risk of bleeding after percutaneous endoscopic gastrostomy (peg)
TRANSCRIPT
ORIGINAL ARTICLE
Risk of Bleeding After Percutaneous Endoscopic Gastrostomy(PEG)
Dushyant Singh • Alexandra S. Laya •
Omkar U. Vaidya • Syed A. Ahmed •
Aaron J. Bonham • Wendell K. Clarkston
Received: 22 November 2010 / Accepted: 29 October 2011 / Published online: 3 December 2011
� Springer Science+Business Media, LLC 2011
Abstract
Background Patients who undergo percutaneous endo-
scopic gastrostomy (PEG) placement are often on antico-
agulation and/or antiplatelet therapy with a potential
thromboembolic risk if these medications are discontinued.
Data on the safety of peri-procedural use of these drugs is
limited.
Aims To assess the risk and to identify any predictive
factors for post-PEG bleeding, and to determine if clopi-
dogrel increases the risk of bleeding following PEG.
Methods A retrospective chart audit was conducted from
January 1, 2002 to June 30, 2011.
Results A total of 1,541 patients underwent PEG place-
ment during this period. Gastrointestinal bleeding after
PEG placement occurred in 51 cases (3.3%) and bleeding
directly attributed to PEG was noted in six patients (0.4%).
Multivariate logistic regression analysis of variables (age,
gender, length of hospitalization, indication for PEG,
antiplatelet or anticoagulant medications) showed that
heparin infusion (P = 0.018) and length of hospitalization
(P = 0.029) were statistically significant predictors of
bleeding. The mean period for cessation and resumption of
clopidogrel with PEG placement were 2.2 and 1.3 days,
respectively.
Conclusion Although PEG is classified as a high-risk
endoscopic procedure, bleeding with PEG placement was
rare, even with use of anticoagulation and antiplatelet
medications. In selected patients on heparin infusion
undergoing PEG, delaying the procedure, alternative use of
low-molecular-weight heparin or close monitoring and
frequent assessments should be considered. Clopidogrel
did not contribute to an increase in bleeding risk, despite
being held for a much shorter peri-procedural period as
recommended by expert consensus.
Keywords Bleeding � Hemorrhage � Complications �PEG � Gastrostomy � Endoscopy
Abbreviations
ASGE American Society for Gastrointestinal
Endoscopy
APTT Activated partial thromboplastin time
GI Gastrointestinal
LMWH Low-molecular-weight heparin
PEG Percutaneous endoscopic gastrostomy
Introduction
In the absence of anticoagulants and antiplatelet agents, the
American Society for Gastrointestinal Endoscopy (ASGE)
D. Singh � A. S. Laya � W. K. Clarkston
Section of Gastroenterology, University of Missouri Kansas
City, Kansas City, MO, USA
D. Singh (&)
Graduate Medical Education (3rd Floor), Saint Luke’s Hospital,
4401 Holmes St, Kansas City, MO 64111, USA
e-mail: [email protected]
O. U. Vaidya � S. A. Ahmed
Department of Internal Medicine,
University of Missouri Kansas City, Kansas City, MO, USA
A. J. Bonham
Department of Biomedical and Health Informatics,
University of Missouri Kansas City, Kansas City, MO, USA
W. K. Clarkston
Section of Gastroenterology, Saint Luke’s Hospital,
Kansas City, MO, USA
123
Dig Dis Sci (2012) 57:973–980
DOI 10.1007/s10620-011-1965-7
[1] has defined a high-risk endoscopic procedure as one
that carries a risk of clinically significant bleeding of
greater than 1%. High-risk procedures include colonic
polypectomy (range 0.2–3% risk) [2, 3], endoscopic
sphincterotomy (0.76–3.2% risk) [4, 5], percutaneous
endoscopic gastrostomy (2–2.5% risk) [6, 7], and variceal
band ligation (3–5% risk) [8, 9].
Since its introduction in 1979, PEG (percutaneous endo-
scopic gastrostomy) has become a widely performed medical
procedure [10]. Use of antiplatelet and anticoagulant medi-
cations for a variety of cardiovascular, cerebrovascular, and
hematologic conditions has become widespread over the past
decade [11–16]. These medications may potentiate gastro-
intestinal (GI) bleeding, and thus the need for endoscopy in
patients taking these medications is increasing. Before the
introduction of antiplatelet drugs, for example clopidogrel,
the incidence of GI bleeding after PEG placement was
approximately 2.5% [6, 17]. A recent report [18] suggests
that the incidence of post-PEG bleeding continues to be
similar, approximately 2.8%.
The major dilemma concerning patients taking these
medications includes the potential risk of bleeding as a
result of endoscopic intervention and the risk of thrombo-
embolic events when such medications are withheld
[1, 19]. Recent ASGE guidelines published in 2009 [20]
for the use of anticoagulant and antiplatelet therapy for
endoscopic procedures recommends that patients who are
taking clopidogrel or ticlopidine should have these medi-
cations discontinued 7–10 days before PEG placement and
their resumption should be individualized.
