Forrest Classification Calculator- Free Upper GI Bleeding Risk Stratification Tool

Forrest Classification of Upper GI Bleeding – Complete Guide to Peptic Ulcer Bleeding Risk Stratification, Endoscopic Findings, and Rebleeding Risk Assessment

Upper gastrointestinal bleeding remains one of the most common and potentially life-threatening gastrointestinal emergencies encountered in clinical practice worldwide. Among the various causes of upper GI hemorrhage, peptic ulcer disease accounts for approximately 40 to 60 percent of all cases, making it the single most prevalent etiology. The ability to rapidly and accurately assess the severity of peptic ulcer bleeding during endoscopy is critical for determining the appropriate management strategy and predicting patient outcomes. The Forrest classification, first introduced in 1974 by J.A. Forrest and colleagues in a landmark publication in The Lancet, provides a standardized endoscopic scoring system that categorizes bleeding peptic ulcers based on their visual appearance during endoscopy. This classification has become the cornerstone of risk stratification in nonvariceal upper gastrointestinal bleeding management worldwide.

The Forrest classification system divides peptic ulcer bleeding into three main categories based on the endoscopic appearance: active bleeding (Forrest I), signs of recent hemorrhage (Forrest II), and lesions without signs of recent bleeding (Forrest III). Each main category is further subdivided to provide more granular risk assessment, resulting in six distinct classifications that guide clinical decision-making regarding the need for endoscopic intervention, intensity of monitoring, and pharmacological management. Understanding these categories and their associated rebleeding risks is essential for gastroenterologists, emergency physicians, surgeons, and all healthcare professionals involved in the management of upper GI bleeding.

History and Development of the Forrest Classification

The Forrest classification was originally proposed by J.A. Forrest, N.D. Finlayson, and D.J. Shearman in their seminal 1974 paper titled “Endoscopy in Gastrointestinal Bleeding,” published in The Lancet. At the time of its introduction, endoscopy was still an evolving diagnostic tool, and there was no standardized terminology for describing the appearance of bleeding peptic ulcers. The classification was initially developed to unify the description of ulcer bleeding for better communication among endoscopists, creating a common language that could be used across institutions and clinical trials.

Over the subsequent five decades, the Forrest classification has evolved beyond its original descriptive purpose to become a powerful prognostic tool. Numerous clinical studies have validated its utility in predicting the risk of rebleeding, need for surgical intervention, length of hospital stay, and mortality. International guidelines from organizations including the European Society of Gastrointestinal Endoscopy (ESGE), the American College of Gastroenterology (ACG), and the Asia-Pacific Working Group now incorporate the Forrest classification as a fundamental component of their recommendations for managing nonvariceal upper GI bleeding. The classification remains remarkably relevant despite being nearly fifty years old, a testament to its clinical validity and practical utility.

The Six Categories of the Forrest Classification

The Forrest classification system organizes peptic ulcer bleeding into six distinct categories across three main groups. Understanding the precise endoscopic appearance of each category is fundamental to proper classification and appropriate clinical management.

Forrest Ia – Active Spurting Arterial Bleeding

Forrest Ia represents the most severe form of peptic ulcer bleeding, characterized by a visible arterial jet of blood spurting from the ulcer base. This finding indicates direct arterial erosion with ongoing high-pressure hemorrhage. On endoscopy, the clinician observes a pulsatile stream of bright red blood emanating from a specific point within the ulcer crater. This is the highest-risk lesion, associated with a rebleeding rate of approximately 55 to 90 percent if left untreated. Forrest Ia lesions require immediate endoscopic hemostasis and typically warrant intensive care unit admission with high-dose intravenous proton pump inhibitor therapy.

Forrest Ib – Active Oozing Bleeding

Forrest Ib lesions demonstrate active but non-spurting hemorrhage, presenting as a diffuse ooze of blood from the ulcer surface without a clear arterial jet. The bleeding appears as a slow, continuous flow of blood from the ulcer base or margins. While less dramatic than the spurting hemorrhage of Forrest Ia, oozing lesions still represent active bleeding that typically requires endoscopic therapy. The untreated rebleeding rate for Forrest Ib lesions ranges from approximately 10 to 55 percent. Recent Doppler probe studies have shown that Forrest Ib lesions have a significantly lower rate of underlying arterial flow compared to Forrest Ia lesions, which partly explains the lower rebleeding risk.

