Keywords
colonoscopy - etiology - lower gastrointestinal bleeding - outcomes - rebleeding
Introduction
Lower gastrointestinal bleeding (LGIB) refers to blood loss from a source distal to
the ligament of Treitz.[1] The clinical spectrum of LGIB can be varied, ranging from self-limited melena or
hematochezia to massive hemorrhage with shock.[2] It accounts for approximately 20% of all cases of gastrointestinal bleeding, which
in many instances requires extensive diagnostic workup and usage of health resources.[3] Though literature from the West has described the clinical spectrum, patient characteristics,
and outcomes of LGIB,[4]
[5]
[6] there is limited prospective data regarding this common clinical condition from
South Asia. Hence, we aimed to describe the clinical characteristics, diagnostic evaluation,
therapeutic interventions, and outcomes of patients with LGIB at our institution,
a tertiary care center located in South India.
Materials and Methods
This was a single-center, prospective, observational study of patients with LGIB from
August 2023 to July 2024 who fulfilled the inclusion and exclusion criteria. We included
patients aged 16 years and above with a history of LGIB who were admitted for evaluation
or who had at least one episode of LGIB while the patient was admitted for another
clinical indication, and gave informed consent. Patients who did not give valid consent,
those below the age of 16, those in whom upper gastrointestinal (UGI) endoscopic evaluation
identified a cause for the LGIB, and those on a nasogastric (NG) tube which drained
altered blood were excluded.
Patient details including demographics, type of presentation, mode of onset, whether
presenting with LGIB or an existing in-patient who developed an LGIB, vitals and inotrope
requirements, comorbid illnesses and Charlson comorbidity index (CCI), medication
history, addictions, health status based on American Society of Anesthesiologists
(ASA) classification, and per rectal (PR) examination findings were recorded on a
pre-designed proforma.[7]
[8] The endoscopic findings, interventions performed, and the time to LGI endoscopy
were recorded, as also whether the endoscopy was repeated. The requirement for contrast-enhanced
computed tomography (CECT) of abdomen, interventional radiological procedures or surgery,
and need for blood products were also recorded.
The diagnostic algorithm followed in the study is depicted in [Fig. 1]. We performed CT angiography (CTA) in patients with hemodynamic instability despite
resuscitation and in those where a bleeding source was not identified after successful
resuscitation and a colonoscopy. CT enterography (CTE) was performed when colonoscopy
failed to identify a bleeding source, in patients with no active hematochezia or hemodynamic
compromise. Capsule endoscopy was performed in patients in whom a small bowel source
was suspected after unyielding gastroscopy, colonoscopy, and CTA/CTE failed to localize
the bleeding source.
Fig. 1 Algorithm for evaluation of patients with lower gastrointestinal (GI) bleed.
Patients were followed up until 28 days post-discharge. Important outcomes such as
duration of hospitalization, intensive care unit (ICU)/high dependency unit (HDU)
support, need for readmission, and occurrence of death were recorded. Management of
patients on antiplatelets and anticoagulation was based on existing guidelines.[9] However, individualized treatment plans were made in some patients based on the
treating clinician's decision. The institutional review board approved the study (IRB
Min No. 12013).
Definitions used in the study were mainly based on the large multi-centric study by
Oakland et al[4]:
-
LGIB: Hematochezia or melena without hematemesis.
-
Acute LGIB: Duration of onset of bleed ≤3 days.[10]
-
Clinically significant bleeding: Systolic blood pressure (SBP) <100 mm Hg, heart rate
(HR) ≥100/minute, and ≥1 unit of packed red blood cell (PRBC) transfusion.
-
Shock: HR ≥100/minute and SBp <100 mm Hg.
-
Rebleeding: Additional PRBC requirement and/or a reduction in hematocrit by ≥20% after
24 hours of hemodynamic stability.
-
Readmission: An unplanned admission due to further LGIB within 28 days of discharge.
-
28-day follow-up mortality: Death within 28 days of discharge from the hospital.
