Introduction
Endoscopy is critical for the management of patients with overt gastrointestinal bleeding (GIB), but some patients do not undergo endoscopic evaluation. We aimed to determine the clinical characteristics and outcomes of patients with overt GIB who did not have endoscopy, as well as the main reasons for not undergoing endoscopy.
Aims & Methods
This was a prospective study of patients admitted with overt GIB between 2013 and 2025 at a tertiary care center. We compared patients who had endoscopic evaluation (Endoscopy group) to those who did not (No Endoscopy group). Multivariate analysis was performed to identify predictors of mortality and rebleeding.
Results
A total of 979 patients were included, 114 (11.6%) of whom did not have endoscopy. Among the latter group, 66 patients were critically ill, or had a do-not-intubate or do-not-resuscitate status, a terminal illness or severe coagulopathy, making the risk of endoscopy too high. A comparison between 865 patients (Endoscopy group) and 114 patients (No Endoscopy group) revealed that the latter group had a higher age-adjusted Charlson Comorbidity Index (CCI), and lower requirement for blood transfusion (Table). On univariate analysis, the No Endoscopy group had higher 1-month and end-of-follow-up mortality but lower end-of- follow-up rebleeding. On multivariate analysis, blood transfusion was a predictor of 1-month and end-of-follow-up rebleeding (OR and CI) [2.7 (1.5-5.0), P <0.001], and [1.7 (1.2-2.5), P = 0.006], respectively. Age-adjusted CCI was a predictor of 1- month and end-of-follow-up mortality, [1.2 (1.1-1.3), P < 0.001] and [1.5 (1.4-1.6), P <0.001], respectively. Blood transfusion was a predictor of 1-month [2.7 (1.5-5.0), P <0.001] and end-of-follow-up mortality. Endoscopic evaluation was protective against 1-month and end-of-follow-up mortality [0.3 (0.2-0.5), P < 0.001] and [0.5 (0.3-0.9), P = 0.02], respectively. However, it had no effect on rebleeding rate.
A subgroup analysis was done to compare the 66 critically ill/terminal patients who did not have endoscopy to 108 controls who had endoscopy and who were matched for age (73 ± 2 vs 75 ± 1 years), sex (64% vs 63% males, P = 0.92), and CCI (6.6 ± 3 vs 5.9 ± 3, P = 0.09). The former group had higher 1 month and end of follow up mortality. On multivariate analysis, endoscopy was protective against 1-month mortality [0.2 (0.07-04), P˂0.001], but not against end of follow up mortality. Age-adjusted CCI was a predictor of end-of follow up mortality [1.3 (1.1-1.6), P = 0.002].
Table. Comparison of clinical characteristics and outcomes in patients with GIB between the Endoscopy Group and the No Endoscopy group
| Characteristics | Total N (%) | Endoscopy Group N (%) | No Endoscopy Group N (%) | P-value |
| Number of patients | 979 | 865 (88) | 114 (12) | - |
| Age (Years) | 68.3 ± 16.2 | 68.1 ± 16.1 | 70.1 ± 17.2 | 0.21 |
| Age-adjusted CCI* | 4.8 ± 2.9 | 4.7 ± 2.9 | 5.5 ± 2.9 | 0.003 |
| Nadir Hemoglobin (g/dL) | 8.3 ± 2.3 | 8.2 ± 2.3 | 8.7 ± 2.5 | 0.04 |
| Blood Transfusion | 584 (60) | 528 (61) | 56 (49) | 0.015 |
| One-month Rebleeding | 85 (9) | 78 (9) | 7 (6) | 0.31 |
| End of Follow-up Rebleeding | 205 (22) | 190 (23) | 15 (14) | 0.03 |
| One-month Mortality | 115 (12) | 84 (10) | 31 (27) | <0.001 |
| End of Follow-up Mortality | 521 (56) | 442 (53) | 79 (73) | <0.001 |
Conclusion
Patients with overt GIB who do not undergo endoscopy represent a higher-risk population with greater comorbidity burden and significantly worse mortality outcomes. Endoscopic evaluation was found to be protective against short and long-term mortality but not rebleeding. The apparent adverse outcomes of not doing endoscopy could be due to hidden confounders. These findings emphasize the importance of careful patient selection for endoscopic evaluation.
Disclosure
The authors have no conflicts of interest to disclose