Application Value Of Risk Assessment Models For Venous Thromboembolism In Patients In The Department Of Internal Medicine

Objective: To study the effectiveness of Caprini risk assessment model and Padua risk assessment model in predicting venous thromboembolism in medical inpatients.Methods: Data of internal medicine inpatients in affiliated hospital of Qingdao university from August to December 2017 were retrospectively collected,and 3268 inpatients who met the study requirements were screened out for retrospective study,among which 172 patients were diagnosed with VTE.Caprini and Padua risk assessment models were used in all patients for thrombosis risk score and VTE risk grading,and multivariate Logistic regression model was used to analyze the risk factors of VTE in internal medicine inpatients with VTE.The relationship between Caprini and Padua risk assessment models and the risk of VTE occurrence in different risk stratification was compared.The value of the two models in predicting the risk of VTE was evaluated by using the subject operating characteristic curve(ROC)and calculating the area under the curve.Meanwhile,the Logistic binary regression analysis was applied to obtain the prediction probability model after the combination of Caprini and Padua,and the ROC characteristic curve was drawn.The area under the curve(AUC)of the three models was compared by Delong method.Results: VTE incidence in internal medicine inpatients was 5.26%,among which 38.98% of patients in Caprini risk assessment model were rated as high risk or very high risk.Patients in Caprini risk assessment model VTE group had significantly higher scores than patients in non-vte group [6(1-15)vs 4(0-15),P < 0.05].Compared with patients in the low-risk group,patients in the high-risk and extremely high-risk groups had a 6.948-fold and 8.333-fold risk of developing VTE,respectively.At the same time,37.82% of the patients were high-risk patients assessed by Padua risk assessment model,and the patients in the VTE group had significantly higher scores than those in the non-vte group [5(1 ~ 13)vs.3(0 ~ 13),P <0.05].The risk of VTE in high-risk patients was 11.380 times higher than in low-risk patients.Exist in hospitalized patients with risk factors as the independent variable of multi-factor analysis showed that age(≥70 years),BMI > 25 kg/m2,brake(> 3 d),active malignant tumor,cerebral apoplexy,central venipuncture,recent surgery or trauma(< 1month),diabetes,D-dimer,past history of VTE and lower limb swelling and the occurrence of VTE was statistically significant(P < 0.05).VTE is most closely related to recent surgery OR trauma(OR=5.418),followed by active malignant tumor(OR=4.416).Caprini model had a sensitivity of 79.1% and a specificity of 66.7% when Youden index was 0.458.The Youden index of Padua model was 0.548,with a sensitivity of 72.1% and a specificity of 82.7%.Thesensitivity and specificity of the prediction probability model were 83.7% and 87.9% when Youden index was 0.716.The area under the curve of Padua model was higher than that of Caprini model [0.817±0.017 VS 0.739±0.020,Z = 9.433,P < 0.05],and the area under the curve of the prediction probability model constructed by the two models was higher than that of the two models(P < 0.05).Conclusion: 1.Both Caprini and Padua risk assessment models can predict the risk of VTE in medical inpatients.2.In internal medicine inpatients,the predictive value of Padua risk assessment model is superior to Caprini.The combined use of Caprini and Padua risk assessment model can improve the accuracy of screening VTE in internal medicine inpatients.