Development And Validation Of PICC Related Upper Extremity Venous Thrombosis Scoring Model

ObjectiveTo explore the risk factors of peripherally inserted central catheters(PICCs)related upper limb vein thrombosis,develop and validate a new scoring model to evaluate the patients' risk level,thus assisting the guidance of clinical assessment and prevention about PICC related thrombosis.Meanwhile,for conventional ultrasound and imaging technology with difficulty to make a precise diagnosis but highly suspected correlation thrombosis patients,implementation of contrast-enhanced ultrasound,to explore the application value of contrast-enhanced ultrasound in the diagnosis of PICC related thrombosis.MethodsThe patients who are fit in with the inclusion and exclusion criteria were enrolled in the study from April 2015 to January 2017 at a hospital in Tianjin,at the same time,divided these patients into training samples and testing samples according to the chronological order.The every research object was carried out for a month follow-up,according to the requirements of PICC catheters' maintenance,received the routine ultrasound examination in every 7 days before change the dressings.The data collected included general data of patients,puncture process related information,physical illness and medical history information,laboratory results and the clinical situation about the limb.The diagnosis of upper extremity venous thrombosis was performed by routine Doppler ultrasound.In order to clarify the status of thrombosis,the patients who was difficult to confirm the diagnosis,implementation of the contrast-enhanced ultrasound examination.On the basis of venous thrombosis,the research objects of training samples and testing samples were divided into non thrombosis group and thrombosis group,then used the SPSS18.0 software to analyze and process the data.For the sake of finding out the statistically significant differences,the variables of the training samples was analyzed by c 2 test,t-test and Mann-Whitney U test.After that,using the binary logistic regression analysis to filter the independent risk factors about PICC related upper limb vein thrombosis and develop the PICC related upper extremity venousthrombosis scoring model.Furthermore,in the light of the odds ratios to determine the value of each factor,at the same time,classified the risk level of the scoring system.The scoring model was evaluated according to the sensitivity,specificity and ROC curve in the training samples and the validation samples.Finally,the image data of contrast-enhanced ultrasound was analyzed to preliminary summarize its application value in patients with PICC related thrombosis.Results1.A total of 375 subjects were enrolled in the study,training sample group of 275 cases,the validation sample group of 100 cases,a total of 72 cases of thrombosis(training group of 53 cases,validation group of 19 cases),thrombosis of the total incidence rate was 19.20%.2.Univariate analysis showed that the PICC related thrombosis' relevant factors including obesity,disease diagnosis,thrombosis history,chemotherapy history,main infusion contents,catheter tip position,CVC or PICC catheter history,diabetes history,activity(KPS score)and D-dimer.3.Logistic regression analysis showed that the independent risk factors of PICC related thrombosis including whether obesity(OR=3.449,95%CI:1.634-7.493,P=0.001),diabetes mellitus history(OR=3.016,95%CI:1.342-6.775,P=0.008),the history of thrombosis(OR=2.887,95%CI:1.256-6.632,P=0.012),history of chemotherapy(OR=2.268,95%CI:1.070-4.808,P=0.033)and D-dimer(OR=2.283,95%CI:1.132-4.604,P=0.021).4.According to the OR value(rounding-off method)to determine the value of the integral model of each independent factor,respectively obesity(BMI ? 25kg/m2)3 points,diabetes history 3 points,thrombosis history 3 points,chemotherapy history2 points,D-dimer(?0.5mg/L)2 points.The scoring model was as following :Y=obesity(BMI ? 25kg/m2)×3+ diabetes history×3+ the history of thrombosis×3+chemotherapy history×2+D-dimer(? 0.5mg/L)×2-3.19.5.By calculating the 95% confidence interval of the thrombus group and the non thrombus group in the training sample,getting the risk grade of the risk about thrombosis as follows: <3 points is divided into low risk,3-5 points is divided into middle risk and >5 points is divided into high risk.6.ROC curve analysis: according to the ROC curve analysis of the PICC related thrombosis scoring model in the training sample,the area under the curve was 0.753,sensitivity was 0.632,specificity was: 0.797,the cutoff point was 4.5.In the validation sample,the area under the curve was 0.758,sensitivity was 0.784,specificity was: 0.691,the cutoff point also 4.5.The prediction effect of the model is in the middle level.7.The effect and value of PICC catheters' contrast enhanced ultrasound: the contrast enhanced ultrasound can make the accurate diagnosis about PICC catheter-related thrombosis suspects in the states of color overflow,no obvious blood flow signal in low flow rate and severe clinical symptoms.Conclusions1.Obesity,diabetes history,thrombosis history,chemotherapy history and Ddimer(more than 0.5mg/L)were the independent risk factors of PICC related thrombosis.2.The PICC related upper extremity venous thrombosis scoring model can well predict the patients' risk level,it is simple and reliable with high sensitivity and specificity.2.The application of contrast-enhanced ultrasound in the accurate diagnosis of PICC related thrombosis is safe and reliable,it is worthy of clinical application.