With regard to aspirin, in the absence of a pre-existing
bleeding diathesis, endoscopic procedures may be per-
formed while the patients are on aspirin and other NSAIDs
at standard doses [20]. This was emphasized by similar
recommendations of the American College of Cardiology
Foundation Task Force [21]. Cardiovascular and neurology
consultants are often reluctant to discontinue antiplatelet
therapy, especially in the face of recent percutaneous cor-
onary and central nervous system vascular intervention.
It should be recognized that the risk of a thromboem-
bolic event varies according to the underlying condition. In
clinical management, it is useful to consider a disease
process in terms of a low risk versus a high risk for
thromboembolization. Overall, anti-coagulation lowers the
risk of arterial or venous thromboembolism by 66–80%
[22–25]. Therefore, the decision to reverse anticoagulation
therapy is not always straightforward. Warfarin therapy
should be discontinued 3–5 days before the procedure and
bridged with LMWH (low-molecular-weight heparin) or
unfrationated heparin as an intravenous infusion. For
patients on LMWH, either as bridge therapy or actual
therapy, LMWH should be discontinued at least 8 h before
the PEG procedure. The decision when to restart therapy
should be individualized. Heparin infusion should be
discontinued 4–6 h before PEG and restarted 2–6 h after
the procedure is completed. These guidelines are based on
expert opinion and best clinical practice; there are no
prospective randomized controlled trials to support them
[20].
In addition, there is a lack of consensus amongst gast-
roenterologists regarding the management of these agents
in the peri-procedural period. An international survey in
2008 [26] concluded that opinions and clinical practice
patterns for management of anticoagulation and antiplatelet
therapy differed significantly among Western and Eastern
endoscopists. This lack of uniformity has been confirmed
by multiple studies [27–29]. We initiated this study with
the objective of determining whether the use of antiplatelet
and anticoagulant agents was associated with a significant
risk of post-PEG bleeding.
Methods
From January 1, 2002 to June 30, 2011, all PEG procedures
performed at a single, large quaternary-care academic
medical center were reviewed. This 623-bed hospital
includes the largest regional heart institute and a stroke
institute, where high volumes of patients receive antico-
agulation or antiplatelet therapy for a variety of vascular
diseases. In addition, patients with dysphagia because of
stroke or neurologic disorders are common, and frequently
require PEG placement to enable maintenance of nutrition.
PEG was performed by using the standard pull tech-
nique adapted from the method first described by Ponsky
and Gauderer [30]. After IRB approval was obtained, a
retrospective on-line chart audit of patients who underwent
PEG placement was conducted.
Two independent gastroenterologists reviewed the data
and there was no conflict related to the cases identified with
bleeding. A third investigator subsequently reviewed all
cases and confirmed the findings of the two reviewers,
including medication exposures.
Data Collection
Data retrieved from the database included patient demo-
graphics, indication for PEG, endoscopic findings, admit-
ting diagnosis, comorbid illnesses, evidence of post-PEG
bleeding, requirement of blood transfusion, intervention for
post-PEG bleeding, peri-procedural antiplatelet and anti-
coagulant medications, and laboratory data including the
trend of hemoglobin, platelet, and coagulation profiles. The
use of each antiplatelet and anticoagulant drug was care-
fully recorded keeping in mind the time of the last dose
before the PEG and the first dose after the PEG.
974 Dig Dis Sci (2012) 57:973–980
123
Definition of Bleeding, Time Interval, and Severity
Gastrointestinal bleeding was defined as any episode of
bleeding (melena, hematochezia, bleeding from PEG site,
peristomal bleeding, occult blood in stool, and unexplained
drop in hemoglobin) occurring after PEG placement. It was
categorized as immediate, early, or late. Immediate post-
PEG bleeding was defined for study purposes as bleeding
occurring within the first 24 h after PEG placement, of
sufficient severity to require endoscopic or local surgical
intervention. Patients who developed bleeding after 24 h
but within the first seven days were classified in the
‘‘early’’ bleeding group. The ‘‘delayed’’ bleeding group
included patients who had bleeding from seven days after
PEG placement until hospital discharge.
The severity of post-PEG bleeding was based on a
modification of a grading system proposed by Cotton and
Williams [31] for post-sphincterotomy bleeding (Table 1).
For all patients who had post-PEG bleeding, their records
were carefully reviewed for the number of blood products
transfused and the methods used for hemostasis.
