Clinical outcomes of lower gastrointestinal bleeding in patients managed with lower endoscopy: A tertiary center results

Abstract

Background:

Lower gastrointestinal bleeding (LGIB) is an urgent presentation with increasing prevalence and remains a common cause of hospitalization. The clinical outcome can vary based on several factors, including the cause of bleeding, its severity, and the effectiveness of management strategies. The aim of this study is to provide a comprehensive report on the clinical outcomes observed in patients with LGIB who underwent lower endoscopy.

Methods:

All patients who underwent emergency lower endoscopy for fresh bleeding per rectum, from May 2015 to December 2021, were included. The primary outcome was to identify the rate of rebleeding after initial control of bleeding. The second was to measure the clinical outcomes and the potential predictors leading to intervention and readmission.

Results:

A total of 84 patients were included. Active bleeding was found in 20% at the time of endoscopy. Rebleeding within 90 days occurred in 6% of the total patients; two of which (2.38%) were within the same admission. Ninety-day readmission was reported in 19% of the cases. Upper endoscopy was performed in 32.5% of the total cases and was found to be a significant predictor for intervention (OR 4.1, P = 0.013). Personal history of inflammatory bowel disease (IBD) and initial use of sigmoidoscopy were found to be significant predictors of readmission [(OR 5.09, P = 0.008) and (OR 5.08, P = 0.019)].

Conclusions:

LGIB is an emergency that must be identified and managed using an agreed protocol between all associated services to determine who needs upper GI endoscopy, ICU admission, or emergency endoscopy within 12 hours.

INTRODUCTION

Lower gastrointestinal bleeding (LGIB) is defined as bleeding originating from the gastrointestinal (GI) tract distal to the ligament of Treitz.[1] LGIB is well recognized as a common cause of hospitalization and can incur costly fees for both the patient and the treating hospital, considering the need for endoscopic evaluation and radiological examinations as a routine part of its management.[2,3,4,5] Upper GI bleeding has been widely investigated and well documented, whereas the existing data on the incidence, risk factors, and outcomes of LGIB is limited.[6] Global incidence in the past few decades has seen a significant decrease in upper GI bleeding.[1] This is likely attributable to the improvement in Helicobacter pylori eradication, wide proton pump inhibitors usage, and endoscopic management.[1,3,4] Meanwhile, LGIB prevalence is rising locally with rising incidences of hemorrhoids, colorectal cancer (CRC), and inflammatory bowel disease (IBD)[7,8,9] in addition to the increase in the aging population and use of anti-thrombotic and non-steroidal anti-inflammatory drugs.[1,3,4,10]

Diverticular disease, anorectal diseases, colonic polyps, colitis, angiodysplasia, and colorectal cancer are common etiology for LGIB.[3,4,5] However, compared to the West, the Middle East generally has a lower prevalence of colonic diverticulosis, especially in left-sided cases, but presents with higher rates of hemorrhoids.[6,11] Initial resuscitation followed by a colonoscopy to identify the source of the bleeding is required in most cases as an initial management of LGIB. In some cases, additional modalities may be necessary to control the bleeding, such as computed tomographic angiography, angioembolization, repeat endoscopic interventions, or surgical procedures.[4,5]

LGIB has a wide array of etiologies and an even wider spectrum of risk factors. Therefore, the definitive management of LGIB involves a complex and complicated hospital course with a risk of rebleeding, rescoping, readmission, reintervention, and mortality.[4,5,10] This study seeks to report the clinical outcomes of LGIB in patients who underwent urgent lower endoscopy and identify the prognostic factors of rebleeding and readmission.

PATIENTS AND METHODS

After ethical Institutional Review Board (IRB) approval at King Saud University, we conducted a retrospective cohort study and reviewed the electronic medical records of all patients who presented with fresh bleeding per rectum to the emergency department at King Khalid University Hospital, Riyadh, Saudi Arabia.

