| Obective:Constructing a recurrence risk prediction model for patients with severe acute pancreatitis(SAP)and verify the model to provide a screening tool for people at high risk of SAP recurrence.Methods:In this study,a retrospective method was used to collect relevant clinical data of SAP patients admitted to a tertiary hospital in Zunyi City from January 2010 to October 2019.According to the case data,the patients were divided into recurrence group and non-recurrence group.A total of 421 patients were included in the study,and they were divided into the modeling group(January 2010-December 2017)according to the length of hospitalization,accounting for 70%of the total number,and the verification group(January 2018 to October 2019),accounting for 30%of the total number of people.Refer to domestic and foreign literature,select the risk factors that may affect the recurrence of SAP,and use the hospital electronic case information system to collect relevant information,including general data and clinical data.Use SPSS18.0 and R3.5 software for statistical analysis.Normal distribution is represented by mean±standard deviation,comparison between groups is by t test,measurement data of non-normal distribution is represented by median and quartile,and comparison between groups is by nonparametric test.The count data use case(%)indicates that the chi-square test is used for comparison between groups.Single-factor and multi-factor Logistic regression analysis to determine the risk factors for SAP recurrence,and then build a nomogram prediction model in the modeling group(drawn by the R software"rms"package)through the established risk factors,and verify the model in the verification group to predict The performance of the model is evaluated by the degree of discrimination and the degree of calibration.Results:(1)A total of 296 patients were included in the modeling group of this study,of which 68 patients relapsed after discharge,the recurrence rate was 22.97%,and the recurrence rate within 1 year after discharge was relatively high,accounting for 89.38%of the total recurrence rate.The main cause of recurrence is hyperlipidemia,followed by biliary,alcoholic,and other causes.(2)There were 6 risk factors for relapse in univariate analysis,including gender,cholelithiasis,diabetes,alcohol consumption,blood glucose,and triglyceride,and the difference was statistically significant(P<0.05).There was no statistical difference in other remaining indicators between the relapse group and the non-relapse group(P>0.05).(3)Based on the results of univariate analysis,factors with P<0.05 were included in the multivariate Logistic regression analysis.The results showed that cholelithiasis,diabetes,alcohol consumption,triglyceride and blood glucose were independent risk factors for SAP patients’recurrence.(4)The area sensitivity and specificity under the ROC curve of the modeling group were 84.1%and 70.6%,and the area sensitivity and specificity under the ROC curve of the verification group were 80.4%and 68.2%.Hosmer-Lemeshow test results showed no statistical significance,the model better goodness-of-fit(?~2=3.721,P=0.881),the calibration curve shows nomogram occurred during the validation group is more consistent with the actual situation of risk.Conclusion:1.The recurrence rate of SAP patients in this hospital was 22.97%,which was more likely to occur within 1 year after discharge.The main reason for recurrence was hyperlipidemia.2.Previous history of diabetes,history of cholelithiasis,alcohol consumption,highest blood glucose and triglyceride level within 24 hours of admission are risk factors for SAP recurrence.3.The prediction efficiency and fitting degree of this model are good,which has certain guiding significance for identifying high-risk groups and predicting recurrence risk,and has important clinical value and innovation. |
PDF Full Text Request