About the Forrest Classification of Peptic Ulcer Bleeding
The Forrest classification describes what a bleeding peptic ulcer looks like at endoscopy and maps that appearance to the risk of rebleeding. Class I is active bleeding (Ia spurting, Ib oozing); class II shows stigmata of recent haemorrhage (IIa non-bleeding visible vessel, IIb adherent clot, IIc flat pigmented spot); class III is a clean ulcer base. Rebleeding risk without endoscopic therapy falls from as high as 55–90% for a spurting vessel and 40–50% for a visible vessel down to roughly 5% for a clean base. High-risk stigmata — Ia, Ib and IIa — should be treated endoscopically; IIc and III need none. It is a descriptive classification, not an additive score, and it predicts rebleeding rather than death.
Interpreting the result
Read the class as a rebleeding risk and a treatment instruction in one. Ia, Ib and IIa are high-risk stigmata: apply endoscopic haemostasis — combination therapy or a mechanical/thermal method rather than adrenaline alone — and give high-dose intravenous proton pump inhibitor afterwards. IIb (adherent clot) is intermediate and handled according to local practice, either clot removal and treatment of what lies beneath or PPI therapy alone. IIc and III are low risk, need no endoscopic therapy, and support oral PPI, early feeding and early discharge. Across all classes, test for and eradicate Helicobacter pylori and review NSAID and antiplatelet use.
| Score | Band | What it means | Action |
|---|---|---|---|
| Ia | Active spurting — high risk | Rebleeding ~55–90% without endoscopic therapy | Immediate endoscopic haemostasis + high-dose IV PPI; admit and monitor |
| Ib | Active oozing — high risk | Rebleeding ~50% without endoscopic therapy | Endoscopic haemostasis + high-dose IV PPI; admit and monitor |
| IIa | Non-bleeding visible vessel — high risk | Rebleeding ~40–50% without endoscopic therapy | Treat endoscopically despite no active bleeding; high-dose IV PPI |
| IIb | Adherent clot — intermediate risk | Rebleeding ~20–30% without endoscopic therapy | Consider clot removal and treat the base, or high-dose PPI alone; admit |
| IIc | Flat pigmented spot — low risk | Rebleeding ~7–10% | No endoscopic therapy; oral PPI, early feeding, consider discharge |
| III | Clean base — low risk | Rebleeding ~3–5% | No endoscopic therapy; oral PPI and early discharge usually appropriate |
What the Forrest Classification needs (6 inputs)
- Class Ia — spurting arterial bleeding
- Active, pulsatile arterial haemorrhage. The highest-risk appearance and an unambiguous indication for immediate endoscopic haemostasis.
- Class Ib — oozing bleeding
- Active non-pulsatile oozing from the ulcer base without a discrete spurting point. Still active bleeding and still treated.
- Class IIa — non-bleeding visible vessel
- A protuberant, pigmented or pearl-like vessel in the ulcer base that is not currently bleeding. High rebleeding risk, so it is treated despite the absence of active bleeding.
- Class IIb — adherent clot
- A clot adherent to the base that resists vigorous washing. Intermediate risk; management is debated between clot removal with treatment of the underlying lesion and intensive PPI therapy alone.
- Class IIc — flat pigmented spot
- A flat red or black spot in the ulcer base — a low-risk stigma that does not require endoscopic therapy.
- Class III — clean ulcer base
- No stigmata of recent haemorrhage. The lowest-risk appearance; no endoscopic treatment needed and early discharge is usually appropriate.
What it returns
- Forrest class
- The single descriptive category that best matches what was seen at endoscopy.
- Rebleeding risk band
- High (Ia, Ib, IIa), intermediate (IIb) or low (IIc, III), which drives whether endoscopic therapy is indicated.
How it is calculated
Forrest is a lookup, not a calculation: the endoscopist selects the appearance that best fits the ulcer, and that category carries an approximate untreated rebleeding risk drawn from decades of natural-history and trial data. The classification splits ulcers into three tiers — actively bleeding (I), recently bled with stigmata (II), and clean (III) — and the class II tier is further graded because a visible vessel, an adherent clot and a flat spot behave very differently. The clinical rule follows directly: the higher tiers are treated endoscopically and the lower ones are not.