All patients who underwent colonoscopy or sigmoidoscopy between May 2015 and December 2021 were initially included. The patients were screened for fresh bleeding per rectum and excluded pediatric patients, those presenting with melena, and patients with incomplete medical records. Their files were reviewed for baseline characteristics and demographics, which included comorbidities and relevant personal and family history, as well as vital signs and laboratory findings upon presentation. Finally, we recorded the parameters related to lower GI bleeding, such as detection of active bleeding, rebleeding, site of bleeding, and etiology, in addition to any interventions or investigations that were required during or after the procedure.

Statistical analysis

All statistical data were calculated using Statistical Packages for Social Sciences (SPSS) version 26 (Armonk, NY: IBM Corp, USA). Descriptive statistics were summarized using frequencies and proportions (%) for all qualitative variables, whereas means and standard deviations were used to elaborate all quantitative variables. Between comparisons, Fischer’s exact tests were applied. Prognostic factors for detection of bleeding, intervention, and readmission were performed with corresponding odds ratios and 95% confidence interval. The level of significance was set to P < 0.05.

RESULTS

Between May 2015 and December 2021, 84 patients presented to the emergency department with fresh bleeding per rectum and underwent colonoscopy or sigmoidoscopy. The demographics and baseline characteristics are presented in Table 1. The mean age of the patients was 49.6 years, with 54 patients being males (64%). Prior history of GI bleeding was positive in 41 patients (48.8%). Nine patients (10.7%) had a history of malignancy. Majority of the patients were hemodynamically stable (80%) with hemoglobin levels above 8 g/dl (60%).

Table 1.

Baseline characteristics of the patients (n=84)

Study variables	n (%)
Age in years (mean±SD)	49.6±18.9
Male	54 (64.3%)
BMI (kg/m2) (mean±SD)	26.1±5.89
Smoking	12 (14.3%)
History of GI bleeding	41 (48.8%)
Personal history of cancer	9 (10.7%)
Personal history of IBD	16 (19.0%)
Family history of cancer	9 (10.7%)
Comorbidities
DM	18 (21.4%)
Hypertension	24 (28.6%)
Dyslipidemia	19 (22.6%)
Cardiovascular disease	25 (29.8%)
Cerebrovascular disease	4 (4.8%)
Asthma	10 (11.9%)
Liver disease	7 (8.3%)
Chronic renal disease	7 (8.3%)
History of GI bleeding	41 (48.8%)
Blood thinner medications
Aspirin	21 (25.0%)
Clopidogrel	8 (9.5%)
DOACs	8 (9.5%)
Warfarin	3 (3.6%)
Vital signs and laboratory findings upon presentation
Heart rate >100	15 (17.9%)
MAP <60	2 (2.4%)
Hgb level <7	13 (16.5%)
INR >l. 1	50 (69.4%)
Creatinine level >100	23 (29.9%)
Shock index >0.9	14 (16.7%)

SD: Standard Deviation, BMI: Body Mass Index, GI: Gastrointestinal IBD: Inflammatory Bowel Disease, DM: Diabetes Mellitus, DOACs: Direct Oral Anticoagulants, MAP: Mean Arterial Pressure, Hgb: Hemoglobin, INR: International Normalized Ratio

Lower endoscopy findings and parameters are presented in Table 2. Among the 84 patients, colonoscopy was done for 71 (86.6%) patients while the remaining underwent sigmoidoscopy (13.4%). Upper GI endoscopy was performed for 27 patients (32.5%). Seven patients (8.3%) were found to have upper GI source of bleeding, six of them had evidence of recent bleeding, while active bleeding was detected in one patient. The majority of the lower endoscopies (71%) were performed 24 hours after presentation. Active GI bleeding was detected in 17 patients (20%). The most common sites of bleeding were rectum (52.9%), followed by left-sided colon bleeding (29.4%). Colorectal cancer and ulcerative colitis were the most common source of active bleeding (23.5%).

Table 2.

