Analysis of a large data set to identify predictors of blood transfusion in primary total hip and knee arthroplasty

• DOI: 10.1111/trf.14783
• Publication Date: 2018-08-26T06:29:16Z
• Authors: ZeYu Huang; Cheng Huang; JinWei Xie; Jun Ma; GuoRui Cao; Qiang Huang; Bin Shen; Virginia Byers Kraus; FuXing Pei
• Keywords: Joint arthroplasty

The aim of this study was to identify the predictors need for allogenic blood transfusion (ALBT) in primary lower limb total joint arthroplasty (TJA).This utilized a large dataset 15,187 patients undergoing unilateral TJA. Risk factors and demographic information were extracted from electronic health record. A predictive model developed by both random forest (RF) algorithm logistic regression (LR). area under receiver operating characteristic curve (AUC-ROC) used compare accuracy two methods.The rate ALBT 18.9% total. Patient-related associated with higher risk an included female sex, American Society Anesthesiologists (ASA) II, ASA III, IV. Surgery-related operative time, drain use, amount intraoperative loss. Higher preoperative hemoglobin tranexamic acid use decreased ALBT. RF had better (area [AUC] 0.84) than LR (AUC, 0.77; p < 0.001).The identified current can provide specific, personalized perioperative assessment patient considering Furthermore, significantly that LR, making it potential tool future prediction