Statistical Analysis
Categorical and continuous demographics and clinical
characteristics (age, gender, indication for PEG, aspirin,
clopidogrel, heparin infusion, prophylactic subcutaneous
heparin, LMWH, and coumadin before and after PEG
placement) were compared between patients who had post-
PEG bleeding and those who did not; chi-squared tests
were used for categorical variables and independent sam-
ples and t tests were used for continuous variables. To
determine the association between post-PEG bleeding and
all the predictors of bleeding, multiple logistic regression
analysis was used. All reported P-values were two-sided,
and a P-value of \0.05 was considered statistically sig-
nificant. All statistical analysis was performed by use of
SPSS version 18 (IBM SPSS 2009; Chicago, IL, USA). For
the post-hoc power analysis, we used PASS software
(Hintze 2007; Kaysville, UT, USA).
Results
Patient Characteristics
A total of 1,541 patients underwent PEG placement from
January 1, 2002 to June 30, 2011. There were 55.2%
(n = 851) males and 44.8% (n = 690) females with a
mean age of 67.6 years (SD, 16.3). The mean and median
number of days in the hospital was 11.1 (SD, 9.6) and 9,
respectively. PEG was placed for benign indication in
91.4% (n = 1,408) of patients and for malignant indication
in 8.6% (n = 133) of patients.
Post-Procedure Bleeding: Incidence and Associations
Bleeding occurred in 51 patients (3.3%) and 1,490 patients
(96.7%) did not bleed. A detailed descriptive analysis was
performed amongst patients who had bleeding and those
who did not bleed (Tables 2, 3).
Of 51 patients with bleeding, 29 (56.9%) had clinically
significant bleeding and underwent repeat endoscopy. Of
these 29 patients, six had bleeding directly related to
PEG. These were all secondary to either PEG-related ulcer
or peristomal bleeding. The remaining 23 patients that
underwent endoscopy had bleeding from gastric or duo-
denal ulcers at a location different from that of the PEG site
(ten patients), normal upper endoscopy and colonoscopy
(eight patients), rectal ulcer from rectal tube trauma (two
patients), bleeding ulcer at the colo-colonic anastomosis
Table 1 Grading of post-PEG bleeding
Severity of
bleeding
Definition
Mild Clinical or endoscopic evidence of bleeding
Decrease in hemoglobin less than 3 g/dL
No need for transfusion
Moderate Transfusion required but of four units or less
No angiographic intervention
No surgery
Severe Transfusion of five units or more
Angiographic or surgical intervention needed
Modified from Cotton and Williams [15]
Table 2 Descriptive statistics and bivariate comparisons of patients
who had no bleeding versus who had bleeding
No bleeding Had
bleeding
P-value
Total, no. (%) 1,490 (96.7) 51 (3.3)
Age (years) mean ± SD 67.6 ± 16.3 68.1 ± 14.7 0.823
Sex, male, no. (%) 826 (55.5) 24 (45.3) 0.141
Hospital days, mean ± SD 11.1 ± 9.6 15.0 ± 8.5 0.003*
Indication, malignant,
no. (%)
132 (99.2) 1 (0.8) 0.075
ASA 542 (96.1) 22 (3.9) 0.450
Clopidogrel 140 (97.9) 3 (2.1) 0.355
Heparin infusion 129 (92.1) 11 (7.9) 0.003*
Heparin (prophylaxis) 217 (96.0) 9 (4.0) 0.628
LMWH 259 (95.9) 11 (4.1) 0.529
Coumadin 176 (94.6) 10 (5.4) 0.122
LMWH, low-molecular-weight heparin
* P \ 0.05
Dig Dis Sci (2012) 57:973–980 975
123
(one patient), bleeding from biopsy site in the stomach for
Helicobacter pylori (one patient), and nasogastric tube
trauma (one patient). Endoscopy records indicated that the
bumper was loosened to assess if there was a source of
bleeding beneath the bumper. The other 22 patients
(43.1%) had very limited or minimum bleeding and a
repeat endoscopy was deemed to be of low diagnostic yield
and thus not attempted.
Bleeding Directly Related to PEG
Six patients (0.3%) had bleeding directly related to PEG
placement. These patients’ charts were closely reviewed
with the following results.
‘‘Immediate’’ bleeding was noted in two patients,
whereas three patients had ‘‘Early’’ bleeding occurring
within the first four days post-procedure. One patient had
‘‘delayed’’ bleeding that occurred seven days after PEG
placement. Because this is a retrospective analysis, the
patients that were discharged were not called/contacted for
follow-up data; it is possible that complications were
‘‘lost’’ to follow-up.
In terms of severity, only one patient had ‘‘severe’’
bleeding requiring multiple blood transfusions. The
bleeding in this patient was adequately controlled by
endoscopic injection of epinephrine at the PEG site. The
other five had post-PEG bleeding classified as ‘‘mild.’’
None required angiographic or surgical intervention. There
was no mortality as a result of post-PEG bleeding.
