This Guideline is an official statement of the European Society of Gastrointestinal
Endoscopy (ESGE). It provides guidance on the diagnosis and management of acute lower
gastrointestinal bleeding. The Grading of Recommendations Assessment, Development,
and Evaluation (GRADE) system was adopted to define the strength of recommendations
and the quality of evidence.
Abbreviations
APC:
argon plasma coagulation
AUROC:
area under the receiver operating characteristic curve
BSG:
British Society of Gastroenterology
CTA:
computed tomography angiography
DAPT:
dual antiplatelet therapy
DOAC:
direct oral anticoagulant
EBL:
endoscopic band ligation
EDSL:
endoscopic detachable snare ligation
ESGE:
European Society of Gastrointestinal Endoscopy
FFP:
fresh frozen plasma
GRADE:
Grading of Recommendations Assessment, Development and Evaluation
Hb:
hemoglobin
LGIB:
lower gastrointestinal bleeding
NICE:
National Institute for Health and Clinical Excellence
NSAID:
nonsteroidal anti-inflammatory drug
OR:
odds ratio
PCC:
prothrombin complex concentrate
PEG:
polyethylene glycol
PICO:
population, intervention, comparison, and outcome
RBC:
red blood cell
RCT:
randomized controlled trial
RR:
relative risk
UGIB:
upper gastrointestinal bleeding
1 Introduction
This European Society of Gastrointestinal Endoscopy (ESGE) Guideline aims to summarize
the available evidence and provide guidance regarding the diagnosis and management
of acute lower gastrointestinal bleeding (LGIB) focusing on the risk stratification
of patients, the role of endoscopy and other modalities (interventional radiology,
surgery) ([Fig. 1 ]), and on the appropriate management of antithrombotic agents in patients presenting
with acute LGIB. All recommendations in this Guideline apply in patients with major
LGIB as defined in section 4 of this document.
Fig. 1 Algorithm for assessment, stratification, and management of patients presenting with
acute lower gastrointestinal bleeding (LGIB).
2 Methods
The ESGE commissioned this clinical Guideline (ESGE Guideline Committee chair, J.v.H.)
and appointed a guideline leader (K.T.). The guideline leader established four task
forces each with its own leader (K.O., I.G., G.M., F.R.). Key questions were prepared
by the coordinating team (K.T., K.O., I.G., G.M., F.R., P.G.) and divided amongst
the four task forces (Appendix 1 s , see online-only Supplementary material). Each task force performed a structured
systematic literature search using keywords in English-language articles until August
31, 2020 in Ovid MEDLINE, EMBASE, Google Scholar, and the Cochrane Database of Systematic
Reviews. The hierarchy of studies included in this evidence-based guideline was, in
decreasing order of evidence level, published systematic reviews/meta-analyses, randomized
controlled trials (RCTs), prospective and retrospective observational studies, case
series.
Evidence on each key question was summarized in tables (Tables 1s-17 s ), using the Grading of Recommendations Assessment, Development and Evaluation (GRADE)
system, wherever applicable [1 ]. Grading of the evidence depends on the balance between the benefits and risk or
burden of any health intervention. Further details on ESGE guideline development have
been previously published [2 ].
The results of the literature search and answers to the PICO questions were presented
to all guideline group members during two online meetings conducted on September 26
and 27, 2020. Subsequently, drafts were created by each task force leader and distributed
between the task force members for revision and online discussion. In November 2020,
a full draft prepared by K.T., P.G. and the four task force leaders was sent to all
guideline group members. After the agreement of all members had been obtained, the
manuscript was reviewed by two independent external reviewers. The manuscript was
then sent for further comments to all ESGE member societies and individual members.
The final revised manuscript, having been agreed upon by all the authors, was submitted
to the journal Endoscopy for publication.
This ESGE Guideline was issued in 2021 and will be considered for update in 2026. Any
interim updates will be noted on the ESGE website: http://www.esge.com/esge-guidelines.html .
3 Definition, epidemiology, and risk factors
3 Definition, epidemiology, and risk factors
For the purposes of this guideline, the term “lower gastrointestinal bleeding” will
be used for any bleeding deriving from a site distal to the ileocecal valve and including
the rectum [3 ]
[4 ]. The majority of LGIB causes are summarized in [Table 1 ]
[4 ]
[5 ] and its most common clinical presentation is hematochezia.
Table 1
Overview of causes of acute lower gastrointestinal bleeding.
Benign diseases
Diverticular disease
Anorectal conditions
Hemorrhoids
Anal fissure
Solitary rectal ulcer
Rectal prolapse
Radiation proctopathy
Trauma
Vascular lesions
Angioectasias
Hereditary hemorrhagic telangiectasia
Dieulafoy’s lesion
Colonic or rectal varices
Colitis
Inflammatory bowel disease (ulcerative colitis, Crohn’s disease)
Ischemic colitis
Infectious colitis
Undetermined colitis
Polyps
Adenomas, hamartomas
Iatrogenic
Post-endoscopic intervention (polypectomy, EMR, ESD)
Post-surgical
Chronic anastomotic ulcer
Malignant diseases
Colorectal cancer
Anal cancer
Metastatic/invasive lesions
EMR, endoscopic mucosal resection; ESD, endoscopic submucosal dissection.
Diverticular bleeding is the commonest cause of LGIB with an incidence exceeding 20 %
among patients admitted to hospital [6 ]. The incidence of definitive diverticular bleeding (high risk stigmata at endoscopy
or bleeding diverticula on computed tomography angiography [CTA] or classic angiography)
was 20 %, but increased to 34 % when presumptive diverticular bleeding (diagnosis
of diverticular disease with lack of any other evident bleeding source in the endoscopy
or complementary work-up) was taken into account [7 ].
Anorectal diseases are the second most frequent cause of LGIB. Hemorrhoidal bleeding
is diagnosed in 12 %–21 % of patients admitted to hospital with a presenting complaint
of LGIB, which is usually small in amount and self-limited [6 ]. However, massive hemorrhoidal bleeding in elderly patients receiving anticoagulants
has been described [8 ].
Other causes of LGIB include different types of colitis (e. g. ischemic), radiation
proctitis, iatrogenic-induced bleeding (e. g. post-polypectomy), vascular malformations
(e. g. angioectasias), and colorectal cancer, among others, while no finding was recently
reported in 22.8 % of patients with acute LGIB [6 ].
Different risk factors may trigger LGIB (Table 1 s ). Alcohol consumption, smoking index ≥ 400, nonsteroidal anti-inflammatory drugs
(NSAIDs), low dose aspirin, and non-aspirin antiplatelet drugs have been identified
as independent risk factors for diverticular bleeding (odds ratio [OR] ≥ 1.9) [9 ], while bilateral diverticular location, nonselective NSAIDs, low dose aspirin, and
anticoagulants were associated with an increased risk of diverticular bleeding (OR ≥ 2.23)
in a case–control study [10 ]. Finally, a meta-analysis of six studies concluded that both NSAIDs and aspirin
significantly increased the relative risk (RR) for diverticular bleeding (RR ≥ 1.73)
[11 ].
