Analysis of factors and risk prediction visualization model construction for recurrence of inguinal hernia after laparoscopic surgery in the elderly
Objective To analyze the factors of recurrence after laparoscopic inguinal hernia repair in elderly patients and construct a risk prediction visualization model.Methods The clinical data of 597 patients who underwent laparoscopic inguinal hernia repair in our hospital from Aug 2017 to Aug 2022 were retrospectively collected and randomly divided into a learning set(398 cases)and a testing set(199 cases)by a computer-generated random number table at a ratio of 2:1.All patients were followed up for 12～72 months after surgery,and the learning set patients who recurred were included in the recurrent group,and those who did not recur were included in the non-recurrent group.The general data of the learning set and the testing set were compared;the general data of the recurrent group and the non-recurrent group of the learning set were compared.Cox regression model was used to analyze the influencing factors of recurrence after laparoscopic inguinal hernia repair in elderly patients,and construct a regression equation and risk prediction nomogram model.Receiver operating characteristic(ROC)curve,Calibration curve,and decision curve analysis(DCA)were used to evaluate the predictive performance,calibration ability,and clinical net benefit of the model.Results There was no statistically significant difference in general data between the learning and testing sets(P<0.05).Among the 597 patients,63(10.55%)had recurrence after surgery.Cox regression analysis showed that male(HR=2.818,95%CI:1.703～4.663),body mass index(BMI)(HR=2.259,95%CI:1.209～4.222),long-term constipation(HR=3.080,95%CI:1.642～5.779),incarcerated hernia(HR=2.614,95%CI:1.159～5.897),inner ring diameter(HR=2.557,95%CI:1.398～4.677),use of absorbable ligation line(HR=2.396,95%CI:1.293～4.443),and time to start getting out of bed after surgery(HR=0.556,95%CI:0.368～0.839)were influencing factors of recurrence after laparoscopic inguinal hernia repair in elderly patients(P<0.05).Cox regression equation was constructed:h(t,x)=h0(t)exp(1.036x1+0.815x2+1.125x3+0.961x4+0.939x5+0.874x6-0.587x7).The area under the curve(AUC)of the risk prediction visualization model was 0.942(95%CI:0.914～0.962),the sensitivity was 92.68%(95%CI:80.13%～98.53%),and the specificity was 94.96%(95%CI:92.13%～97.12%).The AUC for predicting postoperative recurrence in the testing set was 0.926(95%CI:0.880～0.958),the sensitivity was 86.36%(95%CI:65.12%～97.15%),and the specificity was 90.40%(95%CI:85.13%～94.31%).There were no significant differences between the Calibration curves of the learning set and the testing set(χ2=0.854,P=0.165;χ2=0.905,P=0.097),The C-index values of the learning set and testing set were 0.915(95%CI:0.847～0.983)and 0.907(95%CI:0.869～0.945),respectively.The learning set and testing set obtained clinical net benefits within the risk threshold of 0～0.98 and 0～0.96,respectively.Conclusion Male gender,BMI,constipation history,incarcerated hernia,inner ring diameter,use of absorbable ligation line and postoperative recovery time are significant predictors of inguinal hernia recurrence in elderly patients.Based on the factors,the risk prediction visualization model is effecctive.
ElderlyInguinal herniaRecurrenceRisk factorsPrediction model
NETL
NSTL
万方数据