Facts & figures
| Class | Endoscopic appearance | Rebleeding (untreated) | Endoscopic therapy? |
|---|---|---|---|
| Ia | Spurting arterial bleeding | ~55–90% | Yes |
| Ib | Oozing bleeding | ~50% | Yes |
| IIa | Non-bleeding visible vessel | ~40–50% | Yes |
| IIb | Adherent clot | ~20–30% | Consider |
| IIc | Flat pigmented spot | ~7–10% | No |
| III | Clean base | ~3–5% | No |
Percentages are the historical untreated risks reported across natural-history series; endoscopic therapy and high-dose PPI reduce them substantially.
Evidence
Original description
1974Forrest, Finlayson and Shearman proposed the classification in 1974 to standardise the endoscopic description of gastrointestinal bleeding lesions, distinguishing active bleeding, stigmata of recent haemorrhage and clean lesions.
The scheme's enduring value is that the appearance it captures is the dominant predictor of rebleeding, later quantified across many cohorts and used to define which lesions benefit from endoscopic therapy.
Predictors of recurrent haemorrhage — systematic review
2008Systematic review of the predictors of recurrent bleeding after endoscopic haemostatic therapy for bleeding peptic ulcers, pooling the stigmata-specific rebleeding rates.
Active bleeding and a non-bleeding visible vessel carried the highest rebleeding risk after therapy, confirming the class I / IIa grouping as the high-risk tier that justifies treatment.
Natural history and management review
1994Foundational clinical review synthesising the natural history of bleeding peptic ulcers by endoscopic stigma and the effect of endoscopic therapy.
Reported the classic untreated rebleeding gradient — from a clean base at roughly 5% up to an actively spurting vessel at the highest risk — that Forrest classes summarise.
How it compares
Forrest Classification vs Rockall Score
Forrest predicts rebleeding from the endoscopic appearance; Rockall predicts mortality by adding age, shock and comorbidity to the endoscopic findings — use them together after endoscopy.
The full Rockall score actually incorporates the same endoscopic stigmata that Forrest describes, but wraps them in patient-level risk factors to estimate death rather than rebleeding. In practice the endoscopist assigns a Forrest class to drive the immediate treatment decision, and the Rockall score is calculated for prognosis and disposition. They are complementary readings of the same endoscopy.
Forrest Classification vs Glasgow-Blatchford Score
Glasgow-Blatchford is a pre-endoscopy triage score with no endoscopic input; Forrest is an at-endoscopy classification — they operate at different moments and do not compete.
Glasgow-Blatchford decides, before any scope, who needs admission or intervention and who might be discharged. Forrest is only available once the ulcer has been seen, and it decides whether to treat it and how closely to watch afterwards. A patient can have a high Glasgow-Blatchford score and then a reassuring Forrest III clean base, or vice versa.
Pearls & pitfalls
- Forrest predicts rebleeding, not mortality. For a mortality estimate use the Rockall score, which adds age and comorbidity.
- Class IIb (adherent clot) is the contentious one: vigorous washing may expose a treatable Ia/Ib/IIa lesion, and practice varies between clot removal with treatment and high-dose PPI alone.
- A 'visible vessel' (IIa) is a stigma to treat even though nothing is actively bleeding — its untreated rebleeding risk is high.
- Interobserver agreement is only moderate, particularly in telling an adherent clot from a flat spot; when in doubt, treat the higher-risk interpretation.
- The classification applies to peptic ulcers; do not force varices, Dieulafoy lesions or Mallory-Weiss tears into it.
- The quoted rebleeding percentages are untreated natural-history figures — endoscopic therapy plus high-dose PPI markedly lowers them, so they describe risk, not the expected outcome after treatment.
Critical actions
- Apply endoscopic haemostasis to Forrest Ia, Ib and IIa lesions, using a mechanical or thermal method combined with adrenaline rather than adrenaline injection alone.
- Give high-dose intravenous proton pump inhibitor therapy after treating a high-risk stigma.
- For an adherent clot (IIb), decide between clot removal with treatment of the underlying lesion and intensive PPI therapy, per local protocol, and admit for monitoring.
- For IIc and III lesions, withhold endoscopic therapy and consider early feeding and discharge on oral PPI.
- In every case, test for Helicobacter pylori and eradicate if positive, and review NSAID, aspirin and anticoagulant use.
Why this score exists
The original 1974 paper set out to bring order to how endoscopists described bleeding lesions, at a time when flexible endoscopy was new and its findings were recorded inconsistently. The lasting insight is that the look of the ulcer — spurting, oozing, a naked vessel, a clot, a spot, or nothing — encodes its future behaviour, so a shared descriptive language is also a shared risk language. Everything that followed, from the visible-vessel treatment rule to high-dose PPI protocols, is built on that classification.