The incidence of LGIB in patients receiving low dose aspirin in a UK-based, large
(more than 199 000 new low dose aspirin users; mean follow-up of 5.4 years) population
study was 1.22 (95 % confidence interval [CI] 1.16–1.29) per 1000 person-years, being
significantly higher than the incidence rate for upper gastrointestinal bleeding (UGIB)
(0.39 [95 %CI 0.36–0.43]) [12 ]. A study from Taiwan showed that low dose aspirin users presented more often with
LGIB during their first year of follow-up (0.20 %) [13 ]. Finally, a meta-analysis of 43 RCTs showed that the oral anticoagulants dabigatran
and rivaroxaban were related to an increased risk of major gastrointestinal bleeding
compared with conventional anticoagulants (vitamin K antagonists) (OR ≥ 1.27); however,
the overall risk for LGIB did not differ between the two groups (OR 0.88) [14 ].
4 Triage, risk stratification, and blood transfusion
4 Triage, risk stratification, and blood transfusion
4.1 How should patients with lower gastrointestinal bleeding be stratified according
to severity?
4.2 What should be the initial assessment of patients with lower gastrointestinal
bleeding according to the severity of the bleeding?
ESGE recommends that the initial assessment of patients presenting with acute lower
gastrointestinal bleeding should include: a history of co-morbidities and medications
that promote bleeding; hemodynamic parameters; physical examination (including digital
rectal examination); and laboratory markers. A risk score can be used to aid, but
should not replace, clinician judgment.
Strong recommendation, low quality evidence.
Risk factors for poor LGIB outcome include hemodynamic instability at presentation
(tachycardia, hypotension, syncope), ongoing bleeding (gross blood on initial digital
rectal examination, recurrent hematochezia), co-morbidities, older age, laboratory
findings (hemoglobin, creatinine, albumin, prothrombin time), blood transfusion requirement,
and concomitant medication (NSAIDs, antiplatelet agents, and anticoagulants) [158 ]. When stratifying patients with LGIB according to their severity, their vital signs
and the findings of cardiopulmonary, respiratory, abdominal, and digital rectal examination
should be included in the initial physical examination.
Although comparatively less well established than in UGIB, risk stratification scores
do exist for LGIB. Some have been developed to predict adverse outcomes, including
the ABC score [19 ], Strate score [15 ], NOBLADS [20 ], Sengupta score [16 ], BLEED [17 ], Birmingham score [21 ], Severe Acute LGIB (SALGIB) [22 ] score, and the HAKA score [23 ]; whilst others have been developed to identify patients at low risk of adverse outcomes:
Oakland score [24 ] and SHA2 PE [25 ]. Additionally, scores developed for use in UGIB, such as the Glasgow–Blatchford
bleeding score (GBS) [26 ] and Rockall score [27 ] have also been shown to have predictive ability in LGIB. No risk score has been
directly compared with clinician judgment, therefore the clinical data available at
the time of initial patient presentation is the best option to identify patients at
high risk for severe bleeding and other adverse outcomes (Table 2 s ).
4.3 What are the indications to admit a patient with acute lower gastrointestinal
bleeding to the hospital?
4.4 When can a patient with acute lower gastrointestinal bleeding be discharged and
followed-up as an outpatient?
ESGE suggests that no single risk score should be used in isolation to predict adverse
outcomes and determine the need for hospital admission in acute lower gastrointestinal
bleeding.
Weak recommendation, low quality evidence.
ESGE recommends that, in patients presenting with a self-limited bleed and no adverse
clinical features, an Oakland score of ≤ 8 points can be used to guide the clinician
decision to discharge the patient for outpatient investigation.
Strong recommendation, moderate quality evidence.
External validation studies of available tools [15 ]
[17 ]
[19 ]
[20 ]
[26 ]
[28 ] to assess the risk of adverse outcomes in acute LGIB have found that no score reliably
identifies all outcomes of interest [24 ]
[29 ]. Oakland et al. assessed risk scores in a prospective study of 2336 LGIB patients:
the best predictors of mortality, rebleeding, and red blood cell (RBC) transfusion
were AIMS-65 (area under the receiver operating characteristic curve [AUROC] 0.78),
the Oakland and the GBS (both AUROCs 0.74), and the Oakland score (AUROC 0.92), respectively;
however, no score reliably predicted intervention to treat bleeding (AUROCs 0.52–0.65)
[24 ]. [Table 2 ] summarizes the performance of different available scores for the prediction of mortality,
rebleeding, and need for RBC transfusion in patients with LGIB [30 ]. In a multicenter international study, the ABC score was found to be superior to
the AIMS-65 score in predicting mortality (AUROC 0.84 vs. 0.75) [19 ]. The analysis of other scores and other important adverse outcomes, such as severe
bleeding, need for endoscopic hemostasis, embolization, surgery, or RBC transfusion,
has been limited to small single-center studies [29 ]
[31 ]
[32 ].
Table 2
The performance of the BLEED, NOBLADS, Strate, Glasgow–Blatchford, AIM-65, and ABC
scores in the prediction of adverse outcomes in lower gastrointestinal bleeding (LGIB).
Score
Author (year)
External validation population
Population size
Mortality
Rebleeding
RBC transfusion
Sensitivity
Sensitivity
Sensitivity
BLEED
Oakland (2017)
All cases of LGIB, UK
2336
NR
NR
NR
NOBLADS
Oakland (2017)
All cases of LGIB, UK
2336
NR
NR
NR
Aoki (2018)
All cases of LGIB, Japan
511
NR
NR
NR
Strate
Oakland (2017)
All cases of LGIB, UK
2336
NR
NR
NR
Glasgow–Blatchford
Oakland (2017)
All cases of LGIB, UK
2336
NR
NR
NR
AIMS-65
Oakland (2017)
All cases of LGIB, UK
2336
NR
NR
NR
Laursen (2020)
All cases of LGIB with AIMS-65 ≥ 2, UK
2336
58 %
NR
NR
ABC
Laursen (2020)
All cases of LGIB with ABC ≥ 8, UK
2336
22 %
NR
NR
RBC, red blood cell; AUROC, area under the receiver operating characteristic curve;
NR, not reported. Adapted from Oakland K [30 ].
The Oakland [24 ] ([Table 3 ]) and SHA2 PE [32 ] scores have been specifically designed to identify low risk patients. The Oakland
score was validated in a retrospective study of 38 067 patients admitted to 140 hospitals
in the USA [33 ]. It comprises seven variables and has been designed to predict “safe discharge,”
a composite outcome defined as the absence of in-hospital rebleeding, RBC transfusion,
therapeutic intervention, in-hospital death, and readmission with subsequent LGIB
within 28 days. A score threshold of ≤ 8 points has a 95 % probability of safe discharge
and is the threshold recommended to identify patients for discharge [24 ]
[34 ]. Therefore, any self-limited LGIB with an Oakland score ≤ 8 should be considered
as minor, and such patients can be considered for early hospital discharge, while
all others, presenting with or without hemodynamic instability, should be considered
as having a major LGIB.