Among patients for whom PEG-related bleeding was
identified, 100% (n = 6) were on aspirin alone or in
combination with other anticoagulants (Table 4). None of
these patients were on clopidogrel. Five out of the six
patients were on gastrointestinal prophylaxis with either
H2 blockers or proton-pump inhibitors. There were no
concomitant bleeding disorders in these patients and
they did not have thrombocytopenia (platelet range,
156–558 9 103 per lL) or abnormal coagulation profiles
(international normalized ratio range, 1.0–1.3 and pro-
thrombin time range, 26–36 s) that would predispose them
to bleeding.
Primary Analysis: Predictors of Bleeding
Multiple logistic regression analysis was conducted to
determine which demographic and clinical factors were
predictive of post-PEG bleeding among these patients.
Factors included in the analysis were age, gender, indica-
tion for PEG, aspirin, clopidogrel, heparin infusion, pro-
phylactic subcutaneous heparin, LMWH, and coumadin.
Heparin infusion (P = 0.018) and days of hospitalization
(P = 0.029) were the only statistically significant predic-
tors of bleeding in this sample. The odds of a patient on a
heparin drip bleeding were 2.7 times the odds of a patient
who was not on a heparin drip (OR, 2.66). For every
increase of one day hospitalized, the odds of bleeding
increased by 0.02 (OR, 1.02). All other variables were not
statistically significant with P C 0.05 (Table 5).
Table 3 Incidence of post-PEG bleeding with different medications
Therapy Patients
receiving
medication
Number
who bled
Incidence of
bleeding (%)
None 600 16 2.7
Coumadin 186 10 5.4
Heparin infusion 140 11 7.9
Heparin (prophylaxis) 226 9 3.9
Aspirin 564 22 3.9
Clopidogrel 143 3 2.1
LMWH 270 11 4.0
Aspirin ? Clopidogrel 122 3 2.5
Any two medications 355 14 3.9
More than two
medications
112 7 6.3
LMWH, low-molecular-weight heparin
Table 4 Dosing of individual medication in patients with post-PEG
bleeding
Patient/Drug Dose Pre-PEG
dosing* (days)
Post-PEG
dosing* (days)
Patient 1
Aspirin 325 mg -1 2
Warfarin 5 mg NA** 8
Heparin SC 5,000 units -1 Not given
Patient 2
Aspirin 81 mg 0 1
LMWH 30 mg/kg -2 1
Patient 3
IV heparin NA -8 5
Patient 4
Aspirin 325 mg 0 2
Warfarin 5 mg NA** 1
LMWH 40 mg/kg NA** 1
Patient 5
Aspirin 81 mg 0 2
Patient 6
Aspirin 81 mg NA** 5
Warfarin 5 mg -4 1
IV, intravenous; LMWH, low-molecular-weight heparin; SC,
subcutaneous
* Day of PEG placement was considered as Day 0
** Patient never received a dose of the medication
976 Dig Dis Sci (2012) 57:973–980
123
Secondary Analysis: Mortality
Secondary analysis was conducted to determine whether
there were any significant predictors of patient mortality.
One-hundred and seventy-one patients (11.1%) died and
1,370 (88.9%) survived their hospitalization. Multiple
logistic regression analysis was conducted to determine
whether the following variables were predictive of mor-
tality: age, gender, days of hospitalization, indication
(benign or malignant), and bleeding (yes or no). Bleeding
was the only statistically significant predictor of mortality
(P = 0.020); all other variables were not statistically sig-
nificant (P [ 0.190 for each). The odds of a bleeding
patient dying were almost three times that of patients who
had no bleeding (OR, 2.89).
Subgroup Analysis: Clopidogrel
One-hundred and forty-three patients (9.2%) were treated
with clopidogrel. Among these 143 patients, three patients
(2.1%) had bleeding while on clopidogrel. Of the 143
patients, data regarding the number of days the medication
was held before the PEG were available for 78 patients,
and data regarding resumption of clopidogrel after PEG
placement were available for 75 patients. On average,
clopidogrel was held for 2.2 days (mean, 2.23 days; SD,
2.09 days) before PEG placement and resumed 1.3 days
(mean, 1.33 days; SD, 1.27 days) after PEG placement.
Discussion
PEG is the primary procedure for long-term nutrition in
patients with swallowing disorders. Although success
greater than 95% has been reported for PEG, procedure-
related complications are frequent, the most common being
infection [32, 33]. Most series report incidence of mor-
bidity ranging from 9 to 17%, although major complica-
tions occur in only 1–3% of cases [34, 35].
Bleeding is one of the major complications of PEG.
During the procedure, bleeding may be caused by puncture
of gastric wall vessels. The most common cause of post-
PEG bleeding is ulceration of the gastric mucosa beneath
the internal bumper when applied in very close proximity
to the mucosa [36]. Esophageal trauma, gastric erosions,
and unrelated peptic ulcer disease are less common etiol-
ogies of gastrointestinal bleeding after PEG [6, 36]. Cuta-
neous bleeding from the skin incision is common and
usually self-limited.