Forrest IIa – Non-Bleeding Visible Vessel

A Forrest IIa lesion presents as a raised, protuberant discoloration at the ulcer base, representing a non-bleeding visible vessel. The vessel appears as a small, mound-like protrusion that may be red, blue, purple, or even white in color. Despite the absence of active bleeding at the time of endoscopy, the exposed vessel wall carries a high risk of rebleeding, estimated at 43 to 50 percent without treatment. This finding indicates that a significant artery has been eroded and remains vulnerable to re-hemorrhage. Current guidelines strongly recommend endoscopic hemostasis for Forrest IIa lesions.

Forrest IIb – Adherent Clot

Forrest IIb classification is assigned when a blood clot is found firmly adherent to the ulcer base and cannot be dislodged by gentle irrigation or washing. The adherent clot obscures the underlying ulcer surface, making it impossible to determine whether a visible vessel or active bleeding source exists beneath. The rebleeding risk for Forrest IIb lesions ranges from approximately 22 to 35 percent without intervention. Management of these lesions has been debated; current guidelines recommend vigorous irrigation and attempted clot removal followed by treatment of the underlying lesion if high-risk stigmata are identified beneath the clot.

Forrest IIc – Flat Pigmented Spot

A Forrest IIc lesion is characterized by a flat, pigmented (hematin-covered) spot on the ulcer base without any raised component. The spot may appear dark brown, black, or coffee-ground colored, representing degraded hematin from previous bleeding. This finding suggests that bleeding has occurred recently but has stopped spontaneously and the source has begun to heal. The rebleeding risk is low, typically ranging from 0 to 13 percent, and most guidelines recommend conservative medical management without endoscopic therapy for these lesions.

Forrest III – Clean Ulcer Base

Forrest III represents an ulcer with a clean, fibrin-covered base and no visible signs of recent hemorrhage. The ulcer crater is smooth and typically white or pale yellow in appearance, with no evidence of blood, clot, or pigmented spots. This is the lowest-risk category with a rebleeding rate of approximately 0 to 5 percent. Patients with Forrest III lesions generally do not require endoscopic intervention and can often be managed with oral proton pump inhibitor therapy and early discharge planning.

Risk Stratification and Clinical Significance

The primary clinical utility of the Forrest classification lies in its ability to stratify patients into risk categories that directly inform management decisions. Based on extensive clinical evidence, the six Forrest categories are commonly grouped into high-risk and low-risk designations. High-risk lesions include Forrest Ia (active spurting), Forrest Ib (active oozing), and Forrest IIa (non-bleeding visible vessel), all of which carry substantial rebleeding risk and typically require endoscopic hemostasis. Forrest IIb (adherent clot) occupies an intermediate position, with management depending on whether high-risk stigmata are found beneath the clot after removal. Low-risk lesions include Forrest IIc (flat pigmented spot) and Forrest III (clean base), which generally do not benefit from endoscopic therapy.

A landmark study by de Groot and colleagues reassessed the predictive value of the Forrest classification using prospective registry data from 397 patients. Their analysis found that rebleeding occurred in approximately 18.6 percent of all patients, with the highest rebleeding rate (59 percent) observed in Forrest Ia lesions. Interestingly, the odds ratios for rebleeding among Forrest Ib through IIc lesions were similar, leading the authors to propose a simplified three-tier classification: high risk (Forrest Ia), increased risk (Forrest Ib through IIc), and low risk (Forrest III). This simplified approach maintained comparable predictive accuracy while potentially improving clinical practicality.

Treatment Recommendations by Forrest Category

Current international guidelines provide specific treatment recommendations based on the Forrest classification. For Forrest Ia and Ib lesions with active bleeding, immediate endoscopic hemostasis is required. The preferred approach is dual-modality therapy combining mechanical hemostasis (hemoclips) or thermal coagulation with epinephrine injection. Epinephrine injection alone should not be used as monotherapy for high-risk lesions, as evidence consistently shows that combination therapy is superior in preventing rebleeding.

For Forrest IIa lesions with a non-bleeding visible vessel, endoscopic hemostasis is strongly recommended. Studies have demonstrated that endoscopic therapy for these lesions significantly reduces rebleeding rates from approximately 43 to 50 percent to less than 15 percent. The treatment approach is similar to that used for active bleeding, typically employing thermal coagulation, hemoclips, or combination therapy.