Findings and parameters of lower endoscopy (n=84)

Study variables	n (%)
Type of scope
Colonoscopy	71 (86.6%)
Sigmoidoscopy	11 (13.4%)
Timing of lower endoscopy
• Within 12 h  • 12–18 h  • 18–24 h  • >24 h	3 (3.6%) 3 (3.6%) 18 (21.4%) 60 (71.4%)
Detection of active bleeding	17 (20.0%)
Site of active bleeding
Rectum  • Left-sided colonic  • Pan-colonic	9 (52.9%) 5 (29.4%) 2 (17.6%)
Source of bleeding
• Colorectal cancer  • Ulcerative colitis  • Hemorrhoids  • Angiodysplasia  • Polyps  • Diverticular bleed  • Colitis  • Post-polypectomy  • Rectal varices	4 (23.5%) 4 (23.5%) 2 (11.8%) 2 (11.8%) 1 (5.9%) 1 (5.9%) 1 (5.9%) 1 (5.9%) 1 (5.9%)
Aborted procedure	17 (20.23%)
Repeated scope	21 (25.0%)
Detection of bleeding in second scope	8 (38.09%)
Reasons of abortion
Poor bowel preparation  • Patient discomfort  • Risk of perforation	12 (70.5%) 3 (17.6%) 2 (11.8%)
Upper endoscopy	27 (32.5%)
Evidence of upper GI source	7 (8.3%)

Intervention and clinical outcomes are shown in Table 3. Out of 84 patients, 17 (20%) had active bleeding. Only 5 (6%) rebled within 90 days, of which two (2.4%) have rebled within the same admission. Three patients (3.6%) were admitted to the intensive care unit due to hemodynamic instability, and two of the three (66%) died within 30 days. The overall mortality rate was 3.6%. The readmission rate within 90 days was 19%. Five cases (31%) were readmitted for rebleeding. Two (12%) cases were readmitted as a colorectal cancer-related complaint. One (6%) case was readmitted as a UC flare. One case (6%) was readmitted due hospital acquired pneumonia. The remaining cases were admitted as medical causes such as angina, loss of consciousness, and hematology-related conditions.

Table 3.

Intervention and clinical outcomes (n=84)

Study variables	n (%)
Endoscopic intervention	13 (15.5%)
Type of endoscopic intervention
Polypectomy	8 (61.5%)
APC	2 (15.4%)
EGD banding	1 (7.7%)
Single band ligation	1 (7.7%)
Clips	1 (7.7%)
Other types of intervention
Need of CT-angiography	4 (4.8%)
IR intervention	1 (1.2%)
Surgical intervention	4 (4.8%)
Required blood transfusion	16 (19.0%)
pRBC pre-colonoscopy	12 (14.3%)
FFP pre-colonoscopy	1 (1.2%)
pRBC post-colonoscopy	4 (4.8%)
FFP post-colonoscopy	1 (1.2%)
ICU admission	3 (3.6%)
ICU stay in days (mean±SD)	21.3±10.3
Length of stay in days (mean±SD)	10.5±15.5
Rebleeding (within 90 days)	5 (6.0%)
Readmission (within 90 days)	16 (19.0%)
30-day mortality	3 (3.6%)

APC: Argon Plasma Coagulation, EGD: Esophagogastroduodenoscopy, CT: Computed Tomography, IR: Interventional Radiology, pRBC: Packed Red Blood Cells, FFP: Fresh Frozen Plasma, ICU: Intensive Care Unit

Prognostic factors for intervention and readmission are summarized in Tables 4 and 5. Overall, there were no significant factors for intervention [Table 4] except for the use of upper endoscopy, which was found to be a significant predictor for intervention (OR 4.1, P = 0.013) in patients with LGIB. The factors associated with readmission [Table 5] showed no significant association with most of the characteristics, however, a personal history of IBD was found to be a significant predictor for readmission (OR 5.09, P = 0.008). Patients who initially received a sigmoidoscopy were more likely to be readmitted (OR 5.08, P = 0.019).

Table 4.

