The Predictive Value Of Wells And Revised Geneva Score And D-dimer For The Diagnosis Of Acute Pulmonary Embolism In Critical Care Unit

BackgroundAcute pulmonary embolism has non-specific clinical manifestations with a widerange from no symptom to cardiogenic shock or sudden death. Moreover, themortality of un-treated patients with pulmonary embolism can be as high as30%.Therefore, the diagnosis and treatment as early as possible are of great significance.Currently the rules which are used to assess the clinical probability of APTE areGeneva and revised Geneva score,Wells score, simplified revised Geneva score andsimplified Wells score,moreover,the Wells and revised Geneva score are most widelyused.There are many researches and guidelines for the diagnosis of acute pulmonaryembolism for the suspected patients which suggest that assessing the probability withclinic prediction rules should be done prior to further examinations. Low possibilityof a clinical decision rule combined with negative D-dimer can safely excludepulmonary embolism, and30%of the suspected patients avoid further imagingexaminations. The patients in the Intensive Care Unit are in a critical situation andmost of them have multi-system complications. Thus the transformation andexaminations for this special crowd are limited and the diagnosis rate for acutepulmonary embolism is low. Currently, there are a small number of researches about the use of the clinical prediction rules in the ICU.To response to this, this study willevaluate the performance of the Wells and revised Geneva score and D-dimer for theprediction value for diagnosis of acute pulmonary embolism in the ICU in aretrospective analysis method and explore the suitable score to provide guidance forthe diagnosis of acute pulmonary embolism in the ICU.ObjectiveTo evaluate the performance of the Wells and revised Geneva score and D-dimerfor the prediction value for diagnosis of acute pulmonary embolism in the ICU andexplore the suitable score to provide guidance for the diagnosis of acute pulmonaryembolism in the ICU.MethodsCollected consecutive patients presenting to the Respiratory Intensive CareUnit(RICU) of the first affiliated hospital of Zheng Zhou University from the Marchof2012to August of2013with clinical features suspicious for APTE and accepting acomputed tomography angiography (CTPA) and a plasma D-dimer testing.Made thediagnosis of CTPA results for APE strictly in accordant with the imaginationsexamination procedures for pulmonary thromboembolism and deep vein thrombosismade by Chinese Medical Association. According to the guidelines for pulmonaryembolism made by Chinese medical association, set D-dimer0.5mg/L as thediagnostic criterion. Gathered patients’ clinical information and gained the scores byfilling the forms of Wells and revised Geneva score. Then used the curve of ROC toevaluate the prediction value of two scores and D-dimer for suspected APTE andcalculated the Youden Index for the cut-off point.Results1. Of the83suspected patients for APTE, there were2cases whose CTPAimages were not clear and cannot be identified for the diagnosis of APTE,1case with chronic thromboembolic pulmonary hypertension,3cases who had been diagnosed asCTPA in the past half year and accepting anticoagulation therapy,1case with a acutecourse of more than2weeks for the test of CTPA,1case who was taking regular doseof β1blockers during the imagine testing. Finally, confirmed75cases suspected forAPTE for the study,36cases of them were male and39were female.All the subjects’heart rate≥75beats/min,and54cases>95beats/min,47cases>100beats/min.31cases of the75suspected patients were established the diagnosis as APTE byCTPA. The overall prevalence of PE was41.3%.2. The predictive value of Wells score: of75suspected patients for APTE,evaluated by three-level Wells score,17cases with a low Wells score probability,37with a intermediate Wells score probability and21with a high Wells score probability.Of the three groups, the cases diagnosed as APTE by CTPA were2,19,10and the rateof a definite diagnosis were11.8%(2/17),51.4%(19/37),47.6%(10/21) respectively;evaluated by two-level Wells score,32cases in the unlikely PE group,43cases in thelikely PE group. Of the two groups, the cases diagnosed as APTE by CTPA were7,24and the rate of a definite diagnosis were21.9%（7/32）、55.8%（24/43）.3. The predictive value of Revised Geneva score: of75suspected patients forAPTE,7cases with a low Revised Geneva score probability,56with a intermediateprobability and12with a high probability. Of the three groups, the cases diagnosed asAPTE by CTPA were2,21,8and the rate of a definite diagnosis were28.6％(2/7)、37.5％(21/56)、66.7％(8/12) respectively.4. The predictive value of D-dimer: The range of D-dimer level was0.10-57.20（6.07±8.97）mg/L. Of75suspected patients for APTE,64cases with a positiveD-dimer testing and30(46.9%) of them with a definite diagnosis as APTE,11with anegative D-dimer testing and1(9.1%) of them with a definite diagnosis as APTE.5.The predictive value and comparisons of Wells, revised Geneva score andD-dimer for APTE: The area under ROC curve of two decision rules and D-dimerwere0.687±0.061(95%C10.566~0.807, P=0.566),0.582±0.068(95%C10.448~0.716, P=0.448),0.682±0.061(95%C10.562~0.802, P=0.562). Thecomparison of the area under ROC curve of Wells score and D-dimer was nostatistically significant difference (Z=0.06, P=0.48). The cut-off of Wells score was 3.5points and making3.5points the criterion, the sensitivity of predictive value ofWells score for APTE was83.9%, specificity was54.5%.The cut-off of D-dimertesting was1.355mg/L, the sensitivity for pulmonary embolism was90.3%, thespecificity was45.5%.6.The predictive value of Wells score combined with D-dimer for APTE: making0.5mg/L of D-dimer level as the criterion, negative predictive value of a lowprobability of three-level or PE unlikely of two-level of Wells score with a negativeD-dimer testing for pulmonary embolism were all100%; making1.355mg/L ofD-dimer as the criterion, negative predictive value of a low probability of three-levelWells score with a negative D-dimer testing for pulmonary embolism was90.0%;negative predictive value of PE unlikely of two-level Wells score combinedwith negative D-dimer testing for pulmonary embolism was88.2%.Conclusion1. To evaluate the clinical probability for patients suspected for APTE in theintensive care unit, the revised Geneva score may not be suitable and the Wells scoremay be suitable. However, its predictive value is limited.2．D-dimer has a high sensibility and a poor specificity for patients suspected forAPTE in the intensive care unit.3. In the intensive care unit, a low probability of three-level Wells score or PEunlikely of two-level Wells score combined with a negative D-dimer testing canexclude APTE patients.