Value Of Daniel Score Combined With D-dimer And Revised Geneva Rule In Predicting Acute Pulmonary Embolism

Objective: Assess the predictive value for acute pulmonary embolism by combining Daniel score with D-dimer and revised Geneva score.Construct a new prediction model and verify the if the predictor could offer a better predicting efficiency.Methods: Retrospectively collect the case of patients admitted to emergency room or wards from The Second Clinical College of Dalian Medical University,suspected as acute pulmonary embolism(APE)with the symptoms including chest pain,dyspnea,syncope or haemoptysis and were diagnosed by CT pulmonary angiography(CTPA).The patients were divided into APE group and APE control group while APE group was divided into low-probability category,moderate-probability category,highprobability category according to the revised Geneva rule criteria formulated by the Chinese Medical Association in the Guidelines for Diagnosis,Treatment and Prevention of Pulmonary Thromboembolism published in 2018.Comparison results of the general clinical data,laboratory test and previous medical history between experiment group and the control group were recorded.ROC curve was drawn to compare whether the predictive efficacy of combining Daniel score,D-dimer and revised Geneva rule is higher than singly combining revised Geneva rule with D-dimer in diagnosing APE in three groups.New predictors were established by logistic regression analysis combining statistically significant items with high correlation in the low-probability,moderate-probability and high-probability category.Compare the area under ROC curve(AUC)of new predictor with each separated indicator and Ddimer combined with Daniel score,calculate the optimal cut-off value,sensitivity,specificity,positive predictive value and negative predictive value.Results: We enrolled 293 subjects including 133 males(45.4%)and 160 females(54.6%),with an average age of 65.0(56.0-73.0)years in this study.There were 158 cases in APE group(81 in low-probability category,64 in moderate-probability category,13 in high-probability category)and 135 cases in APE control group.Incidence of APE as the dichotomous,it was found that height,weight,heart rate,BNP,autoimmune diseases,diabetes,tumor,hyperlipidemia were statistically correlated with the incidence of APE(P < 0.05).The predictive value of APE by combining Daniel score,D-dimer and revised Geneva rule was higher than that by combinig revised Geneva rule with D-dimer in low-probability category(AUC: 0.868,95%CI: 0.804-0.932,sensitivity: 88.9%,specificity: 72.5%),moderate-probability category(AUC:0.740,95%CI: 0.652-0.828,sensitivity: 73.4%,specificity: 73.8%)and high-probability category(AUC: 0.876,95%CI: 0.735-1.000,sensitivity: 92.3%,specificity:79.3%).In the low-probability category,the T-wave inversion in leads V1-V4,SⅠ,TⅢ were statistically correlated with predicting APE and were used to establish the predictor 1.The AUC of predictor 1 was higher than that of each separated indicator and D-dimer combined with Daniel score.(AUC: 0.929,cut-off: 0.011,95%CI:0.866-0.971,sensitivity: 80.2%,specificity: 92.2%).In the moderate-probability category,the Twave inversion in lead V1-V4,RBBB,SⅠand TⅢ were used to establish the predictive factor 2.The AUC of predictor 2 was higher than that of each separated indicator and D-dimer combined with Daniel score(AUC: 0.830,cut-off: 1.07,95%CI:0.756-0.903,sensitivity: 65.6%,specificity: 88.5%).In the high-probability category,none of the factors were statistically correlated with predicting APE so that the predictor could not be established.Conclusions: Daniel score combined with D-dimer and revised Geneva rule have better predictive value for APE than D-dimer combined with revised Geneva rule.In the low-probability category and moderate-probability category,the occurence of SⅠ,TⅢ and the T-wave inversion in leads V1-V4 in ECG were significantly correlated with the occurrence of APE.