Grading of Gastrointestinal Bleeding: Classification and Scoring Systems
Upper Gastrointestinal Bleeding Classification
The Forrest classification is the gold standard endoscopic grading system for peptic ulcer bleeding, with Forrest Ia (active spurting) carrying the highest rebleeding risk at 59% and mandating immediate dual-modality endoscopic hemostasis. 1
Forrest Classification Categories
The Forrest system stratifies ulcers into three main categories based on endoscopic appearance:
Forrest I (Active Bleeding):
Forrest II (Signs of Recent Hemorrhage):
- IIa: Non-bleeding visible vessel - high risk 1
- IIb: Adherent clot
- IIc: Flat pigmented spot
Forrest III: Clean-based ulcer - lowest risk 1
Endoscopic hemostasis is strongly recommended for Forrest Ia, Ib, and IIa lesions due to high risk of persistent bleeding or rebleeding. 1 Dual modality therapy (mechanical therapy combined with epinephrine injection) is preferred over single modality for these high-risk lesions. 1
Therapeutic Implications by Forrest Grade
- Forrest Ia/Ib/IIa: Require immediate endoscopic therapy with bipolar electrocoagulation, heater probe, clips, or absolute ethanol injection 2
- Forrest IIb: May benefit from clot removal and underlying vessel treatment 2
- Forrest IIc/III: Generally do not require endoscopic therapy but need PPI therapy 2
Clinical Scoring Systems for Upper GI Bleeding
Glasgow-Blatchford Score (GBS) - Pre-Endoscopic Risk Stratification
The Glasgow-Blatchford score of 1 or less identifies patients at very low risk for rebleeding or mortality who may be safely discharged without hospitalization or inpatient endoscopy. 3
The GBS incorporates:
- Blood urea nitrogen
- Hemoglobin level
- Systolic blood pressure
- Heart rate
- Presence of melena or syncope
- Hepatic disease or cardiac failure 3
The GBS demonstrates superior sensitivity (0.99) for detecting high-risk patients compared to other pre-endoscopic scores, making it the preferred tool for emergency department risk stratification. 3 Implementation of GBS-based discharge protocols has proven effective in reducing unnecessary hospitalizations without compromising safety. 3
Rockall Score - Combined Pre- and Post-Endoscopic Assessment
The Rockall scoring system exists in two forms:
- Pre-endoscopic Rockall: Uses age, shock, and comorbidities 3
- Full Rockall: Adds endoscopic diagnosis and stigmata of recent hemorrhage 4
The full Rockall score (incorporating endoscopic findings) predicts mortality with similar accuracy to AIMS65 (AUROC 0.78), while pre-endoscopic Rockall has lower predictive value and cannot be recommended for discharge decisions. 3, 5
Patients with Rockall score >6 have significantly higher mortality risk and require intensive monitoring. 4
AIMS65 Score
The AIMS65 score should NOT be used to identify low-risk patients suitable for discharge, as it has inferior sensitivity (0.78-0.82) compared to GBS. 3
However, AIMS65 demonstrates superior performance for predicting:
The AIMS65 components are:
- Albumin <3.0 g/dL
- INR >1.5
- Mental status alteration
- Systolic BP ≤90 mmHg
- Age 65 years or older 5
Lower Gastrointestinal Bleeding Classification
Oakland Score - Risk Stratification for Lower GI Bleeding
The Oakland score should be calculated immediately for all stable lower GI bleeding patients to guide disposition, incorporating age, gender, previous LGIB admission, digital rectal examination findings, heart rate, systolic BP, and hemoglobin. 6
- Oakland score ≤8: Low-risk patients potentially suitable for outpatient management 6
- Oakland score >8: Major bleeding requiring colonoscopy as primary diagnostic/therapeutic modality 6
BLEED Classification for Lower GI Bleeding
The BLEED classification identifies high-risk lower GI bleeding patients based on:
- Bleeding (ongoing)
- Low systolic blood pressure
- Elevated prothrombin time
- Erratic mental status
- Disease (unstable comorbid conditions) 6
Additional high-risk features include heart rate >100/min, syncope, bleeding during first 4 hours of evaluation, aspirin use, and >2 active comorbidities. 6
Endoscopic Stigmata in Diverticular Bleeding
The presence of a visible vessel or adherent clot within a diverticulum reliably indicates severe hemorrhage requiring intervention, while a clean-based ulcer suggests low rebleeding risk permitting early discharge. 3, 7
Gastric Variceal Bleeding Classification
Recommended Endoscopic Classification
For gastric varices, a simplified classification system is recommended over the traditional Sarin classification, categorizing varices as cardiofundal GV, lesser curve GV, or distal GV based on vascular supply patterns. 3
- Cardiofundal GV: Located on posterior/greater curvature of cardia, distinct vascular supply from esophageal varices 3
- Lesser curve GV: Similar vascular supply to esophageal varices, managed with band ligation 3
- Distal GV: Rare, often associated with splenic vein thrombosis, require different management 3
Endoscopic classification systems for gastric varices should NOT be used for primary prophylaxis decisions due to lack of validated predictive models. 3
Critical Implementation Points
Documentation Requirements
Standard terminology and classification systems must be used when documenting endoscopic findings, with proper photo-documentation essential for future comparison. 1
Hemodynamic Assessment Priority
Before applying any classification system, calculate the shock index (heart rate/systolic BP) immediately—a shock index >1 indicates hemodynamic instability and dictates the entire management pathway, overriding other scoring systems. 6
Upper vs. Lower Source Differentiation
Recognize that 11-15% of patients with brisk hematochezia and hemodynamic compromise actually have an upper GI source requiring upper endoscopy first. 6 Risk factors suggesting upper GI source include elevated BUN/creatinine ratio, antiplatelet drug use, and hemodynamic instability. 6