Aspirin and clopidogrel are crucial in the prevention of
thrombotic vascular events in cardiovascular and cerebro-
vascular diseases, and after coronary artery stent placement
[11–15]. Endoscopists are increasingly performing proce-
dures in individuals who must remain on these antiplatelet
therapies. Furthermore, premature discontinuation of aspirin
and clopidogrel in the setting of a drug-eluting coronary stent
is associated with high incidence of stent thrombosis [37].
There is sufficient evidence to show that there is no statisti-
cally significant association of aspirin and bleeding after
PEG [18] and after colonoscopy with polypectomy [3, 38,
39]. Data on clopidogrel in high-risk endoscopic procedures
is more limited, but three studies [18, 40, 41] have demon-
strated that clopidogrel alone was not an independent risk
factor for post PEG and postpolypectomy bleeding.
Overall, the incidence of post-PEG bleeding in this study
(3.3%) was very close to that previously reported in the lit-
erature (2.5–2.8%) [6, 17, 18]. Bleeding directly attributed to
PEG placement (excluding unrelated peptic ulcer disease
and other lesions) was very low (0.3%). Twenty-two patients
did not undergo repeat endoscopy, on the basis of the clinical
assessment that bleeding was mild. Even if these 22 patients
were assumed to be in the group that had bleeding directly
attributed to PEG placement, the incidence would be still
lower (1.8%). This lower incidence of bleeding could
potentially be because of delayed bleeding that could have
been missed, because the retrospective chart review was
limited to the same hospitalization.
It is interesting to note that heparin infusion was a sta-
tistically significant predictor of bleeding. Although anti-
coagulation has been linked to postpolypectomy bleeding
[3, 38, 39, 42], no studies have found any statistically
significant association between intravenous heparin infu-
sions and post-PEG bleeding.
Heparin remains the most widely used parenteral anti-
thrombotic [43]. It is used to achieve immediate antico-
agulation. It has a short duration of action that can be
advantageous when it is necessary to vary the intensity of
Table 5 Logistic regression analysis for predictors of bleeding
Predictor b df P-value OR 95% CI
Age 0.00 1 0.943 1.00 0.98–1.02
Sex -0.37 1 0.191 0.69 0.40–1.20
Hospital days before PEG 0.02 1 0.029* 1.02 1.00–1.04
Indication -1.33 1 0.191 0.26 0.04–1.95
Aspirin 0.18 1 0.557 1.20 0.65–2.20
Clopidogrel -0.64 1 0.311 0.53 0.16–1.81
Heparin infusion 0.98 1 0.018* 2.66 1.18–5.99
Heparin (prophylaxis) 0.25 1 0.539 1.28 0.58–2.81
LMWH 0.31 1 0.387 1.37 0.67–2.49
Coumadin 0.08 1 0.860 1.08 0.47–2.49
(constant) -3.66 1 \0.001 0.03
LMWH, low-molecular-weight heparin
* P \ 0.05
Dig Dis Sci (2012) 57:973–980 977
123
anticoagulation over a short time period, for example, in
patients undergoing endoscopy or surgery. Determination
of the activated partial thromboplastin time (APTT) ratio
enables monitoring during treatment. A typical regimen
would be to give sufficient heparin to prolong the APTT to
2.0 times normal. The 51 patients that had post-PEG
bleeding had a mean PTT of 35.8 s, with a minimum of
25 s and a maximum of 64 s (normal range, 18–28). The
APTT was measured at 6 to 12-h intervals and the dose of
heparin adjusted accordingly. Patients’ charts were
reviewed for cessation and re-initiation of heparin infusion
during the peri-procedural period. It was noted that heparin
infusion was held for 4–6 h before PEG and resumed 2–6 h
after PEG placement. This is consistent with current
guidelines and recommendations [11, 20, 44]. It may be
assumed that the likely cause of bleeding in these patients
was the increased APTT levels.
Recently, LMWH has become the heparin of choice,
especially in the prevention and treatment of deep-vein
thrombosis, pulmonary embolus, and unstable coronary
disease. LMWH has the advantage that it can be given as a
once daily subcutaneous injection without the need for
monitoring or dose adjustment. Compared with unfrac-
tionated heparin, LMWH has a superior benefit-to-risk
profile with an appreciably lower risk of heparin-induced
thrombocytopenia with thrombosis (HITT syndrome).
Although it is used in daily clinical practice, the efficacy
and safety of bridging therapy with LMWH is uncertain
[45, 46]. However, recent data suggest LMWH is a safe
and effective bridging therapy for patients requiring long-
term anticoagulation and undergoing endoscopy [47–50].
These patients required temporary interruption of elective
invasive procedures or surgery.