-
Current smoker: Adults who have smoked ≥100 cigarettes in his/her lifetime and within
the last 28 days.[11]
-
Heavy alcohol use: Consumption of >14 drinks/week in males or >7 drinks/week in females.[12]
-
Obscure overt GI bleed (OOGIB): When the source of bleeding is not found even after
the performance of standard upper and lower endoscopic examinations, small bowel evaluation
with video capsule endoscopy (VCE) and/or enteroscopy, and radiographic testing.[13]
Statistical Analysis
In the study by Oakland et al, the occurrence of rebleed was 13.6%, and readmission
was 10.3%.[4] Assuming the composite outcome of rebleed or readmission to be approximately 20%
and keeping a precision of 6.5% with a 95% confidence interval, the sample size for
our study was estimated to be approximately 150 patients with LGIB.
Categorical variables were expressed as frequencies and percentages, while continuous
variables were expressed as mean and/or standard deviation or as median with a range
of values. A p-value <0.05 was considered significant. Quantitative data between groups were compared
using Student's t-test for parametric data and Mann–Whitney U test for nonparametric data. Categorical
variables were compared using the appropriate chi-square or Fisher's exact test.
Results
A total of 150 patients with LGIB fulfilling inclusion and exclusion criteria were
included in the study cohort. The cohort included patients predominantly from South
India, mainly from Tamil Nadu (38%) and Andhra Pradesh (16%). The cohort consisted
of 124 (82.6%) patients admitted with LGIB, while the remaining 26 (17.3%) were in-patients
who developed LGIB while being admitted for another ailment (in-patient bleeds). Baseline
characteristics are depicted in [Table 1]. The mean (SD) age was 50 (16) years with a male predominance (72%). Median CCI
was 2 (range: 0–4). Cardiovascular disease and diabetes mellitus were the most common
comorbidities. Comorbidities were significantly higher among in-patient bleed than
those who presented with LGIB (100% vs. 72%, p < 0.001). Clinically significant bleed was found in 12%. Among the 145 patients who
ever underwent endoscopic evaluation for LGIB, the diagnostic yield for LGI endoscopy
was 61%.
Table 1
Baseline characteristics
|
Parameter
|
Admitted with LGIB
(n = 124)
|
In-patient LGIB
(n = 26)
|
Total
(n = 150)
|
p-Valuea
|
|
Mean (SD) age in years
|
50 (17)
|
53 (12)
|
50 (16)
|
|
|
Male sex (%)
|
88 (71)
|
20 (77)
|
108 (72)
|
|
|
Median (IQR) CCI
|
2 (0–4)
|
3 (1.7–4.2)
|
2 (0–4)
|
|
|
Comorbidities (%)
|
89 (72)
|
26 (100)
|
115 (77)
|
0.001
|
|
Chronic liver disease (%)
|
11 (9)
|
1 (3.8)
|
12 (8)
|
|
|
Cardiovascular (%)
|
54 (44)
|
16 (61)
|
70 (47)
|
|
|
Renal (%)
|
18 (15)
|
6 (23)
|
24 (16)
|
|
|
Respiratory (%)
|
12 (10)
|
4 (15)
|
16 (11)
|
|
|
Hematological (%)
|
15 (12)
|
7 (27)
|
22 (15)
|
|
|
Malignancy (%)
|
11 (9)
|
3 (12)
|
14 (9)
|
|
|
Neurologic disease (%)
|
11 (9)
|
4 (15)
|
15 (10)
|
|
|
Diabetes mellitus (%)
|
31 (25)
|
10 (38)
|
41 (27)
|
|
|
Others (%)
