Abstract
Objectives
The strategy of identifying stigmata of recent hemorrhage (SRH) and treating the bleeding source is important for the prevention of rebleeding in colonic diverticular hemorrhage (CDH). However, there are few known reports on SRH identification thus far. This large multicenter study evaluated factors correlated with SRH identification, including observation time during colonoscopy.
Methods
A total of 392 CDH cases were classified into presumptive CDH (n = 276) or definitive CDH with SRH (n = 116) on the basis of colonoscopy results. Multivariate Cox proportional hazards regression was employed to identify factors correlated with SRH identification. For the endoscopic treatment, endoscopic clips (EC), endoscopic band ligation (EBL) or endoscopic detachable snare ligation (EDSL) was performed.
Results
Longer observation time was significantly correlated with SRH identification in multivariate analysis (OR, 10.3 [95% CI: 3.84–27.9], p<.001). Receiver operating characteristic curve (ROC) analysis of the SRH identification rate by observation time indicated a high area under the curve (AUC) (0.79), and the threshold of the observation time was calculated at 19 min using Youden’s index. Moreover, the patients taken endoscopic hemostasis showed significantly lower early rebleeding rate than patients without endoscopic hemostasis (16.4% vs. 31.9%, p=.001), suggesting the importance of identifying SRH and treating the bleeding source for reducing the risk of recurrent bleeding.
Conclusions
Long-observation time correlated with SRH identification in this study, in which bowel preparation and water-jet scope and cap attachment are commonly used. This is the first known study to highlight the significance of observation time in the SRH identification rates.
Acknowledgments
The authors would like to thank Enago (www.enago.jp) for the English language review.
Author contributions
S.W. Study design, data analysis, data interpretation and drafting the manuscript. A.S.; revising the manuscript; S.W., A.S., Katsumasa K., and A.M. collected and interpreted data; H.U., T.M. and Kenichiro K. provided technical assistance for data analysis and interpretation; T.Y. supervised the overall study and assisted with the manuscript drafting.
Disclosure statement
No potential conflict of interest was reported by the authors.
Data availability statement
The datasets generated during and/or analyzed during this study are not publicly available, but are available from the corresponding author on reasonable request.