Table 3
The Oakland score for predicting the safe discharge of patients presenting with acute
lower gastrointestinal bleeding (LGIB).
Variable
Score
Age, years
0
1
2
Sex
0
1
Previous LGIB admission
0
1
Digital rectal examination findings
0
1
Heart rate, bpm
0
1
2
3
Systolic blood pressure, mmHg
5
4
3
2
0
Hemoglobin, g/dL
22
17
13
8
4
0
bmp, beats per minute.
Adapted from Oakland K et al. [24 ].
Oakland et al. assessed the NOBLADS, Strate score, GBS, AIM-65 and pre-endoscopy Rockall
score in predicting safe hospital discharge. All scores had an AUROC < 0.65, except
the Strate score (AUROC 0.69), GBS (0.80), and Oakland score (0.84) [24 ]. The ABC score can be used to identify patients with a low risk of death: a threshold
of ≤ 3 points is associated with a sensitivity of 73 %, specificity of 84 %, with
a mortality rate of 0.6 % [19 ].
4.5 When should patients with acute lower gastrointestinal bleeding be given a blood
transfusion?
ESGE recommends, in hemodynamically stable patients with acute lower gastrointestinal
bleeding and no history of cardiovascular disease, a restrictive red blood cell transfusion
strategy, with a hemoglobin threshold of ≤ 7 g/dL prompting red blood cell transfusion.
A post-transfusion target hemoglobin concentration of 7–9 g/dL is desirable.
Strong recommendation, low quality evidence.
ESGE recommends, in hemodynamically stable patients with acute lower gastrointestinal
bleeding and a history of acute or chronic cardiovascular disease, a more liberal
red blood cell transfusion strategy, with a hemoglobin threshold of ≤ 8 g/dL prompting
red blood cell transfusion. A post-transfusion target hemoglobin concentration of
≥ 10 g/dL is desirable.
Strong recommendation, low quality evidence.
A 2015 UK audit of 2528 patients admitted with LGIB found that 26.7 % received RBC
transfusion, with 80 % of these transfusions being considered, eventually, as avoidable
[35 ]. The American College of Gastroenterology [36 ], British Society of Gastroenterology [34 ], and NICE [37 ] guidelines, and an international consensus conference [38 ] have recommended that restrictive transfusion thresholds (Hb 7–8 g/dl) should be
used in hemodynamically stable patients with acute gastrointestinal bleeding, whilst
the threshold should be higher for patients with cardiovascular diseases.
These recommendations are based mainly on evidence deriving from UGIB studies, which
have shown that a restrictive blood transfusion strategy is associated with higher
survival, lower length of stay, and less RBC transfusion requirement [39 ]
[40 ]
[41 ]. However, a post-hoc analysis of the UK audit of acute LGIB [35 ]
[42 ] found no difference between liberal and restrictive transfusion strategies for the
odds of rebleeding or in-hospital mortality. Similarly, in both a systematic review
of RCTs and an overview of systematic reviews, mortality did not differ between restrictive
and liberal transfusion strategies for most of the populations [43 ]
[44 ] (Table 3 s ).
On the other hand, elderly patients and patients with cardiovascular disease may have
a different response to restrictive transfusion when compared with liberal transfusion.
A systematic review and meta-analysis of outcomes in patients with cardiovascular
disease in a non-cardiac surgery setting showed that the risk of acute coronary syndrome
in patients managed with restrictive compared with liberal transfusion was significantly
increased (RR 1.78 [95 %CI 1.18–2.70]) [45 ]. Finally, in a meta-analysis of nine RCTs evaluating restrictive vs. liberal transfusion
strategies in older adults, the risk of both 30-day and 90-day mortality was significantly
higher in the restrictive transfusion group (RR 1.36 [95 %CI 1.05–1.74] and RR 1.45
[95 %CI 1.05–1.98], respectively) [46 ]. These findings are particularly relevant to patients presenting with acute LGIB
as many of them have either cardiovascular morbidity or are elderly, with a median
age of 74 years [6 ].
5 Diagnosis and management of lower gastrointestinal bleeding: the role of endoscopy
5 Diagnosis and management of lower gastrointestinal bleeding: the role of endoscopy
5.1 When should colonoscopy be the first diagnostic modality in patients with acute
lower gastrointestinal bleeding?
ESGE recommends that colonoscopy should be the first diagnostic modality for hemodynamically
stable patients with acute lower gastrointestinal bleeding because of the therapeutic
options it offers.
Strong recommendation, very low-quality evidence.
Colonoscopy allows diagnosis, tissue sampling, and treatment during the same session
and is proposed by other current guidelines as the first-line procedure for the majority
of patients with acute LGIB [34 ]
[36 ]. Colonoscopy is estimated to have a diagnostic accuracy ranging from 42 % to 100 %,
while hemostatic therapy is performed in 10 % to 63 % of patients [36 ]
[47 ]. Unlike CTA, colonoscopy does not require active bleeding for diagnosis and avoids
radiation exposure and contrast-induced toxicity.
In a meta-analysis of 22 studies, the overall sensitivity and specificity of CTA in
the diagnosis of acute LGIB were 85.2 % (95 %CI 75.5 %–91.5 %) and 92.1 % (95 %CI
76.7 %–97.7 %), respectively [48 ]. The accuracy of tagged RBC scintigraphy is lower than CTA [49 ] and varies widely in the literature [36 ]
[48 ]
[49 ]. Angiography achieves a high rate of immediate hemostasis (86 %–100 %), but is usually
reserved as a second-line procedure owing to its invasiveness and rate of adverse
events (0 %–60 %) [50 ].
An RCT by Green et al. compared urgent colonoscopy (< 8 hours) to a standard protocol
that included tagged RBC scintigraphy, followed by visceral angiography when positive,
or elective colonoscopy when negative [51 ]. A definitive source of bleeding was found more often in the urgent colonoscopy
group, but the two approaches did not differ in safety, rebleeding, mortality, or
transfusion requirements. Early colonoscopy had a significantly higher diagnostic
yield (85 % vs. 45 %; P = 0.005) and was associated with shorter length of stay and lower transfusion requirements
compared with early radiographic procedures in a retrospective study [47 ].
Moreover, a recent systematic review compared the diagnostic and therapeutic yields
of endoscopy, CTA, and angiography [49 ]. Among the included studies that compared CTA with tagged RBC scintigraphy, one
study demonstrated a higher diagnostic yield for CTA, while the other two reported
no difference. A lack of studies precluded the performance of analyses of colonoscopy
vs. CTA and colonoscopy vs. first-line angiography.