Management of Forrest IIb lesions (adherent clot) involves vigorous irrigation to attempt clot removal, followed by treatment of the underlying stigmata if a visible vessel or active bleeding is identified. If the clot can be removed and only a flat pigmented spot (IIc) or clean base (III) is found beneath, endoscopic therapy may not be necessary. If the clot cannot be dislodged, some experts recommend treating the lesion as a high-risk stigma.

For Forrest IIc and III lesions, endoscopic hemostasis is generally not recommended. These patients can typically be managed with proton pump inhibitor therapy and may be candidates for early discharge or outpatient follow-up. However, clinical context including hemodynamic stability, comorbidities, and anticoagulant use should always be considered when making discharge decisions.

Proton Pump Inhibitor Therapy and Forrest Classification

Proton pump inhibitor (PPI) therapy plays a crucial adjunctive role in the management of peptic ulcer bleeding, and the Forrest classification helps guide the intensity of PPI treatment. For high-risk lesions (Forrest Ia, Ib, IIa, and selected IIb), international guidelines recommend high-dose intravenous PPI therapy following successful endoscopic hemostasis. The standard regimen typically consists of an intravenous bolus of 80 mg esomeprazole or pantoprazole followed by a continuous infusion of 8 mg per hour for 72 hours. This aggressive acid suppression strategy maintains gastric pH above 6, promoting clot stability and facilitating ulcer healing.

For low-risk lesions (Forrest IIc and III), high-dose intravenous PPI therapy is generally not required. Oral PPI therapy at standard or double doses is typically sufficient. Interestingly, a large international randomized trial found that PPI therapy did not significantly reduce rebleeding rates among patients with Forrest Ib lesions after successful endoscopic hemostasis, suggesting that the rebleeding risk for oozing lesions may be lower than traditionally thought once hemostasis is achieved.

Inter-Observer Agreement and Reliability

The clinical utility of any classification system depends on its reproducibility among different observers. Studies examining inter-rater agreement for the Forrest classification have generally shown good concordance. A study by Yen and colleagues found good intra-rater correlation (0.92 to 0.98) and good inter-rater correlation (0.639 to 0.859) among endoscopists of different experience levels. Notably, the correlation was higher among pairs of tutor and junior endoscopists than among experienced endoscopists alone, suggesting that formal training in classification may improve consistency.

Despite generally good agreement, some diagnostic categories pose greater challenges than others. Distinguishing between Forrest Ib (active oozing) and Forrest IIa (visible vessel with oozing from the surrounding base) can be difficult in practice. Similarly, differentiating a flat pigmented spot (IIc) from a small visible vessel (IIa) requires careful inspection and experience. These ambiguities highlight the importance of thorough training and standardized protocols for endoscopic assessment.

Comparison with Other Scoring Systems

While the Forrest classification focuses exclusively on the endoscopic appearance of the ulcer, several other scoring systems incorporate clinical and laboratory parameters to provide a more comprehensive risk assessment. The Rockall score, developed in 1996, combines both pre-endoscopic factors (age, comorbidity, hemodynamic status) and endoscopic findings (including the Forrest classification) to predict rebleeding and mortality risk. The Glasgow-Blatchford score uses clinical and laboratory parameters (hemoglobin, blood urea nitrogen, systolic blood pressure, heart rate, presence of melena or syncope, and comorbidities) to predict the need for clinical intervention, and it can be calculated before endoscopy.

The Forrest classification is best used in conjunction with these clinical scoring systems rather than in isolation. A comprehensive risk assessment that combines endoscopic findings with clinical parameters provides the most accurate prediction of patient outcomes. The Rockall score in particular integrates the Forrest classification as one of its components, creating a hybrid approach that leverages both endoscopic and clinical data.

Validation Across Diverse Populations

The Forrest classification has been studied and validated in diverse patient populations across multiple continents. Originally developed in a Scottish cohort, the classification has been subsequently evaluated in studies from North America, Europe, East Asia, Southeast Asia, the Middle East, and Oceania. Large multicenter trials, including the landmark study by Sung and colleagues on intravenous esomeprazole for preventing recurrent peptic ulcer bleeding, enrolled patients from 91 sites across 16 countries, demonstrating the global applicability of the classification system.

Some population-specific considerations do exist. The prevalence of different Forrest categories may vary across populations due to differences in Helicobacter pylori infection rates, NSAID use patterns, anticoagulant prescribing practices, and time to endoscopy. In populations with earlier access to endoscopy, a higher proportion of active bleeding lesions (Forrest I) may be observed, while delayed endoscopy may result in more patients presenting with signs of recent hemorrhage (Forrest II) or clean bases (Forrest III). Despite these epidemiological variations, the prognostic significance of each Forrest category appears to be consistent across populations.