|
23 (19)
|
11 (42)
|
34 (23)
|
|
|
Presentation
|
|
Hematochezia (%)
|
76 (61)
|
21 (81)
|
97 (65)
|
|
|
Melena (%)
|
38 (31)
|
5 (19)
|
43 (29)
|
|
|
Both (%)
|
10 (8)
|
0 (0)
|
10 (7)
|
|
|
Onset
|
|
Acute (%)
|
39 (32)
|
20 (77)
|
59 (39)
|
0.001
|
|
Chronic (%)
|
85 (68)
|
6 (23)
|
91 (61)
|
|
Medications
|
|
Antiplatelets (%)
|
27 (22)
|
6 (23)
|
33 (22)
|
|
|
Anticoagulation (%)
|
10 (8)
|
5 (19)
|
15 (10)
|
|
|
AP & AC (%)
|
7 (6)
|
2 (8)
|
9 (6)
|
|
|
NSAID (%)
|
5 (4)
|
0 (0)
|
5 (3)
|
|
|
Current smoker (%)
|
7 (6)
|
4 (2)
|
11 (7)
|
|
|
Alcohol excess (%)
|
12 (10)
|
2 (8)
|
14 (9)
|
|
|
Shock (%)
|
11 (9)
|
6 (23)
|
17 (11)
|
|
|
Tachycardia alone (%)
|
17 (14)
|
5 (19)
|
22 (15)
|
|
|
Clinically significant bleeding (%)
|
13 (11)
|
5 (19)
|
18 (12)
|
|
|
Previous LGIB (%)
|
41 (33)
|
4 (15)
|
45 (30)
|
|
|
ASA score (n)
|
|
1 (%)
|
34 (27)
|
1 (4)
|
35 (23)
|
|
|
2 (%)
|
48 (39)
|
7 (27)
|
55 (37)
|
|
|
3 (%)
|
40 (32)
|
15 (58)
|
55 (37)
|
|
|
4 (%)
|
2 (2)
|
3 (12)
|
5 (3)
|
|
|
5 (%)
|
0
|
0
|
0
|
|
|
Mean (SD) hemoglobin at presentation in g%
|
8 (2.7)
|
8.5 (2.6)
|
8.1 (2.7)
|
|
Abbreviations: AP&AC, antiplatelet and anticoagulant; ASA, American Society of Anesthesiologists;
CCI, Charlson comorbidity Index; CVA, cerebrovascular accident; DAPT, dual antiplatelet
therapy; LGIB, lower gastrointestinal bleeding; NSAID, nonsteroidal anti-inflammatory
drug; PVD, peripheral vascular disease; SAPT, single antiplatelet therapy.
Note: aRepresents comparison between “admitted with LGIB” and “in-patient LGIB” groups. Figures
are depicted only for significant P values.
At presentation, 33 (22%) patients were on antiplatelet agents, 15 (10%) were on anticoagulation,
while 9 (6%) patients were on both.
Median (IQR) time from admission or primary gastroenterology consultation to LGI endoscopy
was 29 hours (13–72). The most common etiology of LGIB was colitis (26%), followed
by benign anorectal disorders (19%) ([Table 2]). Among colitis, inflammatory bowel disease (IBD)-related colitis was the predominant
etiology (59%). Colitis (27%) was the most common etiology among patients admitted
with LGIB, while benign anorectal disorders (27%) were the predominant source among
inpatient bleeds. The most common etiologies among those >60 years of age were benign
anorectal disorders (22%) and malignancy (14%), followed by angioectasiae (10%) and
drug-induced causes (10%). Among those with <60 years of age common etiologies were
colitis (35%), benign anorectal disorders (17%), and small bowel source (11%).
Table 2
Etiology of LGIB
|
Diagnosis
|
Total (n = 150)
|
|
Colitisa (%)
|
39 (26)
|
|
Benign anorectal disorders[b] (%)
|
28 (19)
|
|
Malignancy[c] (%)
|
14 (9)
|
|
Small bowel source[d] (%)
|
13 (8)
|
|
Angioectasiae (%)
|
10 (7)
|
|
Drug-induced[e] (%)
|
7 (4)
|
|
Diverticular (%)
|
5 (3)
|
|
Polyp (%)
|
3 (2)
|
|
Others[f] (%)
|
10 (7)
|
|
OOGIB (%)
|
5 (3)
|
|
Unknown (%)
|
16 (11)
|
Abbreviations: LGIB, lower gastrointestinal bleeding; OOGIB, obscure overt gastrointestinal
bleeding.