Clerc et al. found that active bleeding was identified significantly more often with CTA compared
with lower gastrointestinal endoscopy (31 % vs. 15 %; P = 0.03) [52 ], whereas Lee et al. reported a similar yield for both modalities [53 ]. Miyakuni et al. performed a nationwide study in Japan selecting patients with severe LGIB who underwent
angiography or urgent colonoscopy within 1 day of admission [54 ]. After propensity score matching, in-hospital mortality was similar (RR 1.14 [95 %CI
0.95–1.36]), but the need for surgery within 1 day was lower in the angiography group
(RR 0.44 [95 %CI 0.29–0.67]).
None of the reviewed studies reported a cost–benefit analysis or showed a significant
difference in rebleeding rates, adverse events, 30-day mortality, 30-day surgery rate,
hospital length of stay, or transfusion requirements (Tables 4s–6 s ).
To conclude, low quality evidence indicates that CTA and colonoscopy have comparable
diagnostic yields and safety profiles. Colonoscopy has the advantage of allowing diagnosis
and treatment simultaneously, whereas CTA does not require bowel preparation and might
be preferred for selected patients with severe LGIB.
5.2 What is the appropriate timing for colonoscopy in patients with acute lower gastrointestinal
bleeding?
ESGE recommends that, in patients with major acute lower gastrointestinal bleeding,
colonoscopy should be performed sometime during their hospital stay because there
is no high quality evidence that early colonoscopy influences patient outcomes.
Strong recommendation, low quality evidence.
Available evidence comparing early vs. elective colonoscopy in the management of patients
with acute LGIB consists of seven systematic reviews with meta-analyses [55 ]
[56 ]
[57 ]
[58 ]
[59 ]
[60 ]
[61 ], four RCTs [51 ]
[62 ]
[63 ]
[64 ], and 16 observational studies [65 ]
[66 ]
[67 ]
[68 ]
[69 ]
[70 ]
[71 ]
[72 ]
[73 ]
[74 ]
[75 ]
[76 ]
[77 ]
[78 ]
[79 ]
[80 ] (Table 7 s ). Patients with “minor” LGIB managed as outpatients and patients with an UGIB source
were excluded from the RCTs [51 ]
[62 ]
[63 ]
[64 ] and most of the observational studies [66 ]
[67 ]
[69 ]
[71 ]
[72 ]
[73 ]
[74 ]
[75 ]
[76 ]
[77 ]
[78 ]. Early or urgent colonoscopy was defined as a colonoscopy performed within 24 hours
of presentation in most studies [62 ]
[63 ]
[64 ]
[65 ]
[66 ]
[67 ]
[68 ]
[69 ]
[70 ]
[71 ]
[72 ]
[73 ]
[74 ]
[75 ]
[76 ]
[77 ]
[78 ]. In RCTs, delayed or elective colonoscopy was defined as that performed between
24 hours and 96 hours from the time of hospital admission [51 ]
[62 ]
[63 ]
[64 ].
Two recent meta-analyses of observational studies suggested that early colonoscopy
reduces all-cause mortality (OR 0.86 [95 %CI 0.75–0.98]), the need for surgery (OR
0.52 [0.42–0.64]), blood transfusion requirements (OR 0.81 [0.75–0.87]), and hospital
length of stay (mean difference −1.7 days), with no significant differences in terms
of rebleeding, identification of the source of bleeding, adverse events, or need for
endoscopic therapy or interventional radiology [55 ]
[56 ]. One RCT also found that early colonoscopy was associated with shorter hospital
length of stay, but with an increased rate of recurrent bleeding [64 ], while another RCT revealed that a definitive source of bleeding was more often
detected in the urgent colonoscopy group [51 ].
However, two RCTs did not show any significant differences in the clinical outcomes
between early and elective colonoscopy [62 ]
[63 ]. Similarly, three meta-analyses that included the four available RCTs did not show
any differences regarding rebleeding, mortality, need for additional therapy, length
of stay, transfusion requirements, or any other clinical outcome [55 ]
[56 ]
[57 ]. Moreover, subgroup analyses assessing colonoscopy performed within 12 hours from
the time of hospital admission and a post-hoc meta-regression intended to determine
the impact of hemodynamic instability on clinical outcomes did not find any differences
between the groups [55 ]
[57 ].
We considered the certainty of evidence to be low, despite the significant number
of studies evaluating the appropriate timing of colonoscopy. All but one [80 ] of the observational studies were retrospective [65 ]
[66 ]
[67 ]
[68 ]
[69 ]
[70 ]
[71 ]
[72 ]
[73 ]
[74 ]
[75 ]
[76 ]
[77 ]
[78 ]
[79 ], and the definitions and selection criteria were heterogeneous. All RCTs were non-blinded,
with some concerns regarding bias (Tables 7 s and 8 s ), and two trials were terminated before reaching the pre-planned sample size [51 ]
[63 ]. The low number of RCTs and their limited sample sizes led to wide confidence intervals
for all outcomes assessed in the meta-analyses and impeded accurate evaluation of
publication bias. Finally, moderate to high heterogeneity was found for the pooled
data of hospital length of stay and units of blood transfused, altogether leading
to imprecision, inconsistency, and uncertain risk of publication bias in the available
evidence (Table 8 s ).
To conclude, studies comparing early (< 24 hours) vs. delayed (> 24 hours) colonoscopy
have focused on patients with major acute LGIB in whom colonoscopy was performed during
hospitalization. Retrospective data suggest that early colonoscopy may reduce all-cause
mortality, the need for surgery, blood transfusion requirements, and hospital length
of stay. However, meta-analyses of the RCTs have not confirmed these findings and
suggest that both groups have similar clinical outcomes. It remains unclear whether
selected acute LGIB patients could benefit from early colonoscopy.
5.3 Is there a role for unprepped sigmoidoscopy/colonoscopy in patients presenting
with acute lower gastrointestinal bleeding?
ESGE does not recommend unprepped lower gastrointestinal endoscopy (e. g. colonoscopy,
sigmoidoscopy) in patients with acute lower gastrointestinal bleeding.
Strong recommendation, low quality evidence.
Comparative studies on colonoscopy with and without bowel cleansing in acute LGIB
patients are lacking (Table 9 s ). Current guidelines recommend that colonoscopy should only be performed following
adequate bowel preparation [34 ]
[36 ]. Two recent prospective cohort studies in patients with severe LGIB reported the
use of “hydro flush colonoscopy” in 12 and 33 patients, respectively [81 ]
[82 ], where colonoscopy was performed after a tap-water enema and the bowel was further
cleansed using water or polyethylene glycol (PEG) solution delivered by a water-jet
pump and suction during colonoscopy. The bleeding source in many cases of acute LGIB
is located proximal to the rectum and sigmoid colon [82 ]
[83 ]; complete colonoscopy should therefore be the aim. However, in cases where CTA has
identified a bleeding source in the rectum or sigmoid colon, flexible sigmoidoscopy
can be considered.
