Development And Evaluation Of Prognostic Models Of Adults With In-Hospital Cardiac Arrest

Background:IHCA(In-hospital Cardiac Arrest)has always been one of the problems faced by emergency physicians at home and abroad,which is characterized by its high mortality and disability rate.CPR(Cardiopulmonary Resuscitation)is currently the most efficient measure for patients with IHCA.Whether achieving continuous ROSC(Return of Spontaneous Circulation)or not,and the level of the neurological function prognosis are essential indicators for outcomes of IHCA.IHCA has become one of the common medical hotspots around the world for its high morbidity,low success rate of resuscitation,low survival discharge rate and poor neurological function prognosis and brought heavy burden to families and society.The development of clinical prognostic model can fast,accurately and efficiently evaluate the disease,predict the outcome,so as to improve the recovery success rate and prognosis of patients and to some extent reduce the burden of patients and their families or society.The study constructed and evaluated Non-ROSC and poor neurological function at 12-month after IHCA prognostic models by retrospectively analyzing the prognosis factors of adult IHCA patients,aiming to reduce the mortality rate of IHCA patients,improve the prognosis and provide a favorable reference for clinical work.Objective:To investigate the clinical characteristics of adult patients with IHCA,analyze the independent factors influencing the prognosis of IHCA patients,so as to construct and evaluate prognosis models of Non-ROSC and poor neurological functional prognosis at 12-month after IHCA.Methods:A retrospective analysis from September 1,2019 to December 31,2020 was done,including all patients with IHCA occurring in The First Hospital of Jilin University.The area included public area,general ward,emergency room,intensive care unit,operating room and so on.The study subjects were determined according to the inclusion and exclusion criteria,and subsequently divided into ROSC group and Non-ROSC group.Patients who had achieved ROSC for more than 20 minutes were followed up to obtain neurological status at 12-month after IHCA,according to the consequence,were divided into good neurological function prognosis group and poor neurological function prognosis group.The primary outcome was Non-ROSC,and the secondary outcome was a poor neurological outcome at 12-month after IHCA.X~2,Fisher’s,Mann-Whitney U tests were used for comparison between groups,and univariate and multivariate Logistic regression analyses were used to select independent factors associated with Non-ROSC and poor neurological prognosis at12-month after IHCA.Graphing nomograms was based on the above results.The discrimination capacity was evaluated using the receiver operating characteristic(R OC)curve,and the area under the curve(AUC);the calibration capacity was evaluated by Bootstrap self-sampling method,calibration curve,C-index,average absolute error(MAE)and Hosmer-Lemeshow goodness of fit test;decision curve analysis and clinical impact curve were used for the clinical validity assessment of prognostic models.Results:1.A total of 2129 patients suffered IHCA in the study period.851 adult patients were included by inclusion/exclusion criteria,564(66.27%)patients had not achieved ROSC,whereas 287(33.73%)patients had achieved ROSC for more than 20minutes,of which only 58(6.82%)patients had been alive at 12 months after IHCA and owned good neurological functional prognosis(CPC 1-2);229(26.91%)patients were with poor neurological outcome(CPC 3-5)at 12-month after IHCA after ROSC.2.Among the included 851 subjects,543(63.81%)were male,308(36.19%)were female(M:F=1.76:1).The youngest age was 18,the oldest was 100 and the median was 65(54,74).3.Comparison between the ROSC and Non-ROSC groups:by the multivariate Logistic regression,men,advanced age,CPR duration over 23 minutes,and total application dosage of adrenaline application over 3mg are independent risk factors for Non-ROSC;IHCA combined with arrhythmia,initial cardiac rhythm type as shockable,and the establishment of advanced airway are independent protective factors;A nomogram prognostic model was constructed based on the above seven variables,the AUC value is 0.904(95%CI:0.882-0.925),the C-index is 0.901(95%CI:0.879-0.922);Bootstrap internal validation showed the MAE is 0.024;Hosmer-Lemeshow test shows X~2=12.36,P=0.136.When the threshold probability was≥0.9,the number of people classified as high risk by the model was consistent with that of true-positive patients.4.Comparison between groups with good neurological outcome 12-month after IHCA:by the multivariate Logistic regression,IHCA combined with respiratory failure or shock,occurrence in the monitoring area,establishment of advanced airway,and application of norepinephrine are independent risk factors for poor neurological function prognosis;cardiac aetiology,initial rhythm as the shockable are independent protective factors,A nomogram prognosis model was constructed based on the above seven variables,the AUC value is 0.912(95%CI:0.870-0.954),C-index of 0.918(95%CI:0.875-0.957);Bootstrap internal validation showed the mean absolute error is 0.035;Hosmer-Lemeshow test shows X~2=11.10,P=0.196.When the threshold probability was≥0.95,the number of people classified as high risk by the model was consistent with that of true-positive patients.Conclusions:1.Among the 851 patients with IHCA included in this study,the majority were elderly male.The mainly presumed aetiology of IHCA was cardiac,and the initial heart rhythm was mainly unshockable.The Non-ROSC rate was 66.27%and the poor neurological function prognosis rate of 12-month after ROSC was 93.18%.2.Men,advanced age,CPR duration over 23minutes,and total application dosage of adrenaline over 3mg are independent risk factors for Non-ROSC;IHCA combined with arrhythmia,initial cardiac rhythm type as shockable and the establishment of advanced airway are independent protective factors.IHCA combined with respiratory failure or shock,occurrence in the monitoring area,establishment of advanced airway,and application of norepinephrine are independent risk factors for poor neurological function prognosis at 12-month after ROSC;cardiac aetiology,initial rhythm as the shockable are independent protective factors.3.Both of prognostic models own good discrimination and calibration,decision curves also confirm their good clinical applicability.