Validation of the new ABC score for predicting 30-day mortality in gastrointestinal bleeding

Marie Christelle Saade1, Anthony Kerbage1, Suha Jabak1

  • 1Division of Gastroenterology and Hepatology, Department of Internal Medicine, American University of Beirut Medical Center, Beirut, Lebanon.

BMC Gastroenterology
|June 21, 2022
PubMed

Insights

The ABC score effectively predicts mortality in gastrointestinal bleeding (GIB) patients. This novel scoring system outperforms existing methods for both upper and lower GIB, aiding clinical decisions.

Area of Science:

  • Gastroenterology
  • Clinical Medicine
  • Predictive Analytics

Background:

  • Gastrointestinal bleeding (GIB) poses significant mortality risks.
  • Accurate prediction of mortality is crucial for effective patient management.
  • Existing scoring systems have limitations in predicting outcomes for both upper and lower GIB.

Purpose of the Study:

  • To validate the novel ABC score for predicting one-month mortality in patients with upper and lower GIB.
  • To compare the performance of the ABC score against established scores like Oakland, AIMS-65, and Rockall.
  • To assess the utility of the ABC score in a Lebanese tertiary care center cohort.

Main Methods:

  • Adult patients with overt GIB admitted to a tertiary care center were retrospectively analyzed.
  • The ABC score's predictive performance for 30-day mortality was evaluated using Area Under the Receiver Operating Characteristic (AUROC) curves.
  • The ABC score was compared with Oakland (lower GIB) and AIMS-65/Rockall (upper GIB) scores.

Main Results:

  • The ABC score demonstrated good performance in predicting 30-day mortality for upper GIB (AUROC: 0.79), outperforming AIMS-65 (0.67) and Rockall (0.62).
  • For lower GIB, the ABC score showed comparable performance to the Oakland score (AUROC: 0.70 vs 0.56).
  • The ABC score proved superior or comparable to existing scores in this patient cohort.

Conclusions:

  • The ABC score is a valuable tool for predicting 30-day mortality in patients with both upper and lower GIB.
  • Its good performance and simplicity support its use in guiding clinical management decisions.
  • The ABC score appears reproducible and applicable across different patient populations.
Abstract