5.4 Should upper gastrointestinal endoscopy be performed in patients presenting with
acute lower gastrointestinal bleeding?
ESGE recommends that upper gastrointestinal endoscopy be performed in patients presenting
with acute lower gastrointestinal bleeding and hemodynamic instability unless computed
tomography angiography has already been performed showing a definitive bleeding source
in the lower gastrointestinal tract.
Strong recommendation, low quality evidence.
There are no studies comparing upper GI endoscopy vs. no upper GI endoscopy in patients
with acute LGIB (Table 10 s ). Overall, in 8 %–9 % of patients presenting with LGIB, the source of bleeding is
found in the upper GI tract [6 ]
[84 ], whereas in patients with severe hematochezia and hemodynamic instability up to
15 % have an upper bleeding source [63 ]
[85 ]. A past medical history of portal hypertension, peptic ulcer, and antiplatelet medication
are known risk factors for UGIB [63 ]
[85 ]
[86 ]. An elevated blood urea/creatinine ratio (> 30) has also been found to be indicative
of UGIB [86 ]. The British Society of Gastroenterology (BSG) recommends that an upper GI endoscopy
should be performed immediately if no source is identified by initial CTA, while gastroscopy
may be the first investigation if the patient stabilizes after initial hemodynamic
resuscitation [34 ]. Similarly, the American College of Gastroenterology recommends upper GI endoscopy
be performed in patients with hematochezia and hemodynamic instability [36 ].
5.5 In patients with acute lower gastrointestinal bleeding undergoing colonoscopy,
what is the recommended bowel preparation?
ESGE suggests bowel preparation using large volume (4–6 L) PEG-based solution. Use
of a nasogastric tube combined with an antiemetic agent may facilitate bowel preparation
in patients who are intolerant of oral intake.
Strong recommendation, moderate quality evidence.
Adequate preparation of the colon in the setting of acute LGIB facilitates endoscopic
visualization, diagnosis, and treatment, and may reduce the risk of bowel perforation.
The available data are mostly from studies on acute LGIB using large volume bowel
preparation (4–6 L of PEG solution within 3–4 hours), with colonoscopy performed within
1–2 hours of the completion of bowel preparation [51 ]
[63 ]
[74 ]
[87 ] (Table 11 s ).
The use of lower volume or alternative colon preparation solutions in the setting
of LGIB has not been specifically addressed, but preliminary data appear encouraging
[88 ]
[89 ]
[90 ]. A prospective study [91 ] used 2 L of PEG solution added to the water-jet tank, starting from the left side
of the colon up to the cecum, in elderly patients (n = 33). The mean Boston Bowel
Preparation Scores during scope insertion and withdrawal were 2.6 and 7.2, respectively;
the mean (standard deviation) withdrawal time exceeded the insertion time (28.7 [6.9]
minutes vs. 17.1 [4.9] minutes), and the source of bleeding was found in 90.9 % of
patients.
In studies of urgent colonoscopy, one-third of patients required a nasogastric tube
to facilitate rapid bowel preparation [87 ]; therefore, a nasogastric tube can be placed to facilitate this process as long
as the risk of aspiration is low. Few studies have addressed bowel preparation-related
adverse events in acute LGIB. In an age- and sex-matched controlled retrospective
study (n = 161) using PEG solution or enema for those who could not completely consume
the PEG solution, 16 LGIB patients (9 %) experienced an adverse event (7 % hypotension,
2 % vomiting) [92 ].
5.6 What are the endoscopic hemostasis treatments for acute lower gastrointestinal
bleeding?
The summary of evidence is available in Table 12 s .
5.6.1 Diverticular bleeding
ESGE suggests mechanical therapy (e. g. through-the-scope/cap-mounted clip or endoscopic
band ligation) as the preferred treatment for diverticular hemorrhage.
Weak recommendation, moderate quality evidence.
Endoscopic treatment for diverticular bleeding has typically included thermal coagulation,
endoscopic clipping (through-the-scope or cap-mounted), endoscopic band ligation (EBL),
ligation using an endoscopic detachable snare (EDSL), and administration of epinephrine
local injection. Owing to the lack of strong, clear evidence on which hemostasis modality
is more effective and/or safer, recommendations depend on a combination of case reports,
case series, and prospective and retrospective studies, rather than RCTs and systematic
reviews.
5.6.1.1 Injection/thermal contact therapy Injection therapy is used in conjunction with other types of therapy, such as thermal
contact methods. Reports have shown their effectiveness for diverticular bleeding
[87 ]
[93 ]. Thermal contact therapies include heater probe therapy and bipolar coagulation,
with or without adrenalin injection [51 ]
[87 ]
[93 ]. However, thermal therapy poses the risk of perforation owing to the thin wall of
the colon. Injection of epinephrine alone should not be used as definitive hemostasis
therapy.
5.6.1.2 Endoscopic clipping Endoscopic clipping is the method used most often and typically poses less risk of
tissue injury. The through-the-scope method of clipping has been the recommendation
in previous guidelines [34 ]
[36 ].
5.6.1.3 Endoscopic ligation An historical control study done by Okamoto et al. showed EBL to be superior to clipping,
based on its significantly lower rebleeding rates after 1 year of follow-up for patients
with bleeding colonic diverticula (P < 0.01) [94 ]. A recent systematic review and meta-analysis compared several endoscopic modalities,
including ligation therapy, coagulation, and clipping, in patients with colonic diverticular
bleeding. The results suggested that ligation therapy was more effective compared
with clipping, in terms of avoiding transcatheter arterial embolization or surgery.
However, there were no significant differences in the rates of initial hemostasis
and early rebleeding (≤ 30 day) between the coagulation (n = 33), clipping (n = 192),
and ligation groups (n = 156). Pooled analysis showed that the efficacy of band ligation
to treat diverticular bleeding was up to 99 % (95 %CI 95 %–100 %), with the early
recurrent bleeding rate being 9 % (95 %CI 4 %–15 %) [95 ].
A recently published review on treatment trends for colonic diverticular bleeding
in Japan, which assessed five studies (n = 510), concluded that EBL is ultimately
superior to endoscopic clipping in terms of short- and long-term rebleeding rates
and that the proportion of patients needing transcatheter arterial embolization or
surgery after EBL is significantly lower than that for patients who underwent endoscopic
clipping [96 ].
While EBL is considered safe and effective [97 ]
[98 ]
[99 ], there have been reports suggesting that EBL carries the risk of serious complications,
such as delayed perforation, especially for right-sided lesions [100 ]
[101 ]
[102 ]
[103 ].
5.6.1.4 Endoscopic detachable snare ligation EDSL has also been used to ligate a bleeding diverticulum, similarly to endoscopic
band ligation. In a retrospective study, sustained hemostasis was achieved in 7/8
patients (88 %), with early rebleeding occurring in one patient [104 ].