Note: aUlcerative colitis (19), radiation proctitis (4), Crohn's disease (4), infective dysentery
(2), ischemic colitis (2), CMV colitis (1), colonic tuberculosis (1), colonic ulcers—non-specific
(1), drug-induced colitis (1), immune-mediated colitis (3), infective colitis (1).
b Hemorrhoids (20), rectal ulcer (5), anal fissure (2), solitary rectal ulcer syndrome
(1).
c Carcinoma colon (7), carcinoma rectum (3), small bowel adenocarcinoma (2), small
bowel diffuse large cell B lymphoma (2).
d Small bowel neoplasm (6), small bowel Crohn's disease (3), small bowel vasculitis
(2), CMV ileitis (1), anastomotic ulcer (1).
e Anticoagulant-induced GI bleed (5), NSAID enteropathy (1).
f Small bowel vascular malformation (4), thrombocytopenia-related bleed (4), GI amyloidosis
(2).
The cause of LGIB could not be ascertained in 21 (14%) patients, of which 5 patients
were considered OOGIB and the remaining 16 did not undergo complete evaluation and
were deemed to have unknown etiology. Among patients with OOGIB, only two patients
had further bleeding and were readmitted. However, these did not require any further
transfusion or intervention up to the end of follow-up. Among the 16 patients with
unknown etiology due to incomplete assessment, 11 had evaluation using upper GI and
lower GI endoscopy along with CT imaging while 5 had endoscopy alone.
Endoscopic therapy was done in 7 (5%) patients, radiological intervention (mesenteric
artery embolization) in 5 (3%), and surgery in 10 (7%) ([Table 3]).
Table 3
Diagnostic and interventional modalities for LGIB
|
Parameter
|
Admitted with LGIB
(n = 124)
|
In-patient LGIB
(n = 26)
|
Total
(n = 150)
|
|
Diagnostic procedures (%)
|
|
Sigmoidoscopy (%)
|
21 (17)
|
6 (23)
|
27 (18)
|
|
Colonoscopy (%)
|
96 (77)
|
14 (54)
|
110 (73)
|
|
Small bowel enteroscopya (%)
|
7 (6)
|
2 (8)
|
9 (6)
|
|
Capsule endoscopy (%)
|
17 (14)
|
0
|
17 (11)
|
|
No endoscopy (%)
|
5 (4)
|
6 (2)
|
11 (7)
|
|
CT of abdomen (%)
|
86 (69)
|
15 (58)
|
101 (67)
|
|
Transfusion requirement (%)
|
79 (64)
|
21 (81)
|
100 (67)
|
|
PRBC transfusion (%)
|
74 (60)
|
20 (77)
|
94 (63)
|
|
>4 units PRBCs (%)
|
10 (8)
|
3 (12)
|
13 (9)
|
|
FFPs (%)
|
6 (5)
|
6 (23)
|
12 (8)
|
|
Platelets (%)
|
7 (6)
|
5 (19)
|
12 (8)
|
|
Cryoprecipitate (%)
|
2 (2)
|
2 (8)
|
4 (3)
|
|
Factor support (%)
|
1 (1)
|
0
|
1 (1)
|
|
Interventional procedures (%)
|
|
Endotherapy (%)
|
6 (5)
|
1 (4)
|
7 (5)
|
|
Adrenaline injection (%)
|
1 (1)
|
0
|
1(1)
|
|
Hemoclips (%)
|
0
|
0
|
0
|
|
APC (%)
|
4 (3)
|
0
|
4 (3)
|
|
Gold probe (%)
|
0
|
0
|
0
|
|
Dual therapy (%)
|
1[b] (1)
|
1[c] (4)
|
2 (1)
|
|
Interventional radiology (%)
|
3 (2)
|
2 (8)
|
5 (3)
|
|
Surgery (%)
|
7 (6)
|
3 (12)
|
10 (7)
|
|
ICU/HDU care (%)
|
14 (11)
|
9 (34)
|
23 (15)
|
Abbreviations: APC, argon plasma coagulation; CT, computed tomography; FFP, fresh
frozen plasma; ICU/HDU, intensive care unit/high dependency unit; IR, interventional
radiology; LGIB, lower gastrointestinal bleeding; PRBC, packed red blood cell.