For high-risk procedures, for example PEG, delaying the
procedure should be considered in cases in which a finite
period of anticoagulation has been prescribed. If the pro-
cedure cannot be delayed, another form of heparin, for
example LMWH, can be considered. Otherwise, in selected
patients where heparin infusion is required as bridge ther-
apy and LMWH cannot be used, cautious use of peri-
procedural heparin infusion and closer monitoring of
anticoagulation status should be considered.
Another statistically significant predictor of bleeding
was the increased length of hospitalization. This could
possibly be explained by the increased severity of the
underlying illness, which might predispose to GI tract
injury and ulcerations, for example, stress ulcers.
The subgroup analysis of our sample showed that post-
PEG bleeding was associated with increased mortality.
This could be attributed to the severity of illness of these
extremely side patients.
Despite having 1,541 patients in this study, the sample
size for subgroup analysis was small, exposing our analysis
to the possibility of type II error. Because we were pri-
marily interested in the relationship between clopidogrel
and post-PEG bleeding, we conducted post-hoc power
analysis to determine how many patients would be needed
to detect a statistically significant effect of clopidogrel on
post-PEG bleeding. We calculated that, for a prospective
study, a total of 11,375 patients (i.e., 1,035 on clopidogrel
and 10,340 not on clopidogrel) would have to be enrolled
to obtain power of 80% with a type I error rate of 0.05. As
such, this study was underpowered. However, given the
large numbers of patients required to conduct an appro-
priately powered study, it is unlikely that such a study
would ever be feasible. Finally, there is the possibility of
unknown confounders that may bias these results. We tried
to control for these confounders by including all major
patient-related and procedure-related variables that might
protect or predispose individuals to post-PEG bleeding in
the multivariate analysis. Despite our efforts, this remains
an exploratory study. A prospective, appropriately powered
trial is needed for further investigation and corroboration of
these findings and associations.
In summary, despite being one of the potential highest-
risk endoscopic procedures associated with bleeding [1],
our study showed that PEG-related bleeding was rare, even
with the widespread use of anticoagulants and antiplatelet
medications. In view of increased post-PEG bleeding with
intravenous heparin infusion, attempts should be made
either to delay the procedure or to use LMWH as bridging
therapy. Where heparin infusion is deemed necessary,
close monitoring for GI bleeding should be considered.
Expert opinion has proposed that clopidogrel be dis-
continued 7–10 days before endoscopic procedures, but
clopidogrel did not contribute to an increase in bleeding in
our study, despite being held for a much shorter peri-
procedural period. In view of a low risk of post-PEG
bleeding in patients who have been exposed to antiplatelet
therapy coupled with a potential risk of cardiovascular and
cerebrovascular complications if these medications are
discontinued, the risk-to-benefit ratio of discontinuing
antiplatelet therapy far in advance of PEG should be
reconsidered.
Conflict of interest Authors do not have any conflicts (financial,
professional, or personal) relevant to the manuscript.
References
1. Eisen GM, Baron TH, Dominitz JA, et al. Guideline on the
management of anticoagulation and antiplatelet therapy for
endoscopic procedures. Gastrointest Endosc. 2002;55:775–779.
2. Nelson DB, McQuaid KR, Bond JH, Lieberman DA, Weiss DG,
Johnston TK. Procedural success and complications of large-scale
screening colonoscopy. Gastrointest Endosc. 2002;55:307–314.
978 Dig Dis Sci (2012) 57:973–980
123
3. Hui AJ, Wong RM, Ching JY, Hung LC, Chung SC, Sung JJ.
Risk of colonoscopic polypectomy bleeding with anticoagulants
and antiplatelet agents: analysis of 1657 cases. Gastrointest En-dosc. 2004;59:44–48.
4. Freeman ML, Nelson DB, Sherman S, et al. Complications of
endoscopic biliary sphincterotomy. N Engl J Med. 1996;335:
909–918.
5. Christensen M, Matzen P, Schulze S, Rosenberg J. Complications
of ERCP: a prospective study. Gastrointest Endosc. 2004;60:
721–731.
6. Schapiro GD, Edmundowicz SA. Complications of percutaneous
endoscopic gastrostomy. Gastrointest Endosc Clin North Am.
1996;6:409–422.
7. Larson DE, Burton DD, Schroeder KW, DiMagno EP. Percuta-
neous endoscopic gastrostomy indications, success, complications
and mortality in 314 consecutive patients. Gastroenterology.
1987;93:48–52.
8. Sarin SK, Lamba GS, Kumar M, Misra A, Murthy NS. Com-
parison of endoscopic ligation and propranolol for the primary
prevention of variceal bleeding. N Engl J Med. 1999;340:
988–993.
