Related Experiment Videos
Pre-endoscopic prognostic factors in non-varicose upper gastrointestinal bleeding. Development of a predictive
A Zaragoza Marcet1, J M Tenías Burillo, M J Llorente Melero
1Gastroenterology Unit, Lluís Alcanyís Hospital, Xátiva, Spain.
Revista Espanola De Enfermedades Digestivas
|August 21, 2002
Summary
A new algorithm predicts serious upper gastrointestinal bleeding (UDB) using simple clinical and lab factors. This tool helps classify patients early for better management of non-varicose UDB.
Area of Science:
- Gastroenterology
- Internal Medicine
- Clinical Prediction Models
Background:
- Upper gastrointestinal bleeding (UDB) is a frequent cause of hospitalization with significant morbidity and mortality.
- Non-varicose UDB requires effective tools for predicting patient outcomes.
- Early risk stratification is crucial for appropriate clinical management.
Purpose of the Study:
- To identify clinical and laboratory factors predicting a serious course of non-varicose UDB.
- To develop a simple algorithm for clinical application in assessing UDB severity.
- To enable early classification of patients with non-varicose UDB.
Main Methods:
- A longitudinal study of 803 hospitalized patients with non-varicose UDB.
- Clinical and laboratory data collected in the Emergency Service prior to endoscopy.
- Logistic regression models developed and simplified for clinical utility, with prognostic performance assessed via ROC curves.
Main Results:
- An initial model with 11 variables showed an AUC of 0.74.
- A simplified dichotomized model also achieved an AUC of 0.74.
- Major factors (shock, liver disease, anticoagulants, high urea/creatinine ratio) or two minor factors (low hematocrit, renal failure, hematemesis/melena, low blood pressure, alcohol use, UDB history, age > 75) predict an unfavorable UDB course.
Conclusions:
- The developed algorithm effectively utilizes pre-endoscopy data.
- It allows for early classification of patients with non-varicose UDB.
- This aids in predicting the clinical course and guiding management decisions.