Note: aIncludes double balloon enteroscopy, push enteroscopy, on-table enteroscopy, power
spiral enteroscopy.
b Hemoclip + APC.
c Adrenaline injection + Gold probe.
Our study showed a rebleeding rate of 7.9% (11/139) among patients who underwent endoscopic
evaluation during current admission ([Table 4]), of which 9 did not undergo a repeat endoscopy (due to improvement with steroids
for colitis or having undergone surgery or were moribund). Two underwent repeat colonoscopy
followed by capsule endoscopy which showed multiple small bowel angioectasiae. Univariate
analysis showed that requirement of more than 4 units of PRBC was significantly associated
with the occurrence of rebleeding, while the use of antiplatelets and/or anticoagulation
did not ([Supplementary Table S1], available in the online version). Angioectasiae was the most common etiology among
those who rebled (4/11). Three of the four patients with rebleeding due to angioectasiae
were from the small bowel.
Table 4
Outcomes of LGIB
|
Parameter
|
Admitted with LGIB
(n = 124)
|
Inpatient LGIB
(n = 26)
|
Total
(n = 150)
|
P value
|
|
Rebleedinga (%)
|
11 (8.9)
|
0 (0)
|
11 (7.3)
|
0.365
|
|
Median (IQR) duration of in-patient stay
|
5 (3–8)
|
18 (9–30)
|
6 (4–10)
|
0.0001
|
|
Readmission (%)
|
10 (8)
|
0
|
10 (6.6)
|
0.211
|
|
In-patient mortality (%)
|
0 (0)
|
3 (11.5)
|
3 (2)
|
0.003
|
|
28-day follow-up mortality
(%)
|
3 (2.4)
|
2 (7.7)
|
5 (3.3)
|
0.207
|
|
Overall mortality (%)
|
3 (2.4)
|
5 (19.2)
|
8 (5.3)
|
0.004
|
Abbreviation: LGIB, lower gastrointestinal bleeding.
Note: aAmong those who underwent endoscopic evaluation for LGIB in the current admission
(n = 139).
Overall, nine patients had a repeat colonoscopy: seven for further bleeding and two
were repeated to perform endotherapy. Readmission was required in 10 (6.6%) patients
for bleeding within 28 days of discharge.
Overall mortality was 5.3% (8/150). Mortality was greater among in-patient LGIB compared
with those admitted with LGIB (19.2% versus 2.4%, p = 0.004). The cause of death in seven patients was sepsis with severe comorbid illnesses.
Only in one patient was death attributed to uncontrolled bleeding. On univariate analysis,
in-patient bleeds, acute onset of bleed, and shock at presentation were significantly
associated with mortality ([Supplementary Table S2], available in the online version).
Discussion
Majority of the patients in our study were males, which was similar to most other
studies with the exception of the study by Oakland et al wherein females were predominant.[2]
[4]
[14]
[15]
[16]
[17]
[18] Indian studies including ours had patients mainly in the 4th and 5th decades of
life while the Japanese and Western studies had those in the 8th decade.[4]
[14]
[15]
[16]
[17]
[18] Key findings across major studies from India and other countries are depicted in
[Table 5].
