Risk Factors for Rebleeding in Peptic Ulcer Bleeding: A Second Look at Second-Look Endoscopy.

Abstract

Listen

Translate article

Rebleeding, which occurs in 10–15 % of patients with peptic ulcer bleeding (PUB) [1], is associated with a twoto fivefold mortality increase, depending on the presence of other risk factors [2]. Therefore, identification of the predictors of rebleeding seems meaningful in order to identify high-risk patients needing close observation and rapid treatment in case of the development of rebleeding. According to previous studies, hemodynamic shock, usually defined as a systolic blood pressure \100 mmHg, often combined with tachycardia [100 beats/min, is the most powerful pre-endoscopic predictor of rebleeding [3, 4]. In a meta-analysis, hemodynamic shock was associated with an odds ratio (OR) of rebleeding of 3.3 [3]. Conversely, studies on the association between anemia and rebleeding have found conflicting results: Some of the existing data indicate that hemoglobin \10 g/L may be associated with an increased risk of rebleeding [3]. Data concerning the risks of transfusion are even more confounded by differing study protocols (e.g., preor postendoscopic transfusion, different categorization of volume) to the point that the rebleeding risk of pre-endoscopic transfusion is unknown. Regarding endoscopic predictors, active bleeding at endoscopy (OR 1.7), ulcer size [2 cm (OR 2.8), posterior duodenal ulcer location (OR 3.8), and high lesser gastric curvature ulcer location (OR 2.9) all predict rebleeding in a meta-analysis [3]. The type of endoscopic treatment applied does also affect the risk of rebleeding. A Cochrane analysis reported that combination of epinephrine injection with a second endoscopic treatment modality reduces the relative risk (RR) of rebleeding or persistent bleeding (RR 0.57) compared to endoscopic treatment with epinephrine alone [5]. Therefore, endoscopic monotherapy with injection of epinephrine should be avoided. In a meta-analysis based on eight randomized controlled trials (RCTs) published from 1994 to 2006, performance of second-look endoscopy within 16–48 h was associated with a significant reduction in rebleeding rate (OR 0.55) [6]. Generalization of this finding to current practice standards can be questioned because only one of the included studies used endoscopic combination therapy combined with high-dose infusion of proton-pump inhibitors [7]. Furthermore, detailed review of the fully published component studies revealed that a significant reduction in rebleeding was only evident in two studies that included patients with a very high risk of rebleeding (up to 47 % of included patients had hemodynamic shock) [6]. When these two trials were excluded from the meta-analysis, the association between performance of second-look endoscopy and rebleeding became statistically insignificant [6]. In a cost-effectiveness analysis, performance of secondlook endoscopy was only cost-effective after therapeutic endoscopy if the risk of rebleeding was greater than 31 % [8]. In this issue of Digestive Diseases and Sciences, Kim et al. [9] published a prospective multicenter study of risk factors for rebleeding among 699 patients with PUB from Forrest classification [10] Ia–IIb ulcers. Using multivariate logistic regression analysis, the authors reported that performance of second-look endoscopy was associated with a lower risk (OR 0.269) of rebleeding. High transfusion volume (above 5 units) and use of nonsteroidal anti-inflammatory drugs (NSAIDs) were both associated with a fourfold increase in risk of rebleeding. The authors & Stig Borbjerg Laursen Stig.laursen@rsyd.dk