| Background: Insufficient postoperative analgesia is a common phenomenon after cesarean section(CS).Pain is often experienced after cesarean section.Incidence of moderate to severe acute pain is as high as78%.Studies have found that about 20% of women experience severe acute pain within 24 hours of surgery.Severe acute pain after CS has been shown to be a strong predictor of chronic persistent pain and postpartum depression.Women with severe acute pain had a 2.5-fold increased risk of chronic persistent pain and a 3.0-fold increased risk of depression,respectively compared with women who experienced mild pain after surgery.In addition,comparing the cesarean section rate between China and the global average,it is found that the cesarean section rate in China is much higher than the global average.In 2015,the average rate of cesarean section in the world was about 21.1%,while the highest rate in China was 62.5%.There are many parturitions in China,but there is still no ideal model to predict severe pain after cesarean section.The efficacy of existing tools for predicting acute pain after cesarean section is mostly weak to moderate,so it is urgent to carry out relevant research to guide clinical pain management.Objective: To establish and verify the predictive model of severe acute pain after CS.Methods and Materials: This study was a case-control study approved by the Institutional Ethics Review Committee(Medical Ethics Committee of Xiangya Hospital,Central South University)and registered in the Chinese Clinical Trial Registry(Registration number:Chi CTR2100042883).A total of 233 pregnant women aged ≥18 who planned to undergo elective cesarean section in Xiangya Hospital of Central South University were given written informed consents.Demographic,psychological state,physiological changes during pregnancy and operation related parameters of these 233 parturient women were collected.Serological indexes were measured by enzymeLinked immunosorbent assay(ELISA).The peak value of pain and the use of remedial analgesics in four different periods of puerpera within 48 hours after operation were recorded.Numerical rating Scale(NRS)was used to evaluate the pain.Among them,those with NRS≥7,unbearable pain and requiring remedial analgesia were defined as severe pain group,while the rest of the women were automatically classified as non-severe pain group.Binary logistic regression method was used for data analysis.Firstly,univariate analysis was conducted on maternal demographic parameters,psychological state,physiological changes during pregnancy,and operative related factors to obtain the indexs of p-value < 0.1.After that,these indexes were included in the multivariate analysis,and the regression methods,including full inclusion,forward and backward were adopted,and the p-value parameters of entry and stay were set as 0.05 and 0.1,respectively.Finally,the intensity(adjust odds ratio and95%confidence interval)between the indexes of p-value < 0.05 and postoperative severe acute pain were obtained.Different models were established.Then,the Bayesian information criterion(BIC)is used to optimize the minimum selection model,and the prediction results are converted into nomogram,to predict postoperative pain.Then,the ROC curve(area under the curve,AUC),Calibration curve and decision curve analysis(DCA)of the modelwere obtained to evaluate the prediction ability and clinical applicability of the model.Finally,bootstrapping was used to verify the model internally to correct the sample overfitting.Result: In this study,10 out of 233 women underwent general anesthesia,5 parturients lost to follow-up after surgery due to various reasons and 17 parturients did not use patient controlled epidural analgesia(PCEA)after CS.The above 32 parturients were excluded and201 qualified parturients were included for data analysis.In this study,a total of 80 women experienced severe acute pain within 48 hours after CS with non-overlapping pain numbers,and the overall incidence of severe pain was 39.8%(80/201).There were 8 indicators with p-value < 0.1 in univariate logistic regression analysis,which were pain catastrophizing(P=0.002),WBCcount(P=0.049),prolactin(P=0.015),oxytocin(P=0.045),CCL21(P=0.019),CCL19(P=0.091),duration of surgery(P=0.01),mother-infant separation(P=0.061).The above 8 statistically significant factors were analyzed by binary logistic regression.Five preliminary models(s1,s2,s3,s4,s5)were established by full inclusion,forward and backward methods.The corresponding BIC values were calculated.Optimization model s5 was selected according to BIC minimum principle(273.26).s5 model was established by backward stepwise regression method.Finally,five factors were included in the model(P < 0.05)indicating that these five factors were independently correlated with severe acute pain after CS.The following are the p-values,odds ratio(OR)and corresponding 95% confidence interval(95%CI)of multivariate regression analysis of 5 indicators,including pain catastrophizing(P=0.011,OR=3.480,95% CI(1.329,9.114)),prolactin(P = 0.030,OR =4.526,95% CI(1.096,6.025)),CCL21(P = 0.007,OR = 2.446,95% CI(1.276,4.691)),Maternal separation phenomenon(P = 0.035,OR = 2.137,95% CI(1.053,4.338))and duration of surgery(P = 0.004,OR = 3.526,95% CI(1.482,8.392)).Finally,these five factors are integrated into a predictive model and transformed into a nomogram for clinical applying.In addition,the ability of the model was comprehensively evaluated in this study.The AUC of the evaluation model was 0.749,95%CI(0.681,0.816),indicating that the model had a good ability of discrimination(AUC between 0.7-0.9).At the same time,the best prediction probability truncation value is 0.394,corresponding sensitivity is 72.5%,specificity is 66.1%.H-l test(chi-square value =14.253,P >0.05)and calibration curve were obtained by the model.The results showed that the model had a good ability to correctly identify severe pain puerparas.In order to correct the possible over-fitting of research samples,this study conducted internal verification through bootstrapping and finally obtained the correction curve.Corrected c-index =0.714,95%CI(0.712,0.716)and optimism =0.070,95%CI(0.066,0.074).The results showed that the model still had a good and correct ability to distinguish between severe pain parturients after corrected fitting.In this study,DCA of this model was obtained based on the existing sample size.It was proved that nomogram could provide positive and net benefits for all pregnant women when the predicted probability P was in the range of 0-0.75,especially when the prediction probability P=0.394 provided by this study was used,the net benefit rate reached 40/100.When 100 women who using the predictive tool were evaluated,at least 40 women could obtain net benefit.Conclusion: The results of this study suggest that nomogram can provide a more convenient and reliable prediction of severe pain after CS.Finally,it can be used to guide the individualized analgesia of severe pain.The five factors in this model showed that when pain catastrophizating,prolactin and CCL21 levels were higher,operation of surgery was longer,and there was mother-infant separation,the women had a higher risk of experiencing severe pain after CS. |
PDF Full Text Request