5.6.1.5 Hemostatic topical agents Only small studies and case series have evaluated the efficacy and safety of hemostatic
topical agents in the treatment of LGIB. In a multicenter prospective study, the EndoClot
polysaccharide hemostatic system (EndoClot Plus Inc., Santa Clara, California, USA)
was used to treat diverticular bleeding; successful hemostasis was achieved in 83 %
of the patients, while the remaining two cases (17 %) re-bled secondary to malignancy
and a cecal ischemic ulcer [105 ]. A systematic review by Chen et al. [106 ] and two small studies [107 ]
[108 ] also described encouraging results for Hemospray (Cook Medical, Bloomington, Indiana,
USA) in cases of actively bleeding LGIB lesions.
5.6.2 Angioectasia
ESGE recommends treatment of bleeding angioectasia using argon plasma coagulation.
Strong recommendation, low quality evidence.
Argon plasma coagulation (APC) is considered the treatment of choice for angioectasia
in the upper and lower gastrointestinal tract because it is associated with lower
complication rates and less need for RBC transfusion [109 ]
[110 ]
[111 ]
[112 ]; however, comparative studies are lacking. Injection of a saline–adrenaline solution
prior to APC is suggested when treating right-sided colonic lesions, which present
a higher risk for perforation [111 ]. The optimal settings in terms of thermal effect intensity, gas flow, and duration
of the application depend on the site and size of the area that is being treated,
but typically the power ranges from 20–60 W and the gas flow rate from 1–2.5 L/minute
[109 ]
[110 ]
[111 ]
[112 ].
5.6.3 Delayed post-polypectomy bleeding
ESGE recommends the use of mechanical therapy (e. g. through-the-scope/cap-mounted
clips) and/or contact thermal coagulation as the primary treatment options of delayed
post-polypectomy bleeding.
Strong recommendation, low quality evidence.
ESGE suggests that hemostatic topical agents be used as a secondary treatment option
(e. g. rescue therapy) in cases of inadequate/failed hemostasis with ongoing bleeding.
Weak recommendation, low quality evidence.
The modality used most often to treat delayed post-polypectomy bleeding is through-the-scope
clips; however, the use of novel modalities, such as topical hemostatic agents and
cap-mounted clips, has also been reported [113 ]. Through-the-scope clips achieve successful hemostasis in most patients, but evidence
is based on clinical experience [113 ]
[114 ]
[115 ]. Treatment using bipolar coagulation, and non-contact coagulation therapy with APC
have also been reported [116 ]. Regarding hemostatic topical agents, a prospective multicenter study of patients
with active LGIB (n = 50) showed that hemostatic powder, as either monotherapy, combination
therapy, or rescue therapy, successfully induced hemostasis in 98 % of the patients;
however, five patients (10 %) experienced recurrent bleeding within 30 days [117 ].
6 Diagnosis and treatment of lower gastrointestinal bleeding: the role of interventional
radiology and surgery
6 Diagnosis and treatment of lower gastrointestinal bleeding: the role of interventional
radiology and surgery
6.1 When should computed tomography angiography be the initial diagnostic modality
in patients presenting with acute lower gastrointestinal bleeding?
ESGE recommends that patients with hemodynamic instability and suspected ongoing bleeding
undergo computed tomography angiography before endoscopic or radiologic treatment
to locate the site of bleeding.
Strong recommendation, low quality evidence.
ESGE does not recommend red blood cell scintigraphy in the setting of acute lower
gastrointestinal bleeding because of its limited accuracy in identifying the location
of the bleeding site and logistical constraints.
Strong recommendation, low quality evidence.
No RCT has been published on the accuracy of CTA in detecting LGIB. Retrospective
clinical studies report the sensitivity and specificity of CTA for LGIB to be 79 %–95 %
and 95 %–100 %, respectively [118 ]
[119 ]. If extravasation of contrast agent is detected at CTA, patients can then undergo
angiography and selective mesenteric embolization. Among 20 patients with LGIB, CTA
was positive in 9/13 patients (69.2 %) who were hemodynamically unstable and only
in 1/7 of the patients (14.3 %) who were hemodynamically stable [120 ].
Diverticular bleeding is diagnosed more often in patients undergoing CTA prior to
endoscopic examination than in those not undergoing CTA (35.7 % vs. 20.6 %) [121 ]. Furthermore, precise identification of the bleeding diverticulum is significantly
higher in patients with extravasation observed on CTA than in those without this (68 %
vs. 20 %; P < 0.001) [122 ]. Three studies in patients undergoing either CTA or RBC scintigraphy prior to selective
angiography did not detect any difference in the incidence of contrast-induced nephropathy
between the two diagnostic approaches [123 ]
[124 ]
[125 ]. Recently, Zink et al. demonstrated that CTA and RBC scintigraphy had similar sensitivities
in terms of LGIB detection (85.2 % vs. 94.4 %) [124 ]. However, CTA had a positive correlation with catheter-guided angiography compared
with RBC scintigraphy (67.7 % vs. 29.3 %). Jacovides et al. reported equivalent sensitivity
and specificity of RBC scintigraphy and CTA, but the bleeding site located by CTA
was more precise and consistent with the angiography findings [123 ]. Similarly, Feuerstein et al. showed that CTA located the site of LGIB more often
compared with RBC scintigraphy (53 % vs. 30 %) [126 ]. Finally, CTA is readily available at most hospitals, while RBC scintigraphy requires
more time to be performed (radiotracer preparation, with 60 to 90 additional minutes
needed for image acquisition after injection) and has more complicated logistics [123 ] (Table 13 s ).
6.2 When should interventional radiology be used for the treatment of patients with
lower gastrointestinal bleeding?
ESGE recommends that transcatheter arterial embolization should be reserved for the
treatment of acute, potentially life-threatening, lower gastrointestinal bleeding
either in hemodynamically unstable patients with active bleeding as demonstrated by
computed tomography angiography or in patients with brisk and ongoing bleeding not
amenable to or not effectively treated by endoscopic interventions.
Strong recommendation, low quality evidence.
ESGE recommends providing embolization within 60 minutes for a hemodynamically unstable
patient, because time has been proven to be a significant factor influencing patient
outcome.
Strong recommendation, low quality evidence.
Selective transcatheter endovascular therapy using microcatheters aims to decrease
arterial perfusion to the bleeding site, ensuring super-selective embolization of
arteries < 1 mm. The choice of the embolizing agent, including absorbable gelatin
sponges, cyanoacrylate glue, ethylene, or polyvinyl alcohol, and microcoils, is based
upon operator experience and local availability.
Transcatheter arterial embolization as the first step in the management of acute LGIB
should be reserved for patients demonstrating brisk and ongoing bleeding not amenable
to or not effectively treated by endoscopic means. Hemodynamic instability, a drop
in hemoglobin of ≥ 5 g/dL from admission, and blood transfusion requirement of ≥ 5
RBC units within 24 hours have been associated with the ability to locate the source
of LGIB at selective mesenteric angiography [127 ].