9. Lo GH, Lai KH, Cheng JS, Lin CK, Hsu PI, Chiang HT. Pro-
phylactic banding ligation of high-risk esophageal varices in
patients with cirrhosis: a prospective, randomized trial. J Hepatol.1999;31:451–456.
10. Gauderer MW. Twenty years of percutaneous endoscopic gas-
trostomy: origin and evolution of a concept and its expanded
applications. Gastrointest Endosc. 1999;50:879–883.
11. Makar GA, Ginsberg GG. Therapy insight: approaching endos-
copy in anticoagulated patients. Nat Clin Pract GastroenterolHepatol. 2006;3:43–52 (Review).
12. Juul-Moller S, Edvardsson N, Jahnmatz B, Rosen A, Sørensen S,
Omblus R. Double-blind trial of aspirin in primary prevention of
myocardial infarction in patients with stable chronic angina
pectoris. The Swedish Angina Pectoris Aspirin Trial (SAPAT)
Group. Lancet. 1992;340:1421–1425.
13. Antithrombotic Trialists’ Collaboration. Collaborative meta-
analysis of randomised trials of antiplatelet therapy for preven-
tion of death, myocardial infarction, and stroke in high risk
patients. BMJ. 2002;324:71–86.
14. Bhatt DL, Fox KA, Hacke W, et al. Clopidogrel and aspirin
versus aspirin alone for the prevention of atherothrombotic
events. N Engl J Med. 2006;354:1706–1717.
15. The SALT Collaborative Group. Swedish Aspirin Low-Dose
Trial (SALT) of 75 mg aspirin as secondary prophylaxis after
cerebrovascular ischaemic events. Lancet. 1991;338:1345–1349.
16. Steering Committee of the Physicians’ Health Study Research
Group. Final report on the aspirin component of the ongoing
Physicians’ Health Study. N Engl J Med. 1989;321:129–135.
17. Luman W, Kwek KR, Loi K, Chiam MA, Cheung WK, Ng HS.
Percutaneous endoscopic gastrostomy-indications and outcome
of our experience at the Singapore General Hospital. SingaporeMed J. 2001;42:460–465.
18. Richter JA, Patrie JT, Richter RP, et al. Bleeding after percuta-
neous endoscopic gastrostomy is linked to serotonin reuptake
inhibitors, not aspirin or clopidogrel. Gastrointest Endosc. 2011;
74:22–34.
19. NapolEon B, Boneu B, Maillard L, et al. Board of the French
Society for Digestive Endoscopy (SFED). Guidelines of the
French Society for Digestive Endoscopy (SFED). Endoscopy.
2006;38:632–638.
20. ASGE Standards of Practice Committee, Anderson MA, Ben-
Menachem T, et al. Management of antithrombotic agents for
endoscopic procedures. Gastrointest Endosc. 2009;70:1061–
1070.
21. Bhatt DL, Scheiman J, Abraham NS, et al. ACCF/ACG/AHA
2008 expert consensus document on reducing the gastrointestinal
risks of antiplatelet therapy and NSAID use: a report of the
American College of Cardiology Foundation Task Force on
Clinical Expert Consensus Documents. J Am Coll Cardiol.2008;52:1502–1517.
22. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW,
Radford MJ. Validation of clinical classification schemes for
predicting stroke: Results from the National Registry of Atrial
Fibrillation. JAMA. 2001;285:2864–2870.
23. Risk factors for stroke and efficacy of antithrombotic therapy in
atrial fibrillation. Analysis of pooled data from five randomized
controlled trials. Arch Intern Med 1994;154:1449–57. (Erratum
in 1994;154:2254).
24. Research Committee of the British Thoracic Society. Optimum
duration of anticoagulation for deep-vein thrombosis and pul-
monary embolism. Lancet 1992;340:873–876.
25. European Heart Rhythm Association, Heart Rhythm Society,
Fuster V, et al. ACC/AHA/ESC 2006 guidelines for the man-
agement of patients with atrial fibrillation—executive summary:
a report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines and the European
Society of Cardiology Committee for Practice Guidelines
(Writing Committee to Revise the 2001 Guidelines for the
Management of Patients With Atrial Fibrillation). J Am CollCardiol. 2006;48:854–906.
26. Lee SY, Tang SJ, Rockey DC, et al. Korean Association for the
study of intestinal disease. Managing anticoagulation and anti-
platelet medications in GI endoscopy: a survey comparing the
East and the West. Gastrointest Endosc. 2008;67:1076–1081.
27. Fujishiro M, Oda I, Yamamoto Y, et al. Multi-center survey
regarding the management of anticoagulation and antiplatelet
therapy for endoscopic procedures in Japan. J GastroenterolHepatol. 2009;24:214–218. Epub. 09/24/2008.
28. Goel A, Barnes CJ, Osman H, Verma A. National survey of
anticoagulation policy in endoscopy. Eur J Gastroenterol Hep-atol. 2007;19:51–56.
