Epidemiology Of Peroperative Venous Thromboembolism And The Development And Validation Of Its Scoring System

1. Objective:(1) To explore the epidemiology of peroperative venousthromboembolism(VTE)；(2) To develop a risk scoring system of peroperative VTE；(3)To validate the risk scoring system of peroperative VTE.2. Methods:(1) We consecutively screened the hospitalized patients of peroperativeVTE in Chinese PLA general hospital within February2004~February2012, wecalculated the prevalence and analyzed the demographics and clinical characteristicincluding manifestation, thrombus position, diagnosis method and laboratory index, etal.(2) We retrospectively reviewed150consecutively collected peroperative VTEpatients mentioned in the first part, and chose150non peroperative VTE matchedpatients in the same period according the following principle: the surgery disease, thesurgery approach, and the underlying diseases except the risk factors of thrombosismatched to the peroperative VTE patients. These patients randomly divided into twogroups(model and validation group):80peroperative VTE patients and the80matchednon peroperative VTE patients were the model group; the other70VTE patients and the70matched non peroperative VTE patients were the validation group. We screened therisk factors of peroperative VTE by univariate and multivariate analysis and assignedpoint values to each independent risk factor according to the odds ratio, then the riskindex values were divided into discrete ranges.(3) The authenticity and reliability of thescoring system was validated in the model and validation group using receiver operatorcharacteristic curve(ROC) and Hosmer-Lemeshow goodness-of-fit.3. Results:(1) VTE occurred in150of256902(0.584‰) undergoing surgery inChinese PLA general hospital from February2004to February2012. The prevalencewas2.280‰in the department of Vascular, which was highest. The department ofThoracic and Orthopedic were lower than vascular department, the prevalence was0.809‰,0.775‰, respectively. The prevalence of the department of Otolaryngologywas0.197‰, which was the lowest.(2) The relevant risk factors by univariate analysiswere age≥65years, BMI≥26kg/m2, haematocrit(HCT)≤38%, platelet(PLT)<150×109/L, albumin≤3.5mg/dl, emergency, operative time, intraoperative transfusion≥4U RBCs, etal.(3) The independent risk factors by logistic multivariate analysis were age≥65years（OR：2.505，95%CI：1.113-5.638，P=0.026）, BMI≥26kg/m2（OR：2.303，95%CI：1.026-5.170，P=0.043）, albumin≤3.5mg/dL（OR：7.371，95%CI：2.562-21.204，P<0.001）, emergency（OR：4.748，95%CI：1.699-13.272，P=0.003）, intraoperativetransfusion≥4U（OR：4.525，95%CI：1.526-13.420，P=0.006）, et al.(4) The point valueswere assigned to each independent risk factor. Age≥65years added3points to the finalscore, BMI≥26kg/m2added2points, albumin≤3.5mg/dL added7points, emergencyadded5points, intraoperative transfusion≥4U added5points.(5) The risk level: scorerange<4points was low, score range4-9points was medium, score range>9points washigh.(6) In model group, the area under curve(AUC) of the peroperative VTE scoringsystem was0.846by ROC, the sensitivity and specificity was73.8%,78.7%,respectively(P<0.001); In validation group, AUC was0.703, the sensitivity andspecificity was80.0%,52.9%, respectively(P<0.001).(7) In model and validation group,we found that the VTE expected value and observed value had no significant differenceby Hosmer-Lemeshow goodness-of-fit(model group: Chi-square=1.935, the criticalvalue of chi-square=14.067, P>0.05; validation group: Chi-square=6.774, the criticalvalue of chi-square=9.488, P>0.05).4. Conclusions:(1) Peroperative VTE is common in the department of Vascular,Thoracic and Orthopedic, while the department of Otolaryngology was less uncommon.(2) The independent risk factors of peroperative VTE are age≥65years, BMI≥26kg/m2,albumin≤3.5mg/dl, emergency, intraoperative transfusion≥4U RBCs.(3)The scoringsystem based on the risk factors helps to identify patients at high risk for peroperativeVTE.