A systematic review found that super-selective angiographic embolization achieved
immediate hemostasis in 40 %–100 % of cases of diverticular bleeding, with rebleeding
rates ranging from 0–50 % [128 ]. The likelihood of identifying active bleeding was eight-fold higher if angiography
was performed within 90 minutes of CTA, as shown in a retrospective study [129 ], and decreased when its performance following RBC scintigraphy was delayed [130 ]. Therefore, embolization should be provided within 60 minutes in hemodynamically
unstable patients wherever an interventional radiology team is available. The risk
of transcatheter embolization-induced bowel ischemia is 1 %–4 % and is related to
the inability to achieve super-selective embolization [131 ]
[132 ] (Table 14 s ).
6.3 When should surgery be used as a diagnostic or therapeutic modality in patients
with acute lower gastrointestinal bleeding?
ESGE recommends that, except under exceptional circumstances, no patient should proceed
to emergency exploratory laparotomy unless every effort has been made to locate the
site of bleeding by endoscopic or radiological modalities.
Strong recommendation, low quality evidence.
ESGE recommends that surgery should only be undertaken if the lower gastrointestinal
bleed is due to underlying pathology that is not amenable to endoscopic or radiological
treatment, or if these modalities have failed.
Strong recommendation, low quality evidence.
No RCTs or non-randomized interventional studies have directly assessed laparotomy
(open or minimally invasive) as the first diagnostic modality in comparison to radiological
or endoscopic modalities in LGIB. Moreover, only a few prospective observational studies
have assessed such management protocols in LGIB [49 ] (Table 15 s ). In the UK prospective audit, only six patients (0.2 %) underwent laparotomy for
LGIB, with one of these following mesenteric artery embolization, and in only one
case had laparotomy been the initial intervention [6 ]. In general, complications following emergency laparotomy for severe LGIB are common,
including death [6 ]
[133 ]; therefore, surgical intervention should be undertaken only once all interventional
radiologic and endoscopic measures have been exhausted. Even though the need for emergency
laparotomy for LGIB is rare, there are indications where surgery may be justified
(e. g. aortoenteric fistula or bleeding Meckel’s diverticulum identified on Meckel’s
scan or at laparoscopy).
7 Management of antithrombotic agents in patients with lower gastrointestinal bleeding
7 Management of antithrombotic agents in patients with lower gastrointestinal bleeding
Anticoagulant and antiplatelet use is reported in up to 30 % of patients with acute
LGIB, with 2 %–5 % of patients receiving complex antithrombotic therapies, including
dual antiplatelet therapy (DAPT) or a combination of anticoagulant and antiplatelet
agents [6 ]
[134 ]. The management of antithrombotic agents often requires a multidisciplinary approach
that considers the severity of bleeding, the risk of rebleeding, and the patient’s
thrombotic risk. The ESGE recommendations in this guideline on the management of antithrombotic
agents are in line with those reported in the ESGE guideline on non-variceal UGIB
[135 ]
[136 ], as the majority of evidence derives from UGIB studies.
7.1 Management of vitamin K antagonists in patients with lower gastrointestinal bleeding
ESGE suggests not interrupting oral anticoagulation with vitamin K antagonists in
patients presenting with minor self-limited bleeding (i. e. Oakland score ≤ 8).
Weak recommendation, low quality evidence.
ESGE recommends withholding vitamin K antagonists in patients with major lower gastrointestinal
bleeding and correcting their coagulopathy according to the severity of bleeding and
their thrombotic risk. In patients with hemodynamic instability, we recommend administering
intravenous vitamin K and four-factor prothrombin complex concentrate (PCC), or fresh
frozen plasma if PCC is not available.
Strong recommendation, low quality evidence.
ESGE recommends restarting anticoagulant therapy following lower gastrointestinal
bleeding in patients with an indication for long-term anticoagulation.
Strong recommendation, moderate quality evidence.
ESGE suggests restarting anticoagulation at the earliest from day 7 after the interruption
of a vitamin K antagonist in patients at low thrombotic risk.
Weak recommendation, low quality evidence.
In those at high thrombotic risk, an earlier resumption of anticoagulation with heparin
bridging, preferably within 72 hours, is recommended.
Strong recommendation, very low quality evidence.
In patients presenting with minor self-limited bleeding (Oakland score
≤
8), oral anticoagulation can be continued, while its discontinuation is the “standard
of care” in patients with major LGIB. Vitamin K, prothrombin complex concentrate (PCC),
or fresh frozen plasma (FFP) can be used for rapid correction of vitamin K antagonist-related
coagulopathy, but the use of reversal agents (e. g. vitamin K) has been associated
with thromboembolism in patients at high thrombotic risk (i. e. those with a mechanical
heart valve) [137 ]. The correction of coagulopathy should not delay urgent therapeutic interventions
[138 ], which can be safely performed at therapeutic levels of anticoagulation [34 ]
[139 ].
Data from observational studies [140 ]
[141 ]
[142 ]
[143 ] and three meta-analyses [144 ]
[145 ]
[146 ] in the management of UGIB or GI bleeding highlight the net clinical benefit of restarting
anticoagulation after the bleeding event, in lowering the risk of thromboembolism
and death, despite increasing the risk of rebleeding (Table 16 s ). Because the thromboembolic risk increases over time, it is reasonable to restart
warfarin as soon as possible from day 7 onward following its interruption. In patients
at high thrombotic risk (prosthetic metal mitral heart valve, atrial fibrillation
with prosthetic heart valve or mitral stenosis, or less than 3 months after venous
thromboembolism) [147 ], cardiology societies recommend resumption of anticoagulation, with rapid titration
of prophylactic doses of low molecular-weight heparin to therapeutic doses within
48–72 hours [148 ]. If the risk of resuming anticoagulation outweighs its benefits, consultation with
a specialist (hematologist, neurologist, and/or cardiologist) is advised [148 ].
7.2 Management of direct oral anticoagulants in patients with lower gastrointestinal
bleeding
ESGE suggests not interrupting direct oral anticoagulants in patients presenting with
minor self-limited bleeding (i. e. Oakland score ≤ 8).
Weak recommendation, low quality evidence.
ESGE recommends temporarily withholding direct oral anticoagulants at presentation
in patients with major lower gastrointestinal bleeding.
Strong recommendation, low quality evidence.
ESGE suggests the use of reversal agents (idarucizumab in dabigatran patients and
andexanet or four-factor PCC in anti-factor Xa-treated patients) in coordination/consultation
with the local hematologist if bleeding is ongoing and/or there is recurrent hemodynamic
instability.
Weak recommendation, low quality evidence.
ESGE suggests restarting direct oral anticoagulant drug treatment following major
lower gastrointestinal bleeding as soon as possible from day 7.
Weak recommendation, low quality evidence.
