Construction And Application Of Risk Prediction Model For Unplanned Readmission In Patients With Non-varicose Upper Gastrointestinal Bleeding

Objective:1.The hospital electronic medical records were collected to collect the demographic,related diseases and treatment information of patients with non-variceal upper gastrointestinal bleeding(NVUGIB).To investigate the status of unplanned readmission in NVUGIB patients within 30 days and analyze the influencing factors.2.The risk prediction model of unplanned readmission of NVUGIB patients was preliminarily constructed by logistics regression method and verified internally and externally,so as to assist medical staff in identifying high-risk groups of unplanned readmission and provide basis for further nursing intervention.3.According to the high-risk controllable factors in the prediction model based on the previous research results,and combined with the characteristics of the disease and expert opinions,the continuity care plan was formulated based on the IKAP model,and observe the application effect of continuous care in NVUGIB patientsMethods:1.By reading domestic and foreign literature and combining with the consultation results of senior medical staff in the department of Gastroenterology,we selected risk factors that may affect unplanned readmission of NVUGIB patients after discharge to develop a data collection questionnaire.Electronic medical records were collected from the department of gastroenterology of three grade A hospitals in Taiyuan,and demographic disease information and treatment information of NVUGIB patients who met the requirements from May 2019 to May 2021 were retrospectively collected,Patients who are hospitalized need to collect the number of rehospitalization,the reason,and the time to be hospitalized.The single factor analysis sifies the risk factors of the non-planned rehospitalization of NVUGIB patients,and the binary logistic regression analyzes the independent risk factors,the risk prediction model was constructed based onβand OR values,the weight assignment of each risk factor is converted into a simple prediction model formula.2.According to the modeling group(70%)and verification group(30%)ratio,convenience sampling method was used to collect NVUGIB patients from two grade A hospitals in Taiyuan city from June 2021 to February 2022 as a validation group,to conduct internal and external tests of the prediction model.The fitting degree of the prediction model was tested by H-LChi-square test,P>0.05 indicates that there is no significant difference between the prediction probability and the actual incidence,so the prediction effect is good and the consistency is high,and vice versa.The AUC was used to judge the discrimination,the closer AUC is to 1,the stronger the discriminant effect of the model will be.3.NVUGIB patients in a third class a hospital in Taiyuan were randomly divided into intervention group and control group.To explore the application effect of the IKAP model based on the continuity nursing program based on the high-risk controllable factors in risk prediction model.The improvement effect of health behavior and unplanned readmission were compared between the two groups one month after discharge.Results:1.The model group was included in 523 patients with NVUGIB,42 cases of non-planned rehospitalization,and 8.03%for non-planned rehospitalization rates.Age>60 years old(OR=2.086)、GBS score>6points(OR=2.802)、CCI score>3points(OR=2.593)、LOS>7days(OR=3.087)、take NSAIDS(OR=3.357)、history of drinking(OR=4.778)and blood transfusion(OR=4.112)were the independent risk factor for unplanned readmission,endoscopic examination(OR=0.350)was a protective factor After weight assignment of all risk factors,the unplanned readmission risk prediction model of NVUGIB patients was finally established:Y=1×age(>60 years old)+1×GBS score(>6points)+1×CCI score(>3points)+2×LOS(>7days)-3×endoscopy+2×history of drinking+2×blood transfusion+2×NSAIDS.2.H-L test analysis results showed x~2=8.533,P=0.383,with no statistically significant difference(P>0.05),and the probability of prediction was 90.2%.The AUC of the modeling group was 0.862(95%CI:0.788～0.936),the Yuden index was 0.665,the specificity was 81.1%,and the sensitivity was 85.4%.The AUC of the validation group was 0.848(95%CI:0.769～0.926),the Yuden index was 0.613,the specificity was 84.8%,and the sensitivity was 76.5%.3.One month after implementing the continuous nursing program,the scores of each dimension(health responsibility,nutrition,exercise,interpersonal support,stress management and self-actualization)and the total score of the intervention group were higher than those of the control group(P<0.05).The nursing satisfaction(94.9%)was better than the observation group(84.0%)(P<0.05),and the unplanned readmission rate(2.6%)in the intervention group was lower than that in the observation group(12.0%)(P<0.05).Conclusion:NVUGIB patients have a high incidence of unplanned readmission and many independent risk factors for unplanned readmission.The constructed prediction model has good distinction and fit degree,and high prediction efficiency,which has important reference significance for identifying unplanned readmission high-risk groups.Continuing care based on the risk controllable factors in the predictive model can significantly improve the health behavior of patients with NVUGIB,increased satisfaction,and reduced the rate of unplanned readmission,which can provide a reference for the clinical continuation care of such patients.