| [Objective]Deep Venous Thrombosis(DVT)refers to abnormal coagulation of blood in Deep veins,which is a disorder of Venous return of lower limbs.Patients with deep vein thrombosis have the possibility of thrombus shedding and pulmonary embolism resulting in death.Studies have shown that the number of patients with DVT increases with age.At present,there are many DVT risk assessment models in clinical use in China,but there is no unified DVT risk assessment model for elderly inpatients.Senior age is an independent risk factor for the occurrence of DVT,and elderly patients are the population with high incidence of DVT.Risk grade assessment is the basis and means of implementing preventive measures.Therefore,it is particularly important to strengthen the research on risk assessment of deep venous thrombosis in elderly inpatients.By comparing the predictive value of three most commonly used DVT risk assessment models(Caprini2010DVT risk assessment model,Padua DVT risk assessment model and Wells DVT risk assessment model)in the risk assessment of deep vein thrombosis in elderly inpatients,this study determined the most suitable DVT risk assessment model for elderly inpatients.To provide a more accurate assessment tool for the standardized prevention of deep venous thrombosis in elderly inpatients.[Methods]1.Sample and data collectionUsing simple random sampling method from HIS one big 3 armour hospital electronic medical record system from January 1,2019 to October 31,2020 in the hospital over 65(including 65),a total of 329 cases of patients,these cases were retrospectively results through intravenous colour to exceed 80 cases of patients diagnosed with DVT,who did not have 249 cases of deep vein thrombosis patients.General data collection including age,gender,primary diagnosis,merge,presence of chronic diseases,malignant tumor,laboratory indicators(red blood cell count,hemoglobin,blood clots elastic figure,homocysteine),the results of the veins B to exceed,history of blood clots,special treatment(chemotherapy,hormone therapy,anticoagulants,diabetes insulin treatment).The Caprini2010 DVT risk assessment model scale,Padua DVT risk assessment model scale and Wells DVT risk assessment model scale were used to evaluate the selected research subjects.2.Statistical analysis(1)Univariate analysis was carried out for 329 elderly patients with DVT risk assessment model by SPSS software,and univariate variance test was used for quantitative variables.Chi-square test was used to explore the significance of 17 factors in the degree of difference between the observation group and the control group.(2)The samples were divided into five age groups according to age: 65-70 years old,70-75 years old,75-80 years old,80-85 years old and above 85 years old.The evaluation results of Caprini,Wells and Padua evaluation models were analyzed.(3)Rank sum test was used to compare the distribution differences of risk assessment grades of the three models among different diseases.(4)The specialized biomedical software Med Calc was used to draw the Receiver Operating Characteristic ROC of the sample evaluation results,and the Area Under the ROC Curve AUC was calculated.Specific and sensitive,and Youden,S Index.Compared with SPSS software,Med Calc has many functions required by medical research,the most representative of which is the AUC comparison function.ROC was used to analyze and evaluate the effect of dichotomous discrimination.The statistical principle was to obtain data pairs of sensitivity and 1-specificity through the variation of cutoff point/cutoff value,with misjudgment rate as the horizontal axis.The sensitivity is the longitudinal axis,and the obtained data pair is plotted into a curve.AUC is the area under the ROC curve and is the most commonly used diagnostic accuracy measure globally.It is often used to summarize diagnostic capabilities from a single number and can also be used to compare the diagnostic quality of two or more diagnostic procedures.Yoden index is the maximum of sensitivity plus specificity minus 1.It aims to determine the cut-off point with the best test accuracy and is widely used to determine the optimal threshold.By drawing the ROC curve and calculating the parameters(AUC,sensitivity,specificity and Youden index),the prediction value of the three evaluation models was analyzed and evaluated by comparison.(5)The risk level assessed by the Wells Thrombosis Risk Model was taken as the dependent variable,and the general basic information of the two groups of patients and the statistically significant risk factors summarized in univariate analysis were taken as the independent variables for logistic regression analysis.(6)Quantitative assessment of the patient’s Activity of Daily Living ADL,using Barthel index assessment of the risk factors for DVT patients to score the ability to move,Youden index was calculated with Med Calc software to determine the optimal threshold value of daily living ability corresponding to DVT diagnosis.