Direct oral anticoagulants (DOACs) have a relatively short half-life, so that their
anticoagulant effect rapidly wanes over 12–24 hours. Most cases of major LGIB can
be managed by withholding the drug and waiting for the anticoagulant effects to dissipate.
However, in hemodynamically unstable patients, acute reversal of anticoagulation may
be required [6 ]
[134 ]
[148 ]. Vitamin K, FFP, and protamine administration are ineffective. Specific antagonists
are available as first-line reversal agents in DOAC patients presenting with life-threatening/uncontrolled
bleeding or requiring emergency surgery. Idarucizumab reverses dabigatran-related
coagulopathy within a few minutes and lasts for about 24 hours in more than 98 % of
patients, and has a low thrombotic complication rate (6 % at 90 days) [149 ]. Andexanet alfa, an inactive form of factor-Xa that neutralizes circulating factor-Xa
inhibitors, has recently been approved as an antidote to apixaban and rivaroxaban
in patients with life-threatening bleeding. Its clinical use is hindered by its limited
availability, high cost, and safety concerns regarding its procoagulant effect [150 ]. Four-factor PCC at a fixed dose of 2000 IU may represent an alternative to andexanet
alpha, with similar efficacy, yet with a lower thromboembolic risk [151 ]
[152 ]
[153 ].
Data regarding the optimal timing of DOAC resumption following LGIB cessation are
lacking, but similarly to warfarin, restarting the DOAC as soon as possible from day
7 onward after its interruption seems reasonable. DOAC resumption results in full
re-anticoagulation within 2–4 hours, therefore early resumption should be undertaken
with caution.
7.3 Management of antiplatelet agents in patients with acute lower gastrointestinal
bleeding
ESGE does not recommend routine platelet transfusion for patients with lower gastrointestinal
bleeding taking antiplatelet medications.
Strong recommendation, low quality evidence.
ESGE recommends withholding aspirin during the bleeding event in patients taking low
dose aspirin for primary cardiovascular prevention and considering its permanent discontinuation
unless clinically indicated after discussion with the referring specialist.
Strong recommendation, low quality evidence.
ESGE does not recommend withholding aspirin in patients taking low dose aspirin for
secondary cardiovascular prevention. If withheld, low dose aspirin should be resumed,
preferably within 5 days or even earlier if hemostasis is achieved or there is no
further evidence of bleeding.
Strong recommendation, moderate quality evidence.
ESGE does not recommend routinely discontinuing dual antiplatelet therapy (low dose
aspirin and a P2Y12 receptor antagonist) before cardiology consultation. Continuation
of the aspirin is recommended, whereas the P2Y12 receptor antagonist can be continued
or temporarily interrupted according to the severity of bleeding and the ischemic
risk. If interrupted, the P2Y12 receptor antagonist should be restarted within 5 days,
if still indicated.
Strong recommendation, low quality evidence.
There is limited evidence to guide the management of antiplatelet therapy in LGIB
(Table 17 s ). No drugs directly reversing platelet dysfunction exist and higher mortality, with
a similar risk of rebleeding, has been reported in GI bleeding patients on antiplatelet
therapy receiving platelet transfusion in a retrospective study [154 ].
A retrospective study of 295 LGIB patients on aspirin showed that continuing aspirin
was associated with an almost three-fold increased risk of recurrent LGIB, but also
with a 1.6-fold reduced risk of serious cardiovascular events and more than three-fold
reduced risk of death within 5 years [155 ]. A prospective analysis (n = 2528) evaluated the short-term outcomes of antithrombotic
drug interruption in patients hospitalized for LGIB. The in-hospital rebleeding rate
was higher in patients on antiplatelet therapy, with most bleeding events occurring
within 5 days from the time of admission. This incidence was comparable for patients
who continued antiplatelet therapy throughout their hospitalization and those who
had it withheld for fewer than 5 days [18 ]. Another cohort study, including 416 patients with gastrointestinal bleeding (162
LGIB), found no difference in rebleeding rates when the cutoff for resuming the antiplatelet
agent was set at ≤ 7 days [156 ].
According to these studies, continuing antiplatelet therapy during hospitalization
may be appropriate in most patients with high cardiovascular risk, who cannot discontinue
aspirin therapy, even for a short time. However, when temporary interruption is necessary
(i. e. severe and persisting bleeding), antiplatelet therapy should be resumed within
5 days, after which time about 50 % of circulating platelets are new and capable of
producing thromboxane [157 ]. In patients at low thrombotic risk on primary cardiovascular prevention, discontinuation
of aspirin at admission is recommended to reduce rebleeding without increasing the
risk of cardiovascular events. Permanent discontinuation of aspirin should also be
considered in liaison with the referring specialist.
Data regarding the management of LGIB patients taking DAPT are lacking. DAPT is mainly
prescribed in patients undergoing percutaneous coronary intervention with stent placement.
The management of such patients requires a careful assessment of their ischemic risk
and a cardiology consultation is mandatory. DAPT is associated with a five-fold increased
risk of in-hospital rebleeding, but not with bleeding-associated mortality [18 ]
[158 ]. However, discontinuing DAPT during the first 30 days following coronary stenting
and during the first 90 days following acute coronary syndrome is associated with
an increased risk of myocardial infarction and death [159 ]. Therefore, in patients at high ischemic risk, every effort should be made to continue
antiplatelet therapy. Similarly to acute UGIB, in cases of severe LGIB, continuing
aspirin as a single antiplatelet therapy appears to be reasonable, while withholding
the non-aspirin antiplatelet agent for no more than 5–7 days [136 ]. A large systematic review examined the safety of short-term antiplatelet discontinuation
among patients with drug-eluting stents and found very few cases of stent thrombosis
within 10 days of thienopyridine interruption. Because the risk of rebleeding associated
with DAPT is high, the required duration of DAPT should be reassessed after an LGIB
event [160 ].
7.4 Is there any role for antifibrinolytic medications in patients with acute lower
gastrointestinal bleeding?
ESGE does not recommend the use of tranexamic acid in patients with lower gastrointestinal
bleeding.
Strong recommendation, high quality evidence.
In a large (n = 78 291), nationwide, retrospective, propensity score-matched cohort
study, tranexamic acid administration did not reduce in-hospital mortality among patients
with diverticular bleeding [161 ]. Moreover, an RCT (the HALT-IT study) that evaluated 12 009 patients with gastrointestinal
bleeding (1328 LGIB) showed that intravenous tranexamic acid was associated with an
increased risk of venous thromboembolic events, without reducing mortality [162 ].
Disclaimer
The legal disclaimer for ESGE guidelines [163 ] applies to this Guideline.
Diagnosis and management of acute lower gastrointestinal bleeding: European Society
of Gastrointestinal Endoscopy (ESGE) Guideline
In the above-mentioned article, the institutions of Daniele Regge have been corrected.
This was corrected in the online version on July 8, 2021.