Limitations of the Forrest Classification

While the Forrest classification remains an indispensable clinical tool, it has several recognized limitations. First, the classification was specifically designed for peptic ulcer bleeding and may not be directly applicable to other causes of upper GI hemorrhage such as Mallory-Weiss tears, Dieulafoy lesions, or tumor bleeding. Although clinicians sometimes extrapolate the classification to these conditions, the prognostic accuracy may differ from what has been established for peptic ulcers.

Second, the Forrest classification represents a snapshot in time and does not account for the dynamic nature of ulcer bleeding. A lesion classified as Forrest IIc at the time of initial endoscopy may have been actively bleeding shortly before the procedure. Conversely, a Forrest III clean-base ulcer may progress to rebleeding due to underlying patient factors not captured by the endoscopic appearance alone.

Third, the classification does not incorporate patient-level risk factors such as hemodynamic instability, anticoagulant use, comorbidities, or ulcer size and location, all of which independently influence rebleeding risk and mortality. This limitation underscores the importance of integrating the Forrest classification with comprehensive clinical assessment tools.

Fourth, inter-observer variability, while generally acceptable, can lead to misclassification in borderline cases, particularly between adjacent categories. The distinction between some categories (such as IIa and IIc, or IIb and IIa) can be subtle and observer-dependent.

Endoscopic Hemostasis Techniques for High-Risk Lesions

When the Forrest classification identifies a high-risk lesion requiring endoscopic therapy, several techniques are available. Injection therapy with diluted epinephrine (typically 1:10,000 or 1:20,000) achieves initial hemostasis through a combination of tamponade, vasoconstriction, and platelet aggregation effects. However, as noted previously, injection therapy alone has unacceptably high rebleeding rates and should always be combined with a second modality for high-risk lesions.

Thermal coagulation methods include bipolar or multipolar electrocoagulation, heater probe application, and argon plasma coagulation. These techniques achieve hemostasis by causing tissue coagulation and vessel sealing through direct application of thermal energy. Mechanical hemostasis using hemoclips (through-the-scope clips) physically compresses the bleeding vessel, providing immediate and often definitive hemostasis. Newer hemostatic modalities include topical hemostatic powders and over-the-scope clips for refractory bleeding or difficult-to-access lesions.

Role of Second-Look Endoscopy

The role of routine second-look endoscopy after initial successful hemostasis remains debated. Some studies have suggested that scheduled repeat endoscopy at 24 to 48 hours may identify lesions that have deteriorated in Forrest classification or developed new high-risk features. However, routine second-look endoscopy is not universally recommended by current guidelines. Instead, clinicians should consider repeat endoscopy based on clinical indicators of rebleeding, such as recurrent hematemesis, melena, hemodynamic instability, or a significant drop in hemoglobin.

A study examining Forrest IIc lesions at second-look endoscopy found that patients with Rockall scores of 6 or higher had significantly increased rebleeding risk even when their initial high-risk stigmata had faded to IIc after treatment. This finding suggests that clinical risk factors can modify the prognostic significance of the Forrest classification, further supporting an integrated approach to risk assessment.

Emerging Technologies and Future Directions

Artificial intelligence and machine learning are being applied to improve the accuracy and consistency of Forrest classification. Computer-aided detection systems using deep learning algorithms can analyze endoscopic images in real time and suggest the appropriate Forrest category, potentially reducing inter-observer variability and improving classification accuracy, particularly for less experienced endoscopists. Early studies have shown promising results, with AI systems achieving accuracy comparable to or exceeding that of experienced clinicians.

Advanced imaging technologies such as magnification endoscopy, narrow-band imaging, and linked color imaging may provide enhanced visualization of ulcer stigmata, potentially improving the precision of Forrest classification. Additionally, endoscopic Doppler probes can detect arterial blood flow beneath the ulcer surface, providing hemodynamic information that complements the visual assessment of the Forrest classification.

Special Considerations in Clinical Practice

Several practical considerations affect the application of the Forrest classification in real-world clinical settings. Timing of endoscopy significantly influences the distribution of Forrest categories observed. Early endoscopy (within 12 to 24 hours of presentation) is more likely to detect active bleeding or high-risk stigmata, while delayed endoscopy may result in a higher proportion of low-risk findings as bleeding resolves spontaneously in many patients.

