Application of a virtual ruler in predicting postoperative rebleeding from esophageal varices in patients with liver cirrhosis.

ABSTRACT

BACKGROUND: Esophageal variceal diameter (EVD) is one of the most important predictors of variceal bleeding, as well as an important predictor of the effectiveness of endoscopic esophageal varices (EV) treatments. EVD is currently determined using visual inspection by endoscopic operators, meaning that results can vary widely between operators. This approach also means that cases unsuitable for endoscopic variceal ligation (EVL) can be complicated by postoperative hemorrhage. Thus, the purpose of this study was to explore the value of a virtual ruler (VR) in predicting rebleeding after the endoscopic treatment of EV in patients with cirrhosis. METHODS: We enrolled 588 patients with cirrhosis and EV (with and without gastric varices), who were treated with EVL or endoscopic injection sclerotherapy across 3 hospitals. We categorized participants into 2 groups, a nonbleeding group and a rebleeding group, according to whether they bled again after surgery. We compared basic demographic and clinical data, laboratory tests, EVD, and treatment modalities between the 2 groups. Potential risk factors for rebleeding after EV operations were analyzed using univariate and multivariable regression analyses. Correlations between esophageal variceal rebleeding and EVD were also analyzed, as was the consistency between visual EVD estimates and EVD measured using a VR. RESULTS: Child-Pugh class, albumin (ALB) levels, prothrombin time (PT), EVD (visual value), EVD (VR value), red sign, and the number of laps used for EVL showed statistically significant differences between the rebleeding and nonbleeding groups. Univariate regression analysis showed that Child-Pugh classification, ALB levels, PT, EVD (VR value), and red sign were strongly associated with rebleeding after endoscopic treatment of EV, whereas multivariable regression analysis showed that Child-Pugh classification, ALB levels, and EVD (VR value) were predictive factors for rebleeding after endoscopic treatment of EV. Differences between visual EVD estimates and VR EVD measurements were large. (Kappa value 0.391, P < .001). However, the 2 methods showed high agreement for EVD >1 cm (87/95) CONCLUSION: EVD (VR value) can more accurately predict rebleeding rates. It can also provide a basis for selecting appropriate endoscopic treatment modalities for EV and effectively circumvent postoperative EV rebleeding.