The Study On Relationship Between P-wave Dispersion In Electrocardiogram And Paroxysmal Atrial Fibrillation

Objective:The relationships of the Paroxysmal Atrial Fibrillation(PAF)with P-wave dispersion(Pd),P-wave maximum duration(Pmax),P-wave minimum duration,(Pmin)in electrocardiogram(ECG)and Left atrial diameter(LAD)will be Investigated.Meanwhile,verify whether the ECG Pd,Pmax and the combination of them have predictive value for the occurrence of PAF.Methods:This was a retrospective study and selected randomly the 196 patients with PAF in the cardiology department of the second hospital of Tianjin Medical University as an observation group,and 194 patients without PAF another as an control group.The same type of 12-lead synchronous electrocardiogram automatic analyzer was used to collect them eligible ECG,the Pmax and Pmin were measured and the Pd was calculated.The data will be processed by Kolmogorov-Smirnov(K-S)Test for normal distribution.To be expressed by±S and analyzed by T Test about normal distribution data,and expressed by median±interquartile range(M±Q)and analyzed by Non-parametric Test about abnormal distribution data.Count data was described by percentage(%),made test with chi-square.The size of the test was?=0.05,P<0.05,indicated the difference was statistically significant.The critical value was 110ms in Pmax,40 ms in Pd and when Pmax≥110 ms+Pd≥40 ms,calculated their respective sensitivity(Se),specificity(Sp),positive predictive value(PV(+)),negative predictive value(PV(-)).Pd≥40 ms+Pmax≥110 ms had two hand means:one mean was that it was positive diagnosis so long as one of two indicators reached to the critical value,named“parallel tests”;another mean was the positive diagnosis must accord with that both indicators reach their critical values at the same time,named“serial tests”.Analyzed the accuracy of the indicators prediction above test diagnosis with receiver operating characteristic(ROC)curve and ROC area under curve(AUC).Results:The average age between two group were(66.7±10.54)vs.(67.73±10.04)years,P>0.05,differences was not statistically significant.The chi-square test shows the difference on common information data such as gender and comorbidity was not statistically significant(P>0.05).Pmax、Pd and LAD in observation group were greater than in control group,respectively(124.47±18.10)vs.(118.09±14.78)ms,(45.40±21.50)vs(25.6±17.40)ms,(40.75±7.58)vs(38.05±5.95)mm,P<0.001,Pmin in observation group less than in control group,(77.51±17.09)vs(90.51±15.98)ms,P<0.001,the difference between two group was statistically significant.When Pmax≥110 ms,Se=78.1%,Sp=27.3%,PV(+)=52%,PV(-)=55.2%.When Pd≥40 ms,Se=67.3%,Sp=82%,PV(+)=79%,PV(-)=71.3%.Apply“parallel tests”,Se=85.7%,Sp=25.3%,PV(+)=53.7%,PV(-)=63.6%,Apply“serial tests”,Se=59.7%,Sp=84%,PV(+)=79.1%,PV(-)=67.4%.The A_Z value of ROC-AUC was respectively:Pmax≥110 ms,A_Z=0.527,95%CI:0.470～0.584,P=0.358,compared with null hypothesis A_Z=0.5,the difference was not statistically significant and predictive diagnostic value was not high.Pd≥40 ms,A_Z=0.747,95%CI:0.697～0.796,P<0.001,compared with null hypothesis A_Z=0.5,the difference was statistically significant and predictive diagnostic value was higher.Pd≥40 ms+Pmax≥110 ms,A_Z=0.719,95%CI:0.667～0.770,P<0.001,compared with null hypothesis A_Z=0.5,the difference was statistically significant and predictive diagnostic value was higher.Conclusion:1.Pmin decreased,Pmax,Pd and LAD increased had relationships with PAF,It may be an electrocardiogram manifestation of atrial structural remodeling.2.The prediction value of the indicator Pmax≥110ms for occurrence of PAF was not high.3.The prediction value of the indicator Pd≥40 ms for occurrence of PAF was higher,and high specificity,lower false positive rate.4.When Pmax≥110 ms+Pd≥40 ms,the“parallel tests”shows higher sensitivity,lower false negative rate.Was a good predictive indicator for census and screening of PAF.the“serial tests”shows higher specificity,can be used to exclude prior screened patients without PAF.5.P-wave dispersion was a good,fast,convenient,noninvasive and economic indicator of diagnostic test for predicting occurrence of PAF.