29. Mosler P, Mergener K, Denzer U, Kiesslich R, Galle PR, Kanzler
S. Current practice in managing patients on anticoagulants and/or
antiplatelet agents around the time of gastrointestinal endos-
copy—a nation-wide survey in Germany. Z Gastroenterol.2004;42:1289–1293.
30. Ponsky JL, Gauderer MW. Percutaneous endoscopic gastros-
tomy: a nonoperative technique for feeding gastrostomy. Gas-trointest Endosc. 1981;27:9–11.
31. Cotton P, Williams C. Practical Gastrointestinal Endoscopy. 4th
ed. Oxford: Alden Press; 1996:168.
32. Pender SM, Courtney MG, Rajan E, McAdam B, Fielding JF.
Percutaneous endoscopic gastrostomy; results of an Irish single
unit series. Ir J Med Sci. 1993;162:452–455.
33. Gibson SE, Wenig BL, Watkins JL. Complications of percuta-
neous endoscopic gastrostomy in head and neck cancer patients.
Ann Otol Rhinol Laryngol. 1992;101:46–50.
34. Disario J. Endoscopic approaches to enteral nutritional support.
Best Pract Res Clin Gastroenterol. 2006;20:605–630.
35. Lin HS, Ibrahim HZ, Kheng JW, Fee WE, Terris DJ. Percuta-
neous endoscopic gastrostomy: strategies for prevention and
management of complications. Laryngoscope. 2001;111:1847–
1852.
36. Cappell MS, Abdullah M. Management of gastrointestinal
bleeding induced by gastrointestinal endoscopy. GastroenterolClin North Am. 2000;29:125–167.
37. Iakovou I, Schmidt T, Bonizzoni E, et al. Incidence, predictors,
and outcome of thrombosis after successful implantation of drug-
eluting stents. JAMA. 2005;293:2126–2130.
Dig Dis Sci (2012) 57:973–980 979
123
38. Kim HS, Kim TI, Kim WH, et al. Risk factors for immediate
postpolypectomy bleeding of the colon: a multicenter study. Am JGastroenterol. 2006;101:1333–1341.
39. Sawhney MS, Salfiti N, Nelson DB, Lederle FA, Bond JH. Risk
factors for severe delayed postpolypectomy bleeding. Endoscopy.
2008;40:115–119.
40. Singh M, Mehta N, Murthy UK, Kaul V, Arif A, Newman N.
Postpolypectomy bleeding in patients undergoing colonoscopy
on uninterrupted clopidogrel therapy. Gastrointest Endosc.
2010;71:998–1005.
41. Ruthmann O, Seitz A, Richter S, et al. Percutaneous endoscopic
gastrostomy. Complications with and without anticoagulation.
Chirurg. 2010;81:247–254.
42. Witt DM, Delate T, McCool KH, et al. WARPED Consortium.
Incidence and predictors of bleeding or thrombosis after poly-
pectomy in patients receiving and not receiving anticoagulation
therapy. J Thromb Haemost. 2009;7:1982–1989. Epub. 08/28/
2009.
43. Baglin TP, Keeling DM, Watson HG. Guidelines on oral
anticoagulation (warfarin): third edition—2005 update. Br JHaematol 2006;1:277–285. Gut. 2008;57:1322-1329. Epub.
05/9/2008.
44. Veitch AM, Baglin TP, Gershlick AH, et al. Guidelines for the
management of anticoagulant and antiplatelet therapy in patients
undergoing endoscopic procedures. Gut. 2008;57:1322–1329.
Epub. 05/9/2008.
45. Ansell JE. The perioperative management of warfarin therapy.
Arch Intern Med. 2003;163:881–883.
46. Turpie AG, Antman EM. Low-molecular-weight heparins in the
treatment of acute coronary syndromes. Arch Intern Med.
2001;161:1484–1490.
47. Kadayifci A, Haznedaroglu I, Savas M. Risk of withdrawing
chronic anticoagulation without antithrombotic prophylaxis. Am JGastroenterol. 1997;92:361–362.
48. Spyropoulos AC, Turpie A. Perioperative bridging interruption
with heparin for the patient receiving long-term anticoagulation.
Curr Opin Pulm Med. 2005;11:373–379.
49. Constans M, Santamaria A, Mateo J, Pujol N, Souto JC, Fon-
tcuberta J. Low-molecular-weight heparin as bridging therapy
during interruption of oral anticoagulation in patients undergoing
colonoscopy or gastroscopy. Int J ClinPract. 2007;61:212–217.
50. Kwok A, Faigel DO. Management of anticoagulation before
and after gastrointestinal endoscopy. Am J Gastroenterol.2009;104:3085–3097; quiz 3098. Epub. 08/11/2009. Review.
980 Dig Dis Sci (2012) 57:973–980
123