Prognostic factors for intervention (n=84)

Factor	OR	95% Cl	P
Age in years	0.995	0.967–1.024	0.748
Gender
• Male	Ref	0.083–1.211	0.093
• Female	0.317
BMI (kg/m2)	1.022	0.930–1.123	0.651
Smoking	0.318	0.038–2.654	0.290
Comorbidities	0.969	0.333–2.815	0.953
History of GI bleeding	0.679	0.231–1.997	0.482
Aspirin	0.905	0.260–3.154	0.875
Heart rate >100	2.375	0.687–8.215	0.172
MAP <60	4.125	0.245–69.56	0.326
Hgb level <7	2.222	0.580–8.521	0.244
INR >I. I	0.850	0.252–2.863	0.793
Creatinine level >100	0.300	0.062–1.457	0.135
Shock index >0.9	1.754	0.475–6.478	0.399
Required blood transfusion	3.109	0.936–10.33	0.064
Rebleeding (within 90 days)	2.844	0.436–18.56	0.275
Upper GI endoscopy	4.118	1.354–12.52	0.013 *

*Significant at P<0.05 level. BMI: Body Mass Index, GI: Gastrointestinal, MAP: Mean Arterial Pressure, Hgb: Hemoglobin, INR: International Normalized Ratio

Table 5.

Prognostic factors for readmission (n=84)

Factor	OR	95% Cl	P
Age in years	1.031	1.000–1.064	0.052
Gender
• Male	Ref	0.244–2.510	0.679
• Female	0.782
BMI (kg/m2)	1.014	0.919 -–1.118	0.785
Personal history of IBD	5.099	1.518–17.12	0.008 *
Comorbidities	0.614	0.205–1.840	0.384
History of GI bleeding	1.061	0.357–3.152	0.916
History of cancer	0.5	0.058–4.312	0.528
Aspirin	1	0.284–3.517	1
Clopidogrel	1.476	0.569–8.098	0.654
DOACs	0.581	0.066-5.089	0.624
Warfarin	9.571	0.811–112.9	0.073
Heart rate >100	2.636	0.754–9.222	0.129
MAP <60	4.467	0.264–75.52	0.3
Hgb level <7	0.283	0.034–2.360	0.244
Shock index >0.9	1.196	0.291–4.906	0.804
Detection of active bleeding	0.709	0.197–2.558	0.599
Cause of bleeding determined	3.033	0.790–11.64	0.106
Required blood transfusion	0.236	0.029–1.931	0.178
pRBC pre-colonoscopy	0.345	0.041–2.891	0.327
Upper GI endoscopy	1.314	0.422–4.094	0.637
Type of scope:
• Colonoscopy	Ref	1.302–19.85	0.019*
• Sigmoidoscopy	5.083

*Significant at P<0.05 level. BMI: Body Mass Index, IBD: Inflammatory Bowel Disease GI: Gastrointestinal, DOACs: Direct Oral Anticoagulants MAP: Mean Arterial Pressure, Hgb: Hemoglobin, INR: International Normalized Ratio. pRBC: Packed Red Blood Cells

DISCUSSION

LGIB is one of the critical emergency presentations with increasing prevalence globally.[1,10] The outcome varies among patients and depends on many factors such as the cause of bleeding, severity, and appropriate management.[3,5] Prompt intervention is usually required in most of the patients.[4,5] There are limited data reported locally and this study aimed to report the clinical outcomes of patients with LGIB who underwent urgent lower endoscopy.

There is little doubt as to the importance of early diagnosis in cases presenting with acute GI bleeding. Early colonoscopy has been reported to play a role in decreasing the total length of stay and increasing the rate of detection of recent bleeding.[12] However, our study did not find any association between the rate of active bleeding and the timing of the colonoscopy. Active bleeding can be discovered in around 4–21% of colonoscopies according to recent literature.[13,14,15,16] Our study reported active bleeding in 20.2% of the colonoscopies. This wide range could be attributed to the variance in inclusion criteria, timing of colonoscopies, and definition of active bleeding.

