内容
DOI：
10.1109/icmla.2010.50
发表时间：
2010-12-12
期刊：
Proceedings of the ... International Conference on Machine Learning and Applications. International Conference on Machine Learning and Applications
影响因子：
--
作者：
Wang Y;Rosé CP;Ferreira A;McNamara DM;Kormos RL;Antaki JF
通讯作者：
Antaki JF
DOI：
10.1016/j.jtcvs.2009.11.020
发表时间：
2010-05-01
影响因子：
6
作者：
Kormos, Robert L.;Teuteberg, Jeffrey J.;Farrar, David J.
通讯作者：
Farrar, David J.
影响因子：
24
作者：
Matthews, Jennifer Cowger;Koelling, Todd M.;Aaronson, Keith D.
通讯作者：
Aaronson, Keith D.
DOI：
10.1109/titb.2012.2187458
发表时间：
2012-05-01
影响因子：
--
作者：
Wang, Yajuan;Simon, Marc;Antaki, James F.
通讯作者：
Antaki, James F.
影响因子：
8.9
作者：
Dang, NC;Topkara, VK;Naka, Y
通讯作者：
Naka, Y