The presence of blood or clots in the upper GI tract can obscure the ulcer surface and make accurate classification difficult. Thorough lavage and aspiration before detailed inspection are essential for proper assessment. In some cases, the endoscopist may need to use repositioning techniques or even a second endoscope to achieve adequate visualization.

Anticoagulant and antiplatelet medications significantly impact bleeding risk and should be considered alongside the Forrest classification when making management decisions. Patients on dual antiplatelet therapy or direct oral anticoagulants may have elevated rebleeding risk even with low-risk Forrest categories, necessitating more cautious management and potentially longer periods of observation.

Global Application and Practice Variation

Despite the widespread adoption of the Forrest classification in clinical guidelines worldwide, significant practice variation exists in its application. A Canadian registry of patients with upper gastrointestinal bleeding reported that only 47.8 percent of patients with high-risk stigmata underwent endoscopic therapy, while 9.8 percent of those with low-risk stigmata received endoscopic therapy. This finding reveals a gap between guideline recommendations and real-world practice that may result from misclassification, logistical constraints, or differing clinical judgment.

Efforts to improve guideline adherence include structured endoscopy reporting templates that prompt clinicians to document the Forrest classification, quality improvement programs that track endoscopic hemostasis rates for high-risk lesions, and training programs that emphasize standardized assessment of ulcer stigmata. Digital photo-documentation of endoscopic findings is also recommended to allow for retrospective review and quality assurance.

Understanding Rebleeding Risk Factors

Beyond the Forrest classification itself, several independent risk factors for rebleeding have been identified in clinical studies. These include advanced age (particularly over 65 years), hemodynamic instability at presentation, hemoglobin level below 10 g/dL, large ulcer size (greater than 2 cm), posterior duodenal bulb or high gastric body location, concurrent use of NSAIDs or anticoagulants, and significant comorbidities. The presence of multiple risk factors in combination with a high-risk Forrest classification substantially increases the probability of adverse outcomes.

The Forrest classification provides the endoscopic component of this multifactorial risk assessment. When a high-risk Forrest lesion is identified in a patient with additional clinical risk factors, the cumulative rebleeding risk may exceed that predicted by the Forrest category alone. This is particularly relevant for patients with Forrest IIb (adherent clot) or IIc (flat pigmented spot) lesions who have high Rockall scores, as these patients may benefit from more aggressive management than would typically be recommended based on the endoscopic findings alone.

Impact on Hospital Resource Utilization

The Forrest classification has significant implications for hospital resource utilization and patient disposition. Patients with high-risk Forrest lesions (Ia, Ib, IIa) typically require intensive care unit admission, continuous hemodynamic monitoring, and intravenous PPI infusion for at least 72 hours. In contrast, patients with low-risk Forrest lesions (IIc, III) who are hemodynamically stable may be candidates for early discharge or even outpatient management with oral PPI therapy and follow-up.

Proper application of the Forrest classification can therefore optimize resource allocation by ensuring that high-intensity monitoring and intervention are directed toward patients most likely to benefit, while avoiding unnecessary hospitalization and procedures for low-risk patients. This risk-stratified approach has been shown to be cost-effective in multiple healthcare systems worldwide.

Frequently Asked Questions

Conclusion

The Forrest classification remains the gold standard for endoscopic risk stratification of peptic ulcer bleeding, providing a simple yet powerful framework for guiding clinical management decisions. Since its introduction in 1974, the classification has been extensively validated across diverse populations and clinical settings worldwide, demonstrating consistent prognostic value for predicting rebleeding risk, need for intervention, and patient outcomes. By categorizing ulcer bleeding into six distinct categories from active spurting hemorrhage to clean ulcer base, the system enables clinicians to rapidly identify patients who require immediate endoscopic intervention and those who can be safely managed conservatively.

While the Forrest classification has limitations, including its focus on endoscopic appearance alone and potential inter-observer variability, it remains most valuable when integrated with comprehensive clinical risk assessment tools such as the Rockall and Glasgow-Blatchford scores. Emerging technologies including artificial intelligence and advanced imaging techniques promise to further enhance the accuracy and clinical utility of the Forrest classification in the years ahead. Healthcare professionals managing upper GI bleeding should maintain proficiency in the Forrest classification system and apply it consistently within the framework of current evidence-based guidelines to optimize patient outcomes.