Table 5
Summary of studies on LGIB
|
Parameters
|
Present study (n = 150)
|
Goenka et al (n = 166)
|
Dar et al (n = 300)
|
Lakhanpal et al (n = 138)
|
Bai et al (n = 720)
|
Nagata et al (n = 10,342)
|
Oakland et al (n = 2528)
|
Radaelli et al (n = 1198)
|
|
Year of study
|
2023–2024
|
1987–1991
|
2011–2013
|
2019
|
1996–2010
|
2010–2019
|
2015
|
2019–2020
|
|
Location
|
Tamil Nadu, India
|
Chandigarh, India
|
Kashmir, India
|
Himachal Pradesh, India
|
China
|
Japan
|
United Kingdom
|
Italy
|
|
Study design
|
Prospective
|
Prospective
|
Retrospective (n = 180) and prospective (n = 120)
|
Prospective
|
Prospective
|
Retrospective
|
Prospective
|
Prospective
|
|
Number of centers
|
Single center
|
Single center
|
Single center
|
Single center
|
Single center
|
Multi-center
|
Multi-center
|
Multi-center
|
|
Mean age in years
|
50
|
13–20 years: 30
21–40 years: 71
>40 years: 65
|
40
|
49
|
50
|
74
|
74
|
78
|
|
Male sex
n (%)
|
108 (72)
|
96 (58)
|
177 (59)
|
75 (54)
|
425 (59)
|
6,319 (61)
|
1,209 (47)
|
626 (52.2)
|
|
On antiplatelets
|
22%
|
−
|
−
|
−
|
−
|
20.6%
|
29.4%
|
32.1%
|
|
On anticoagulants
|
10%
|
−
|
0.67%
|
−
|
−
|
6.8%
|
15.9%
|
30.5%
|
|
Shock
|
11%
|
−
|
−
|
7.9%
|
−
|
29.5%
|
2.3%
|
32%
|
|
Most common diagnoses
|
- Colitis (26%)
- Benign anorectal condition (19%)
|
- UC (19%)
- Other colitis (12%)
|
- Polyps (23%)
- IBD (17.7%)
|
- Hemorrhoids (19%)
- IBD (19%)
- Carcinoma in colon (17%)
|
- IBD (30%)
- Polyps (23%),
|
- Diverticular (63.6%)
- Ischemic colitis (9.1%)
|
- Diverticular (26.4%)
- Benign anorectal conditions (16.7%)
|
- Diverticular (22%)
- Benign anorectal disorders (21.6%)
|
|
Endotherapy
|
5%
|
−
|
−
|
−
|
|
30.7%
|
2.1%
|
−
|
|
Interventional radiological procedure
|
3%
|
−
|
−
|
−
|
−
|
1.4%
|
0.8%
|
|
|
Surgery
|
7%
|
−
|
−
|
−
|
−
|
1%
|
0.2%
|
7.2%
|
|
Median length of IP stay (days)
|
6
|
−
|
−
|
−
|
−
|
7
|
3
|
7
|
|
Rebleed
|
7.3%
|
−
|
−
|
−
|
−
|
- In hospital(15.2%)
- Out of hospital(25.6%)
|
13.6%
|
8.8%
|
|
Readmission
|
6.6%
|
−
|
−
|
−
|
−
|
−
|
4.4%
|
−
|
|
Overall mortality
|
5.3%
|
−
|
−
|
−
|
−
|
- In hospital(0.9%)
- Out of hospital(6.8%)
|
3.4%
|
−
|
Abbreviations: UC, Ulcerative colitis; IBD, inflammatory bowel disease; IP, in-patient;
LNH, lymphoid nodular hyperplasia.
The etiological spectrum of LGIB is diverse across populations and study settings.
The large UK multicenter audit by Oakland et al had shown diverticular bleeding and
benign anorectal conditions as the leading causes of LGIB.[4] Diverticular bleeding was also the most common cause of LGIB in other studies from
Europe and the USA.[5]
[14]
[19] However, Bai et al in their cohort of 720 patients and an accompanying systematic
review including 160 studies (53,951 patients) from China showed colitis and colorectal
carcinoma, respectively, as the leading causes.[2] Although studies from Japan in the past had shown colitis as the leading cause of
LGIB, diverticulosis was the predominant etiology in a recent multicentric study from
the same region.[15]
[20]
[21]
Studies from northern, eastern, and southern parts of India have shown varying non-diverticular
etiologies such as IBD colitis, benign anorectal disorders, and polyps respectively,
as the major causes of LGIB.[16]
[17]
[18]
[22]
[23] Our results are also in tandem with those from Goenka et al and the Chinese cohort
with colitis (IBD-related and other causes of colitis) being the major etiology of
LGIB.[2]
[16] The difference in etiologies of LGIB in Western studies compared with our cohort
could be attributed to older patients in their study group and differences in diet.[16]
The inability to localize the source of bleeding in our study was similar to the other
studies, ranging from 3 to 30%.[2]
[4]
[14]
[15]
[16]
[17]
[18]
[24]
[25] This situation creates considerable anxiety among both patients and gastroenterologists
alike. Data on outcomes in patients with OOGIB are unavailable from India or South
Asia. However, multiple studies from the West have shown similar results as ours,
wherein patients with negative capsule endoscopy have favorable outcomes with low
rebleeding rates, fewer transfusion requirements, and lower healthcare utilization
in future.[26]
[27]
[28] We suggest that this be highlighted to patients while discussing the need for capsule
endoscopy evaluation in LGIB, when endoscopy and radiologic imaging are unyielding.
