To Evaluate The Clinical Value Of Systemic Inflammation Such As NLR And PLR Identify Epidural-Related Maternal Fever And Histological Chorioamnionitis Based On ROC Curve

Objective:To evaluate the value of joint detection such as NLR,PLR and other systemic inflammatory parameters in the identification of epidural-related maternal fever and histological chorioamnionitis,within pregnant women experiencing intrapartum fever during epidural analgesia.Methods:Using retrospective analysis,screening 119 cases of singleton full-term women receiving epidural analgesia during delivery from June 1,2018 to December 31,2018 in the Women and Children Hospital of Xiamen University.According to placental pathological examination,two groups were divided,67 cases were combined with HCA(observation group)and 52 cases were not combined with HCA(control group).Data statistics on the clinical data of the two groups,including:general conditions at the time of admission(age,height,gestation and parity,Days of pregnancy,body mass index etc.);maternal outcome(delivery ways,stage of labor,postpartum hemorrhage,urinary retention,etc.);Neonatal outcomes(Apgar score,neonatal hospitalization days,neonatal hospitalization costs,Admission rate of neonatal intensive care unit,neonatal-related diseases);laboratory indicators:blood routine parameters at admission,fever and within 24hours of fever after epidural labor analgesia and serum inflammatory indicators.The laboratory parameters with statistical differences between the observation group and the control group were compared and screened out,and the effects of these indicators on identifying HCA and ERMF were further analyzed.Statistical method:Use SPSS25.0 software to process the data,use receiver operating characteristic curve to analyze the performance of individual and joint prediction of each of the above indicators,and establish the relevant prediction mathematical model through the binary logistic regression equation.Results:Comparison of general conditions(age,height,gestation and parity,Days of pregnancy,body mass index,etc.),there was no statistical difference in general conditions between the two groups,p>0.05;maternal and neonatal outcomes,delivery ways(Assisted vaginal delivery,vaginal delivery,Cesarean section):the control group had a Assisted vaginal delivery,cesarean section of 84.6%,9.6%vs the observation group’s 58.2%,28.4%,the difference was statistically significant(p<0.05).Neonatal outcomes:the two groups had statistical differences in the Admission rate of neonatal intensive care unit,neonatal pneumonia,and neonatal metabolic acidosis.The rate of the Admission rate of neonatal intensive care unit,neonatal pneumonia,neonatal metabolic acidosis in observation group:83.6%,49.3%,49.3%vs control group:48.1%,26.9%,30.8%,p<0.05.Laboratory indicators:the two groups had statistical differences in the LYM%,NEU%,NEU,NLR,PLR,p<0.05.observation group:LYM%,NEU%,NEU,NLR,PLR:(12.22±5.52)%,(81.28±6.65)%,11.21(8.92-14.67)×109/L,6.24(4.87-9.2),121.66(91.11-165.12)vs control group(16.97±4.32)%,(75.5±5.90)%,9.80(8.66-10.93)×109/L,5.15(4.32-6.23),106.34(89.42-127.47).Among the single test results,the area under the curve of NEU%is the largest,with AUC:0.742,sensitivity:65.7%,and specificity:76.9%;the combined detection of LYM%+PLR in the two tests has a good effect,AUC:0.757,sensitivity:58.2%,specificity:88.5%;among the three tests,LYM%+NEU+PLR combined detection results are good,AUC:0.760,sensitivity:64.2%,specificity:80.8%;among the four tests.LYM%+NEU+NLR+PLR combined detection results are good,AUC:0.795,sensitivity:59.7%,specificity:92.3%;five combined detection results:LYM%+NEU%+NEU+NLR+PLR area under the curve is 0.799,sensitivity It is 76.1%,and the specificity is 76.9%.In the combined detection,the four indicators of LYM%+NEU+NLR+PLR have the best combined detection effect.The sensitivity is 67.2%,the specificity is 90.4%,and the Cut-off value is 0.619.In the multi-factor joint detection,the four joint detection results of LYM%+NEU+NLR+PLR are the best.The binary logistic regression equation is used to analyze and finally the mathematical model is:Logit(P)=Ln(P/1-P)=-0.011-0.205LYM%-0.586NLR+0.024PLR+0.367NEU,the regression analysis model prediction value(i.e.,LYM%+NEU+NLR+PLR combined detection)is used as a new variable for ROC curve analysis and diagnostic value comparison.The sensitivity of the combined new variable P at the maximum Youden index(0.575)was 67.2%,the specificity was 90.4%,the coincidence rate was 77.4%,the positive predictive value:90%,and the negative predictive value:68.1%.Use the predicted probability value of the model to perform ROC curve analysis,and calculate the AUC to be 0.801.Conclusion:1.Among pregnant women experiencing intrapartum fever during epidural labor analgesia,compared with the control group,the observation group had a higher cesarean section rate,a lower rate of vaginal delivery,the rate of the Admission rate of neonatal intensive care unit,neonatal pneumonia infection rate,and neonatal metabolism the incidence of neonatal metabolic acidosis is higher,and HCA should be recognized clinically as soon as possible to reduce adverse effects.2.The combined detection of LYM%+NEU+NLR+PLR four indicators is of great significance to the early identification of ERMF and HCA,and has good clinical application value.