Outpatient management of low-risk patients with upper gastrointestinal bleeding: can we safely extend the Glasgow Blatchford Score in clinical practice?

Eur J Gastroenterol Hepatol. 2015 May;27(5):512-5. doi: 10.1097/MEG.0000000000000333.


Background and aim: The Glasgow Blatchford Score (GBS) is a validated prognostic score for patients presenting with upper gastrointestinal (GI) bleeding (UGIB). The score predicts the need for therapeutic intervention or death, and studies have suggested that outpatient management is safe for patients with a GBS of zero. Our aim was to assess whether we could safely extend the threshold for outpatient management to patients with GBS≤1.

Methods: Following assessment of our historical data, our UGIB protocol was changed to recommend outpatient management for patients with a GBS≤1, unless required for other reasons. Data on all patients presenting with UGIB over the following 12 months were prospectively recorded, including GBS and clinical Rockall scores. Adverse outcomes were defined by a 30-day combined endpoint of death, endotherapy, interventional radiology, surgery or transfusion. Negative predictive value (NPV) of GBS≤1 for adverse outcomes in UGIB was calculated.

Results: A total of 514 patients presented with UGIB in the 12 month study period. Of the patients, 183 (35.6%) had GBS≤1 (111, GBS=0; 72, GBS=1). Of these, 88 (48.1%) were managed as outpatients, and none had an adverse outcome. Of the 95 (51.9%) patients with GBS≤1 managed as inpatients, 80 (84.2%) had comorbidities requiring inpatient care. Within this admitted group with GBS≤1, one patient required transfusion and one died from a nongastrointestinal malignancy. GBS≤1 had an NPV of 99.45% (95% confidence interval 95.53-99.97%) in predicting adverse outcomes within 30 days.

Conclusion: GBS≤1 has a high NPV for adverse outcomes in UGIB. This suggests outpatient management of patients with UGIB and that GBS≤1 is safe in our population.

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Ambulatory Care*
  • Clinical Protocols*
  • Comorbidity
  • Female
  • Hematemesis / therapy*
  • Hospitalization
  • Humans
  • Male
  • Melena / therapy*
  • Middle Aged
  • Predictive Value of Tests
  • Prospective Studies
  • ROC Curve
  • Risk Assessment
  • Severity of Illness Index*
  • Treatment Outcome