Majority of our patients (85%) were managed conservatively. However, the Japanese
multicentric study showed high rates of endotherapy (30%). The low endotherapy rate
(5%) in our study could be attributed to colitis being the predominant etiology, which
is mostly managed medically. Dar et al from India have also reported high rates of
endotherapy (38%), possibly due to polypectomy required for managing polyps, the most
common etiology in their series.[18]
Compared with our cohort, the UK audit by Oakland et al had higher rebleeding rate
(13.6%) but a similar readmission rate (4.4%).[4] Rebleeding rates across other studies ranged from 8 to 15.2%, similar to our cohort.[14]
[15]
[29] The severity of initial bleeding may be associated with a higher incidence of rebleeding,
as evidenced by the significant association between the requirement of more than four
packed red blood cell transfusions and subsequent rebleeding episodes. Unlike the
UK audit,[4] use of antiplatelet or anticoagulation was not associated with rebleed in our study.
Mesenteric artery embolization is considered a life-saving tool in the therapeutic
armamentarium for LGIB in patients who fail endotherapy and those in whom endoscopy
may not be feasible due to hemodynamic instability before surgery is considered. Embolization
rates similar to ours ranging from 0.3 to 1.4% were described in various other studies.[4]
[14]
[15]
A surgical approach is generally the last resort in managing LGIB, except for a few
conditions, such as active hemorrhoidal bleeding, in which it is considered the first
line. Based on the predominant pattern of etiology of LGIB and available expertise,
various studies have reported rates of surgical intervention ranging from 0.2 to 21%.
Recent studies have reported lower surgical intervention rates than older cohorts.
Longstreth et al in 1997 reported a 16% surgical intervention rate, while an Italian
study had only 7.1% of subjects requiring the same.[14]
[30] Our cohort had a surgical intervention rate of 7%.
The mortality rate in our study was 5.3%. The rate of mortality was significantly
higher among in-patient bleed patients as compared with those admitted with LGIB,
further reiterating the impact of underlying comorbid illnesses on fatal outcomes
in patients with LGIB. In-patient bleeds, acute onset of bleed, and shock at presentation
were significantly associated with mortality. Longstreth et al have also reported
similar findings wherein mortality rates were higher in in-patient bleeds than those
who were admitted for LGIB, which was likely linked to a nosocomial complication or
severe underlying illness rather than the LGIB per se.[30] The mortality rate in our study was comparable to the UK and Italian cohorts (3.4%
in both)[4]
[14] but higher than the Japanese study (0.9%). This difference might have likely resulted
owing to the exclusion of those with those with in-patient bleeds in the latter.[15]
The strengths of our study were that it was a prospectively conducted study that captured
data from admission and assessed outcomes up to 28 days post-discharge in all patients.
Most of our patients were evaluated reasonably in detail for the source of bleed utilizing
advanced endoscopic and radiological modalities. This is the first study from South
Asia to assess outcomes of LGIB comprehensively.
Our study had limitations. The study was restricted to a single center and had a small
sample size. The data may not be truly reflective of the LGIB at the general population
level as ours was conducted in a tertiary care center setting. Conclusive prediction
of risk factors for rebleeding and mortality was affected by the low numbers.
In conclusion, the etiological spectrum of LGIB in our center differs from those described
in the West, with colitis being the predominant etiology in the former. The mortality
rate was similar to the West and was mostly attributed to the underlying comorbid
illnesses. Majority of our patients could be managed conservatively. Future multicentric
studies with larger sample sizes will give more insight into the actual characteristics
and outcomes of LGIB in the South Asian region, which can help create appropriate
management guidelines.
