| Objective To evaluate and compare the diagnostic value of different clinicalpredictive scores to patients with suspected VTE, and construct a clinical predictionmodel which suitable for chinese patients and compare the diagnostic value.Methods The hospitalized patients who were clinically suspected VTE underwentlower extremity compression sonography, CTPA (CT pulmonary angiography), CTvenography, and combined CTPA and CT venography were consecutive reviewedbetween July2006and March2009in affiliated hospital of NingXia medicaluniversity. To collect the clinical data, laboratory examination data and imaging dataof patients, and input information into database by Epidata3.0. We scored eachpatient in accordance with seven clinical scores. Three categorized patients intogroups with low, intermediated and high risk for VTE. In this study, patients with lowpre-test probability as VTE absent, with intermediated and high pre-test probability asVTE present. Imaging data was used as the reference standard. SPSS statisticalpackage was taked for statistical analysis. Screening of clinical indicators whichclosely related to VTE. Logistic regression model analysis was used to calculate theregression coefficients of the corresponding indicators and risk, respectively.Difference was regarded as statistical significant when P＜0.05.Results1. The records of213consecutive PTE patients were reviewed, dyspnea chestdiscomfort141(80.28%), cough105(49.29%), pleural chest pain91(42.72%), moistrale81(38.02%), unilateral leg swelling or pain80(37.55%), sputum80(37.55%),cyanosis62(29.10%), difficulty in breathing50(23.47%), lower extremity varicosevein38(17.84%). And there were143cases with DVT: unilateral leg swelling76(53.14%), unilateral leg pain53(44.05%), lower extremity varicose vein44(30.76%), low extremity enlargement greater than3cm30(20.97%), deep venous tenderness25(17.48%), the local skin of the lower extremity temperature increased22(15.38%), lower extremity pigmentation13(9.09%).2. The independent risk factor for PTE patients were pleural pain, leg swelling or pain,previou PTE or DVT, syncope, recent brake or in bed, active cancer, lower extremitysurgery within four weeks, protective factors were coronary heart diasese, cardiacinsufficiency. The independent risk factor for DVT patients were unilateral lowerlimb swelling, unilateral lower limb pain, low extremity enlargement greater than3cm,the local skin of the lower extremity temperature increased, deep venous tenderness,recent brake or in bed, previou DVT, lower extremity surgery within four weeks. Theindependent risk factor for VTE patients were unilateral lower limb swelling,unilateral lower limb pain, pleural pain, low extremity enlargement greater than3cm,deep venous tenderness, recent brake or in bed, previou PTE or DVT, lower extremitysurgery within four weeks, protective factors were coronary heart diasese.3. The actual incidenc rate in group with low, intermediated and high risk for PTEwere15％～18％、33％～48％、35％～85％by Wells score, Geneve score andimprove Geneve score. Respectively, sensitivity, specificity, NPV and PPV of Wellsscore were70.1%,79.1%,86.4%,73.1%; sensitivity, specificity, NPV and PPV ofGeneve score were37.0%,31.0%,53.0%,33.4%; sensitivity, specificity, NPV andPPV of improve Geneve score were35.2%,84.9%,81.5%,84.0%. The AUC ofWells score, Geneve score and improve Geneve score were0.783,0.569and0.658.The actual incidenc rate in group with low, intermediated and high risk for DVT were14％～56％,46％～66％,78％～100％by lower extimety Wells score, Kahn score,Andre score and constans score. The AUC were0.845,0.526,0.761and0.755. TheAUC of diagnosis VTE by Wells score, Geneve score and improve Geneve scorewere0.818,0.510and0.669. The AUC of diagnosis VTE by lower extimety Wellsscore, Kahn score, Andre score and constans score were0.759,0.518,0.695and0.690.4. The actual incidenc rate in group with low, intermediated and high risk for PTEwere10%,43%and72%by the new clinical prediction model. Respectively,sensitivity, specificity, NPV, PPV and AUC of this clincial prediction model were 84.0%,61.7%,48.8%,89.9%and0.780. Moreover, the different between the newclinical prediction model and Wells score was not significant, but the new clinicalprediction model has higher accuracy compared with Geneva score and improveGeneva score (P<0.05).The actual incidenc rate in group with low, intermediated and high risk for DVTwere11%,61%and95%by the new clinical prediction model. Respectively,sensitivity, specificity, NPV, PPV and AUC of this clincial prediction model were91.6%,66.7%,74.4%,88.2%and0.896. The new clinical prediction model hashigher accuracy compared with lower extimety Wells score, Kahn score, Andre scoreand constans score (P<0.05).The actual incidenc rate in group with low, intermediated and high risk for VTEwere22%,79%and98%by the new clinical prediction model. Respectively,sensitivity, specificity, NPV, PPV and AUC of this clincial prediction model were90.9%,67.8%,86.3%,77.0%and0.878. The different between the new clinicalprediction model and Wells score was not significant, but the new clinical predictionmodel has higher accuracy compared with Geneva score and improve Geneva score(P<0.05).Conclusion1. Summary of the clinical manifestations of pulmonary embolism anddeep venous thrombosis in patients, the study found that dyspnea chest discomfort,cough, pleural chest pain is the most common clinical signs and symptoms of patientswith pulmonary embolism, and should be alert to the possibility of pulmonaryembolism for this patients. Moreover, unilateral leg swelling and unilateral leg pain isthe most common clinical signs and symptoms of patients with DVT, and should beattach important to the possibility of DVT.2. Through the analysis of independent risk factors for PET and DVT, the resultsshowed that pleural pain, syncope, active cancer, previou PTE or DVT, lowerextremity surgery within four weeks, recent brake or in bed, the local skin of thelower extremity temperature increased, deep venous tenderness, unilateral lower limbswelling, unilateral lower limb pain, low extremity enlargement greater than3cmwere the independent risk factors for VTE, and should pay attention to prevent the occurrece of VTE. The OR value prompt unilateral lower limb swelling is the highestrisk factors for VTE diagnosis, followed by pleural pain. Coronary heart diasese,cardiac insufficiency was the protective factors.3. In our study, the PET Wells score apply to the patients with suspected PET, lowerextiremity Wells score is more suitable for patients with suspected DVT, thepulmonary embolism Wells score and the improve Geneva score is more applicable topatients with suspected with VTE, and the pulmonary embolism Wells score hashighest diagnositc value.4. To bulid the new PTE, DVT and VTE clinical prediction model and compared withthe foreign clinical predictive score, we found that the value of the new clinicalprediction model was similar to the previous model. However, the actual value of thenew model has yet to be confirmed by further studies. |
PDF Full Text Request