Rebleeding is one of the major morbidity following initial control of LGIB. Compared to upper GI bleeding, LGIB has a more than twofold rebleeding rate.[17,18] Our study found the overall rate of rebleeding within 90 days to be only around 6%. This can likely be explained by the lower incidence of diverticulosis and the effective rate of treatment for IBD-related diseases.[11,19] Active rebleeding in the second scope reached 38.09%. While some studies have found similar rates of rebleeding of 4–8.8%,[20,21,22] most studies report higher rates of rebleeding. Someko et al. (2022)[23] found lower rebleeding rates in patients who received delayed colonoscopies. A local study by Mosli et al. (2020)[6] reported rebleeding rates of 25.6% and detected bleeding in 52% of their second scopes. Additionally, a meta-analysis of 12 studies found the overall rate of rebleeding to be 13.2–20.7% and 12.8–21.9% between urgent and elective colonoscopies, respectively.[24]

The most common culprits for bleeding in our study were colorectal cancer and ulcerative colitis (23.5%), followed by hemorrhoids and angiodysplasia (11.8%), whereas in a local study they previously reported lower rates of malignancy (18.6%) and higher rates caused by hemorrhoids (23.5%).[6] This can be explained by the nature of our hospital, which is a high-volume oncology and IBD center. In addition, the hospital policy indicated that all LGIB patients should be evaluated by the general surgery team to rule out local anorectal causes before GI team referral. Internationally, a study conducted in the United States reported the highest rates of diverticulosis (33.5%), followed by hemorrhoids (22.5%), and carcinoma (12.7%).[25] This discrepancy is expected as the prevalence of diverticulosis is low in the Middle East.[11] The incidence of colorectal cancer has been on the rise over the last decades in Saudi Arabia.[9] LGIB is one of the most alarming presentations for colorectal cancer.[26] With the local rise in the incidence rates of colorectal cancer,[27] the importance of a national screening program is crucial for early detection of colorectal cancer before presenting with complications such as LGIB.

The use of upper endoscopy in LGIB is valuable as an investigative tool. Its main utility is detecting upper GI sources which can present as hematochezia in some patients.[28] Upper scopes were performed in 32.5% of our patients. Bleeding from upper GI sources was discovered in 8.3% of our patients, with similar rates documented in the literature.[4,29] Emergency upper endoscopy should always be taken into account when presenting with hematochezia and hemodynamic instability in order to rule out acute upper GI bleeding.[28,30] However, in our center, there was no clear indication for upper endoscopy in patients with LGIB. The decision was mainly individualized based on the assessment of the on-call team.

Overall, our study did not find any factors to be significantly associated with increased risk of rebleeding. Previous studies have shown that there is no significant effect on mortality when receiving early colonoscopy.[12,31] Shiratori et al. (2023)[12] found that patients who received early colonoscopy had a curiously higher rate of rebleeding. A randomized clinical trial by van Rongen et al. (2019)[14] supports this and additionally found a higher rate of readmission in the early group. However, in contrast, a nationwide analysis by Sharma et al. (2022)[32] concluded that early colonoscopies did not affect the rate of readmission. Our study found 16 (19%) patients required readmission within 90 days, with no significantly associated risk factors. Similarly, Patel et al. (2019)[33] reported a 30-day readmission rate of 14.4% for LGIB, and metastatic disease was a significant predictor for readmission within 30 days (OR 2.15 P < 0.001). The appropriate interval of early versus late colonoscopy could not be appreciated in our study due to the small sample size. Most of our patients were hemodynamically stable upon presentation. Although the proper timing of colonoscopy still remains unclear, most studies generally recommend early colonoscopies in unstable patients.[12,14,34]

This study contains several limitations. First, the nature of the study being retrospective, potential bias can be expected. Also, being a tertiary hospital, the majority of such cases usually present secondary to CRC. In addition, most of the acute presentations from the emergency department did not present with active bleeding, which may be the reason for the variance in terms of protocol or investigations. This also explains the low intervention rate and clinical outcomes. Finally, the sample size is substantially smaller compared to other studies, which can impact the level of certainty from statistical analysis. Therefore, this topic can benefit from a multicenter approach which can help to generalize these findings.

In summary, LGIB is a life-threatening emergency that should be treated based on an agreed protocol that uses objective criteria agreed upon between gastroenterology, colorectal surgery, and emergency medicine, to specify who should receive an upper GI endoscopy and who should be admitted to ICU as well as who should undergo an emergency endoscopy within 12 hours.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.