The Value of Rockall and Forrest Classification in the Prognosis of Patients With Gastrointestinal Bleeding Applying to the Emergency Department

Abstract

Purpose Gastrointestinal system bleeding (GIB) is a common and highly morbid condition with high mortality rates frequently observed in emergency departments. Several scoring systems have been used to predict mortality and rebleeding in these cases. Among these scoring systems, Rockall Score (RS) and Forrest classification (FS) are commonly used. In our study, we aimed to evaluate the predictive power of RS for mortality, rebleeding, and the need for blood transfusion, and to provide benefits to clinicians. Methods The study was designed as a retrospective and single-center study, conducted between January 1, 2017, and January 1, 2018. Demographic data, RS and FS scores, clinical outcomes (mortality, rebleeding, blood transfusion requirement, and final outcomes) of the cases were evaluated. Results Of the 109 patients in our study, 60.6% were male, and the mean age was 67.83 ± 16.75 years. The most common presenting complaint was black stool, and the most frequently detected finding was melena. The most common endoscopic diagnosis was erosive gastritis (24.8%). The use of acetylsalicylic acid and anticoagulants was significantly higher in cases of lower gastrointestinal bleeding. There was a negative correlation between the RS score and hemoglobin level of the patients (r=-0.537; p < 0.001), while a positive linear correlation was observed between RS and mortality (r = 0.352; p < 0.001), rebleeding (r = 0.381; p < 0.001), and blood transfusion requirement (r = 0.489; p < 0.001). Additionally, there was a positive linear correlation between the amount of blood product given to the patient and RS (r = 0.432; p < 0.001). The discriminatory power of RS for predicting mortality, rebleeding, and blood transfusion was evaluated, and our results showed an AUC of 0.802 (95% CI = 0.684–0.921; p < 0.001) for mortality, an AUC of 0.764 (95% CI = 0.641–0.886; p < 0.001) for rebleeding, and an AUC of 0.815 (95% CI = 0.731–0.899; p < 0.001) for blood transfusion requirement. Conclusions RS provides risk stratification in cases of GIB and aids in early treatment and prognosis prediction for high-risk patients. Furthermore, it has a high predictive power in determining rebleeding, mortality, and the need for blood transfusion. We believe that RS can be beneficial for clinicians in the management, prognosis, and mortality risk assessment of GIB cases, particularly in clinics such as emergency departments where a large number of patients need to be evaluated and treatment and discharge decisions are made within a short period of time.