Construction Of A Clinical Prediction Model For Real-time Assessment And Prognosis Of Acute Upper Gastrointestinal Bleeding

Background and ObjectiveAcute upper gastrointestinal bleeding(UGIB)is a common problem that associated with significant morbidity and mortality.Across the last two decades several studies have demonstrated the incidence of UGIB has fallen,the proportion of critically ill UGIB patients has increased,and the mortality rate has not changed significantly.This result lacks epidemiological data for Chinese UGIB patients.At present,the evaluation and diagnosis strategies of different clinicians for UGIB patients are still not uniform.Identify high-risk and low-risk patients based on the clinician's experience and limited relevant indicators.There is an urgent need for an effective and easy-to-use risk assessment model to identify those acute UGIB patients who are at high risk of death in a Chinese population given the prevalence and mortality of cases in China.This model can not only early identify patients at high risk of death early,but also evaluate the effectiveness of treatment.There are a variety of risk scores that can be used for acute UGIB patients,each with multiple studies analyzing their outcomes.Studies of risk scores have been developed in both geographically and pathologically diverse populations.Each was powered to detect different outcomes and the robustness of external validation varies considerably.None of the existing scores has been widely accepted in clinical practice.The objective of our research is to establish a predictive model that can identify AUGIB patients at high risk of death early and quickly,dynamically guide doctors in the selection of treatment measures,evaluate the treatment effect,and predict the prognosis of UGIB patients.Material and MethodsThis study is a multi-centre,prospective,real-world study.This research consists of 3 parts:1.This study collected the epidemiology data of acute UGIB patients admitted from the emergency department in 20 centres across the country;2.We established a clinical risk prediction model for early assessment of the condition and prognosis of acute UGIB patients in the establishment cohort,and verified the effectiveness of the model in the validation cohort of patients;3.We compared the effectiveness of the clinical risk prediction model established in this study with the existing international scoring system in predicting death.ResultsA total of 1072 patients with acute upper gastrointestinal bleeding were included in our cohort.The results of the study are as follows:1.The survey of epidemiological data shows that among the population of upper gastrointestinal bleeding in China,the proportion of male patients is more than that of females,and people over 60 years old account for more than half of the total population.The existing comorbidities and the severity of the comorbidities are related to the death of patients with upper gastrointestinal bleeding.Among all comorbidities,patients with liver cirrhosis and gastrointestinal tumors have higher rebleeding and mortality.25%of patients did not undergo endoscopy,and the deaths of patients who did not undergo endoscopy accounted for 50%of all deaths.For all patients,gastrointestinal ulcers are still the most common cause of gastrointestinal bleeding,which is related to the widespread use of antiplatelet drugs.Rebleeding is not only an evaluation result of patients with gastrointestinal bleeding,but also an important risk factor for death.Patients with bleeding?4 times have an increased risk of death.The all-cause mortality(10.91%)of all patients was consistent with other reports in the world,and deaths from gastrointestinal bleeding accounted for 50%of all deaths2.The UGIB evaluation model(hibag model)we constructed can effectively predict the 90-day mortality of patients with upper gastrointestinal bleeding.The two risk scores in the evaluation model are consistent and excellent in both the establishment cohort and the validation cohort(AUROCs:0.77-0.81).In UGIB,patients with low hibag-1 score(?4),medium hibag-1 score(5-9)and high hibag-1 score(?10)at admission had 90-day mortality rates of 4.10%,15.56%and 44.29%,respectively;patients with low hibag-2 score(<4),medium hibag-2 score(5-9)and high hibag-2 score(?10)within 48 hours of admission had 90-day mortality rates of 4.88%,26.20%,and 100%,respectively.3.In comparison with existing international scores(pRS,RS,AIMS65,ABC),our UGIB assessment model has excellent predictions of 90-day death at the time of admission and 48 hours after admission(AUROCs:0.62-0.72,P?0.02).ConclusionThis study is a large-scale,multi-center,prospective real-world study of upper gastrointestinal bleeding.We collected epidemiological data on patients with upper gastrointestinal bleeding in China,verified the risk factors related to death,and discovered new predictive factors related to death.This study established a risk assessment model for dynamically assessing the condition and predicting 90-day death from two time points(on admission,48 hours after admission).This risk assessment model is superior to the existing scoring system in predicting 90-day mortality.This model can be used to 1)Identify patients at high risk of death when they are admitted to the hospital,and give corresponding treatment measures in time;2)Evaluate the effect of the previous treatment within 48 hours after the patient is admitted to the hospital,and make dynamic adjustments to improve the prognosis;3)Make accurate and complete predictions of the death of patients,and conduct effective doctor-patient communication as soon as possible.In summary,this model can initially realize individualized treatment based on evidence-based medicine,thereby improving medical quality,reducing medical risks,reducing medical costs,and completing early risk prediction,evaluation and management of patients with upper gastrointestinal bleeding.The model provides a basis for establishing a complete diagnosis and treatment system for upper gastrointestinal bleeding.