Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Rules for predicting safe early hospital discharge for patients with upper gastrointestinal bleeding: A phase I (derivation phase) study

Wong, Grace L., Siu, Priscilla L., Wong, Vincent W., Lai, Larry H., Chow, Dorothy K., Rainer, Timothy, Chan, Francis K., Sung, Joseoph J. and Jau, James Y. 2006. Rules for predicting safe early hospital discharge for patients with upper gastrointestinal bleeding: A phase I (derivation phase) study. Gastrointestinal Endoscopy 63 (5) , AB153. 10.1016/j.gie.2006.03.295

Full text not available from this repository.


Backgrounds: Upper gastrointestinal bleeding (UGIB) is a common cause of hospitalization and resource utilization. The condition is often self-limiting in most patients. We aim to derive at decision rules to identify patients who may be discharged same day from the emergency rooms (ER) after endoscopic triage. Methods: Consecutive admissions of adult patients via ER with acute UGIB from April 2004 to April 2005 were prospectively entered into a database. We defined the high risk group as those requiring endoscopic therapy, those continued to bleed or developed re-bleeding, or the need for re-hospitalization, surgery and death within 30 days of admissions. Clinical and endoscopic parameters were analyzed using logistic regression model. Results: There were 432 admissions of adult patients for acute UGIB, and 99 (22.9%) were classified as the high risk group. Fresh hematemesis at presentation, systolic BP ≤ 100 mm Hg, low hematocrit ≤ 0.3, INR > 1.2, previous history of ulcer, ASA grade > 3 and presence of liver cirrhosis were independent factors that predicted events defined as above. These constituted a risk score. Those with a risk score ≤ 1 were categorized as low risk. The area under the ROC curve of our score was 0.71 (95% CI 0.65-0.77), which was significantly higher than that of Blatchford score (0.67 [0.61-0.74]), and the Rockall admission (0.60 [0.54-0.66]) and post-endoscopy scores (0.67 [0.61-0.73]) applied to the same cohort. Low risk patients may receive upper endoscopy at the level of ER, such that patients with high risk lesions warranting endoscopic therapy would be triaged to hospitalization, while others would be discharged. Screening with the risk score in combined with endoscopic triage at the ER allow 265/432 (61.3%) patients to be discharged on the same day. Conclusion: Our risk score is a simple method to identify low risk patients and to reduce admission when combined with endoscopic triage. It will be validated in another prospective cohort.

Item Type: Article
Date Type: Publication
Status: Published
Schools: Medicine
Subjects: R Medicine > R Medicine (General)
Publisher: Elsevier
ISSN: 0016-5107
Last Modified: 04 Jun 2017 09:15

Actions (repository staff only)

Edit Item Edit Item