[Results]1.Distribution of risk levels of DVT in elderly patients of different ages(1)of this study,a total of 329 cases of elderly patients with DVT risk assessment model of risk factors for consideration on the single factor analysis of observation group and control group in gender,age,red blood cell count,hemoglobin,homocysteine,presence of malignant tumor,chemotherapy,hormone therapy,with or without the use of anticoagulant drugs,history of hypertension,lung disease history and history of heart disease,The differences were not statistically significant(all P > 0.05).There were statistically significant differences in gender,D-dimer,activity,thrombosis history,insulin use and diabetes history between the two groups(all P < 0.05).(2)The samples are divided into five age groups: 65-70 years old,70-75 years old,75-80 years old,80-85 years old,and over 85 years old.The results of Caprini,Wells and Padua evaluation model were analyzed.The scores of Caprini evaluation model in five age groups were(3.52±0.46),(4.13±0.5),(5.11±1.01),(5.09±1.44)and(5.85±0.52),respectively.The scores of Wells evaluation model in five age groups were(0.29±0.13),(0.57±0.17),(0.96±0.4),(1.00±0.34),(1.22±0.2),respectively.The scores of Padua assessment model in five age groups were(1.00±0.41),(1.95±0.42),(2.74±0.78),(3.32±0.87),and(3.72±0.53).According to the proportion of risk levels of the three models in different age ranges of the samples,it can be seen that with the increase of age,The proportion of high-risk patients showed an increasing trend,and the chi-square test showed that the proportion of risk assessment results of the three models in each age group had statistical significance(P < 0.001).2.Comparison of DVT risk among elderly patients with different underlying diseases The observation group was classified into hypertension,heart disease,diabetes,malignant tumor,cerebral infarction,pulmonary disease,Alzheimer’s disease,urinary system disease,digestive system disease and other diseases according to the disease diagnosis.In this part,rank sum test was used to compare the distribution differences of risk assessment grades of the three models among different diseases.The chi-square values(H)of the three models were 30.13,22.82 and 22.26,respectively,and the differences were all less than 0.05.Therefore,in general,the distribution difference of risk grade of elderly patients with DVT among different diseases was statistically significant(P < 0.05).3.ROC curves of Caprini,Wells and Padua thrombus assessment models were compared and analyzed(1)The AUC of Caprini,Wells and Padua thrombus risk assessment models were(0.718±0.079),(0.827±0.067)and(0.714±0.083),respectively.The comparison of AUC of the three models was P < 0.05.The differences were statistically significant,indicating that the three models were all significant for the prediction of DVT.(2)Youden’s index was used to determine the optimal threshold of the three models.The optimal threshold of Caprini model was 6.5(sensitivity 48.8%,specificity 80.8%,Youden index 0.295).The optimal cut-off value of Wells model is 1.5(sensitivity is 70.0%,specificity is 82.7 %,Youden index is 0.572).The optimal cut-off value for Padua model was 1.5(sensitivity 82.5%,specificity 62.8%,Youden index 0.453).(3)Pair comparison analysis of the three evaluation models showed that the AUC of Caprini evaluation model and Padua evaluation model was compared(Z=1.563,P=0.1181),and that of Wells evaluation model and Caprini evaluation model was compared(Z=6.614,P <0.0001).The AUC of Wells model was compared with that of Padua model(Z=4.548,P<0.0001),and the AUC of Wells model was compared with that of Caprini and Padua model.4.Logistic regression analysis of thrombus risk factors in elderly patients Wells thrombosis risk model to assess risk levels as the dependent variable(in patients at high risk of low-risk patients Y = 0,Y = 1),with two groups of patients with significant factors in the general clinical data in the part of the single factor analysis and summary of statistically significant risk factors as the independent variable in logistic regression analysis,the results show that the diagnosis of diseases,daily life ability,D-dimer was the main risk factor of Wells model for thrombosis assessment,with a high correlation(P <0.05).5.Critical value analysis of daily living ability The ROC analysis of Med Calc software was used to determine the optimal cut-off value of daily living ability corresponding to the prediction of DVT.The cut-off value was divided according to the Youden index.The optimal cut-off value of daily living ability was 85 points,and the Youden index was 0.3658.By drawing the line chart of Youden index and activity ability,it can be intuitively seen that when the score of daily living ability of patients is 85 points,the Youden index reaches the highest value,indicating that this threshold value has the best DVT screening ability for patients.[Conclusions]1.Wells model has the highest prediction efficiency for DVT risk in elderly inpatients and is most suitable for DVT risk assessment in elderly inpatients.2.The disease diagnosis,daily living ability and D-dimer level of elderly inpatients were highly correlated with Wells thrombosis assessment model.3.When the score of daily living ability of patients was 85 points,the Youden index reached the highest value,and this threshold value had the best DVT screening ability for patients.Patients were at increased risk of DVT when their daily living ability score was less than 85. |
PDF Full Text Request