About the creator
First author, 1974 classification
Published the endoscopic classification of bleeding peptic ulcer stigmata still used to guide haemostatic therapy. Not to be confused with Ewan H. Forrest, who derived the Glasgow Alcoholic Hepatitis Score.
Limitations
- It is descriptive and subjective, with only moderate interobserver agreement for the intermediate stigmata.
- It predicts rebleeding but not mortality, and takes no account of age, comorbidity or haemodynamics.
- The rebleeding percentages attached to each class are historical, untreated figures and vary between series.
- It applies to peptic ulcers only, and does not describe other bleeding sources.
- The management of the adherent-clot class (IIb) remains genuinely uncertain, so the classification does not settle every treatment decision.
If you are the patient
During an endoscopy for a bleeding stomach or duodenal ulcer, the doctor looks closely at the ulcer and puts its appearance into one of the Forrest categories, from a vessel that is actively spurting (highest risk of bleeding again) through to a clean-looking base (lowest risk). This tells the team whether to treat the ulcer there and then — for example by sealing a vessel with a clip or heat and injecting medication — and how carefully to watch you afterwards. High-risk appearances are treated and usually need a night or two of monitoring; low-risk ones often need no treatment and mean you can eat and go home sooner. It describes the ulcer's risk of bleeding again, rather than any other kind of risk, and it usually comes with tests for the Helicobacter pylori bug and a review of any anti-inflammatory or blood-thinning medicines.
Frequently asked questions
What is the Forrest classification?#
It is an endoscopic classification of bleeding peptic ulcers that groups them by appearance — active spurting (Ia) or oozing (Ib) bleeding, a non-bleeding visible vessel (IIa), an adherent clot (IIb), a flat pigmented spot (IIc), or a clean base (III) — and links each to a rebleeding risk and a treatment decision.
Which Forrest classes need endoscopic treatment?#
The high-risk stigmata — Forrest Ia, Ib and IIa — should be treated with endoscopic haemostasis and high-dose proton pump inhibitor therapy. IIc and III are low risk and need no endoscopic therapy. IIb (adherent clot) is intermediate and managed according to local protocol.
What is the rebleeding risk for each Forrest class?#
Without endoscopic therapy, approximate rebleeding risks are ~55–90% for Ia, ~50% for Ib, ~40–50% for IIa, ~20–30% for IIb, ~7–10% for IIc and ~3–5% for III. These are untreated natural-history figures; treatment lowers them considerably.
Does the Forrest classification predict death?#
No — it predicts rebleeding, not mortality. To estimate mortality after an upper GI bleed, use the Rockall score, which combines the endoscopic findings with age, shock and comorbidity.
What is a Forrest IIa lesion?#
A non-bleeding visible vessel: a protuberant, pigmented vessel in the ulcer base that is not currently bleeding. Despite the absence of active bleeding it carries a high rebleeding risk (about 40–50% untreated) and is treated endoscopically.
Why is the adherent clot (IIb) controversial?#
Because washing the clot away may reveal a higher-risk lesion that benefits from treatment, but doing so can also provoke bleeding. Some endoscopists remove the clot and treat what is underneath, while others manage it with high-dose PPI therapy alone; the evidence does not clearly favour one approach.
References
Original / primary reference
Validation and evidence
- Elmunzer BJ, Young SD, Inadomi JM, Schoenfeld P, Laine L. Systematic review of the predictors of recurrent hemorrhage after endoscopic hemostatic therapy for bleeding peptic ulcers. Am J Gastroenterol. 2008;103(10):2625-2632.
- Laine L, Peterson WL. Bleeding peptic ulcer. N Engl J Med. 1994;331(11):717-727.
Clinical practice guidelines
- Laine L, Barkun AN, Saltzman JR, Martel M, Leontiadis GI. ACG Clinical Guideline: Upper Gastrointestinal and Ulcer Bleeding. Am J Gastroenterol. 2021;116(5):899-917.
- Gralnek IM, Stanley AJ, Morris AJ, et al. Endoscopic diagnosis and management of nonvariceal upper gastrointestinal hemorrhage (NVUGIH): ESGE Guideline – Update 2021. Endoscopy. 2021;53(3):300-332.
- Barkun AN, Almadi M, Kuipers EJ, et al. Management of Nonvariceal Upper Gastrointestinal Bleeding: Guideline Recommendations From the International Consensus Group. Ann Intern Med. 2019;171(11):805-822.