Analysis On The Predictive Value Of Noninvasive Urodynamic Parameter Model For Detrusor Contractility

Objective:The value of constructing a non-invasive urodynamic parameter model to predict the detrusor contractility was analyzed.Methods:A retrospective analysis of the clinical data of 310 patients who underwent the first urodynamic examination at the Department of Urology,Shaoyang Hospital,South of China University from 2016 to 2019.With WFmax7w/m~2 as the boundary,it was divided into group A with WFmax≤7w/m~2(decreased detrusor contractility),a total of 54 cases,group B with WFmax>7w/m~2(detrusor contractility was acceptable),256 cases.Spearman correlation analysis was performed on all non-invasive urodynamic parameters except age and IPSS,and then single factor analysis and multi-factor regression analysis were performed on each parameter.Then,clinical prediction models were constructed based on the analysis results and clinical experience.Finally,ROC curve and AUC were plotted respectively,and then pound-wise comparative analysis was conducted.Results:Spearman correlation analysis was performed on non-invasive parameters except age and IPSS.The results show that:(1)Qmax is weakly negatively correlated with BWT(Spearman correlation coefficient is less than-0.4)and PVR is strongly negatively correlated(Spearman correlation coefficient is-0.58);(2)BWT is weakly positively correlated with PVR(Spearman correlation coefficient is less than 0.4).Univariate analysis showed no statistically significant differences in parameters such as age,BWT,IPP,PV,IPSS,and fPSA between the two groups(P>0.05),and significant differences in parameters such as Vv,PVR,tPSA,and Qmax between the two groups(P<0.05).The univariate regression analysis showed statistically significant differences in PVR,Qmax,and Vv parameters(P<0.05),but no significant differences in age,PV,BWT,IPP,tPSA,fPSA,and IPSS parameters(P>0.05).Multivariate regression analysis showed statistically significant differences in PVR,Qmax,and BWT parameters(P<0.05),and no significant differences in age,PV,Vv,IPP,tPSA,fPSA,and IPSS parameters(P>0.05).Construct a clinical prediction model based on univariate andmultivariateregressionanalysisresults Y=-0.510-0.005PVR+0.152Qmax+0.345BWT,PVR,BWT,Qmax parameters have statistical differences(P<0.05);included according to clinical experience Age,BWT,PVR,Qmax were used to construct a clinical prediction model L=0.938-0.02age-0.005PVR+0.338BWT+0.15Qmax.There was no statistically significant difference in patient age parameters(P>0.05),and the differences in PVR,BWT,and Qmax parameters were statistically significant(P<0.05).AUC of prediction model L=0.732(95%CI:0.679-0.780),AUC of prediction model Y=0.729(95%CI:0.676-0.778),AUC of PVR=0.686(95%CI:0.608-0.764),Qmax AUC=0.659(95%CI:0.583-0.735),AWT of BWT=0.576(95%CI:0.478-0.665);AUC pairwise comparison of the parameters of the regression model L and Qmax,BWT parameters have statistical significance(P<0.05),with marginal significance with PVR(P=0.062);no significant difference with prediction model Y(P>0.05);differences between prediction model Y and non-invasive urodynamic parameters Qmax and BWT are statistically significant Significance(P<0.05),withmarginalsignificancewithPVR parameters(P=0.077);there was no statistical difference between the two pairs of Qmax,BWT,and PVR parameters(P>0.05).The optimal intercept point L of the prediction model L=1.26,the specificity of the predicted detrusor contractility(DU)reduction of 78.21%,and the sensitivity of 59.26%(P<0.01);the optimal intercept point value Y of the prediction model Y=1.46,prediction The specificity of weakened detrusor contraction(DU)was 69.53%,and the sensitivity was 66.67%(P<0.01).Conclusion:(1)Compared with PVR,Qmax and BWT,the predictive models L and Y have higher specificity for predicting DU,which is helpful for the clinical diagnosis of DU.(2)The prediction model L has the highest specificity and lower sensitivity,while the prediction model Y is more balanced,but the former has greater clinical value.(3)PVR and Qmax have high sensitivity and low specificity,which can be used as a clinical initial screening index of DU and combined with the prediction model L,which is expected to improve the sensitivity and specificity of DU diagnosis.