Clinically Significant Gastrointestinal Bleeding Events: Development and Validation of an Electronic Detection Algorithm and Application in a Population-Based Retrospective Cohort Study of Adult Critically Ill Patients in Alberta
CCCF ePoster library. Zuege D. 11/12/19; 283375; EP65 Disclosure(s): Research supported by CIHR grant.
Dr. Danny Zuege
Dr. Danny Zuege
Login now to access Regular content available to all registered users.

You may also access this content "anytime, anywhere" with the Free MULTILEARNING App for iOS and Android
Rate & Comment (0)
Topic: Retrospective or Prospective Cohort Study or Case Series

Danny J. Zuege1,5; Niven, Daniel1,4; Soo, Andrea1 ;Harris, Jo3; Mathew, Stanley3; Sargento, Avan3; Agyemang, Malik3; Bagshaw, Sean M2,5
1. Department of Critical Care Medicine, Cumming School of Medicine, University of Calgary, and Alberta Health Services, Calgary, Canada
2. Department of Critical Care Medicine, Faculty of Medicine and Dentistry, University of Alberta, and Alberta Health Services, Edmonton, Canada
3. eCritical Alberta Program, Alberta Health Services, Calgary, Canada.
4. Department of Community Health Sciences, Cumming School of Medicine, University of Calgary, Calgary, Canada
5. Alberta Critical Care Strategic Clinical Network, Alberta Health Services, Alberta, Canada.

ICU-acquired gastrointestinal bleeding (GIB) is estimated to occur in 1-3% of admissions (1, 2).  Clinically significant GIB events have been defined as events associated with hypotension, initiation or increased dose of vasoactive agents, significant drop in hemoglobin or transfusion of ≥ 2 units of blood (1, 2).  In Alberta, all ICUs are supported by the eCritical program which provides a comprehensive bedside clinical information system (eCritical MetaVision) and data warehouse and clinical analytics system (eCritical TRACER) (3).  All data to detect clinically significant GIB events are part of routine clinical documentation within eCritical systems.


  1. To create and validate an automated algorithm within eCritical systems to detect clinically significant GIB events
  2. To define the clinical epidemiology of GIB events in a retrospective population-based cohort of adult ICU patients in Alberta

An automated electronic algorithm was developed using SQL coding within eCritical systems to detect clinically significant GIB events. The algorithm was tested from technical and calculation perspectives using test cases. Initial clinical testing was undertaken via manual review of archived eCritical charts enriched with GIB events to test for positive and negative cases.  Discordance between manual and electronic GIB identification was sourced to optimize the algorithm.
A final independent clinical validation was undertaken using a random sample of 60 adult ICU admissions enriched with GIB cases. Archived charts were independently reviewed by 2 investigators blinded to electronic algorithm results. Disagreements were resolved via discussion.  The final consensus manual results were compared to the electronic algorithm to establish validation. 
A retrospective population-based cohort study was then performed including all adult patients admitted to Alberta ICUs from Jan 1, 2015 to Feb 28, 2019, excluding than those admitted with a  admission diagnosis of GIB. GIB events occurring ≥24 hours from ICU admission were identified using the validated detection algorithm.  The incidence and timing of GIB events is described. 

The electronic detection algorithm was successfully developed and optimized via technical and clinical testing.  In the final validation, the agreement between the two reviewers prior to discussion for the presence of a GIB event was 97.5% (Kappa 0.95).  The agreement between the final consensus of manual case review and electronic algorithm results for the presence or absence of a GIB event was 97.5% (Kappa 0.95).   Cases with discrepancies between algorithm and manual review related to human errors with the manual review, in particular reviewing high fidelity (per minute) physiologic data. 
Clinically significant GIB events were detected in 3002 out of 48368 (6.2% 95%CI 6.0–6.4) ICU admissions (13.1 events per 1000 patient-days) with significant variation between ICUs.  Events occurred a median of 4.0 (IQR 2.1-7.9) days after admission.  29.4% of events occurred ≥7 days after admission.
We successfully developed and validated an automated electronic algorithm to detect events of clinically significant GIB, which occurred in 6.2% of adult ICU admissions in Alberta.  Automated detection of GIB events will enable real-time measurement of this complication of critical care and allow opportunity to perform pragmatic clinical trials with this endpoint being captured automatically.

1.  Krag M, Perner A, Wetterslev J, et al. Prevalence and outcome of gastrointestinal bleeding and use of acid suppressants in acutely ill adult intensive care patients. Intensive Care Med 2015;41:833-45.  
2.  Cook DJ, Fuller HD, Guyatt GH, et al. Risk factors for gastrointestinal bleeding in critically ill patients. Canadian Critical Care Trials Group. N Engl J Med 1994;330:377-81.
3.  Brundin-Mather R, Soo A, Zuege DJ, et al. Secondary EMR data for quality improvement and research: A comparison of manual and electronic data collection from an integrated critical care electronic medical record system. J Crit Care 2018;47:295-301.
    This eLearning portal is powered by:
    This eLearning portal is powered by MULTIEPORTAL
Anonymous User Privacy Preferences

Strictly Necessary Cookies (Always Active)

MULTILEARNING platforms and tools hereinafter referred as “MLG SOFTWARE” are provided to you as pure educational platforms/services requiring cookies to operate. In the case of the MLG SOFTWARE, cookies are essential for the Platform to function properly for the provision of education. If these cookies are disabled, a large subset of the functionality provided by the Platform will either be unavailable or cease to work as expected. The MLG SOFTWARE do not capture non-essential activities such as menu items and listings you click on or pages viewed.

Performance Cookies

Performance cookies are used to analyse how visitors use a website in order to provide a better user experience.

Save Settings