JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY VOL. 68, NO.

2, 2016

ª 2016 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION ISSN 0735-1097/$36.00

PUBLISHED BY ELSEVIER http://dx.doi.org/10.1016/j.jacc.2016.04.042

REVIEW TOPIC OF THE WEEK

Management of Periprocedural
Anticoagulation
A Survey of Contemporary Practice

Greg C. Flaker, MD,a Paul Theriot, BSBA,b Lea G. Binder, MA,b Paul P. Dobesh, PHARMD,c Adam Cuker, MD,d
John U. Doherty, MDe

ABSTRACT

Interruption of oral anticoagulation (AC) for surgery or an invasive procedure is a complicated process. Practice guidelines
provide only general recommendations, and care of such patients occurs across multiple specialties. The availability
of direct oral anticoagulants further complicates decision making and guidance here is limited. To evaluate current
practice patterns in the United States for bridging AC, a survey was developed by the American College of Cardiology
Anticoagulation Work Group. The goal of the survey was to assess how general and subspecialty cardiologists, internists,
gastroenterologists, and orthopedic surgeons currently manage patients who receive AC and undergo surgery or an
invasive procedure. The survey was completed by 945 physicians involved in the periprocedural management of AC. The
results provide a template for educational and research projects geared toward the development of clinical pathways and
point-of-care tools to improve this area of health care. (J Am Coll Cardiol 2016;68:217–26) © 2016 by the American
College of Cardiology Foundation.

A nnually, 10% to 15% of patients who receive

oral anticoagulation (AC) therapy require
treatment interruption for surgery or an inva-
sive procedure (1,2). Parenteral AC, typically with
unfractionated heparin or low molecular weight hepa-
rin, is thought to prevent thromboembolic events (TE)
during the time when patients do not receive oral AC.
The safety and efficacy of this practice of “bridging
anticoagulation” has been called into question with
several developments in AC therapy.
First, it has been difficult to show that bridging AC
prevents TE. A systematic review and meta-analysis
in patients who required interruption of vitamin K
antagonist (VKA) therapy showed no significant
differences in TE between those patients who
received parenteral AC and those who did not. A sig-
nificant excess of major bleeding was noted in those
receiving parenteral AC (3). In the recently published,
randomized, double-blind BRIDGE (Bridging Anti-
coagulation in Patients who Require Temporary
Interruption of Warfarin Therapy for an Elective Pro-
cedure or Surgery) study, patients receiving VKAs for
atrial fibrillation, who were at moderate risk for TE
and who were undergoing surgery, were randomized
to dalteparin or placebo. A low rate of TE, not signifi-
cantly different between placebo and dalteparin, was
noted. Significantly higher rates of major bleeding
occurred with dalteparin (4).

Listen to this manuscript’s
audio summary by
JACC Editor-in-Chief
Dr. Valentin Fuster.
From the aUniversity of Missouri School of Medicine, Columbia, Missouri; bAmerican College of Cardiology, Washington, DC;
c
College of Pharmacy, University of Nebraska Medical Center, Omaha, Nebraska; dPerelman School of Medicine, University of
Pennsylvania, Philadelphia, Pennsylvania; and the eSidney Kimmel Medical College, Thomas Jefferson University, Philadelphia,
Pennsylvania. The American College of Cardiology provided funds for this project. Dr. Flaker is a consultant for Boehringer
Ingelheim, Pfizer, Bristol-Myers Squibb, and Daiichi-Sankyo. Dr. Dobesh is a consultant for Janssen, Daiichi-Sankyo, Pfizer, Bristol-
Myers Squibb, and Boehringer-Ingelheim. Dr. Cuker is a consultant for Amgen, Biogen-Idec, Bracco, and Genzyme; and receives
grant support from Spark Therapeutics and T2 Biosystems. All other authors have reported that they have no relationships
relevant to the contents of this paper to disclose.

Manuscript received December 3, 2015; revised manuscript received April 5, 2016, accepted April 12, 2016.
218 Flaker et al.
Bridging Anticoagulation Practices
ABBREVIATIONS Second, a number of surgical procedures
AND ACRONYMS with a lower risk for bleeding can be per-
formed with brief or no interruption of
AC = anticoagulation
warfarin. These include pacemaker and
ACC = American College of
implantable cardioverter-defibrillator im-
Cardiology
plantation, dental extraction, and cataract
b.i.d. = twice daily
surgery (5–9). The ability to perform pro-
DOAC = direct-acting oral
anticoagulant
cedures at lower risk of bleeding without
interruption of oral AC reduces the need for
INR = international normalized
ratio parenteral AC and the additional risk of
TE = thromboembolic event bleeding.
VKA = vitamin K antagonist
Third, direct-acting oral anticoagulants
(DOACs) have been incorporated into clinical
practice. Unlike warfarin, which inhibits the synthesis
of several clotting factors, DOACs directly inhibit
selected components of the clotting cascade and have
a much more rapid onset and offset of action than
VKAs. On the basis of these pharmacological proper-
ties, many have questioned the need for the admin-
istration of parenteral AC when DOACs are
interrupted. However, an increased frequency of
stroke after cessation of DOACs has been reported
(10–13), leading to the inclusion of a Food and Drug
Administration recommendation in the prescribing
information, stating that coverage with another AC
should be considered if dabigatran, rivaroxaban,
apixaban, or edoxaban are discontinued. In point of
fact, this recommendation arose from the observation
of excess stroke rates at the end of pivotal clinical
trials, when patients were transitioned from a DOAC
back to warfarin. This was not meant to endorse
bridging when patients were taken off a DOAC for a
procedure, but the impact of this recommendation in
clinical practice is uncertain.
Finally, there is the realization that management of
AC in a patient requiring surgery or an invasive pro-
cedure is complex. The interruption and reinstitution
of oral AC, and the initiation and discontinuation of
parenteral AC requires coordination between a num-
ber of health care providers (14).
Because of these developments, and to better un-
derstand current practice patterns for patients
requiring interruption of AC therapy, a survey was
developed by members of the American College of
Cardiology (ACC) Anticoagulation Initiative Work
Group and completed by a variety of health care
providers in the United States who care for patients
receiving AC.
METHODS
The ACC’s Anticoagulation Initiative Work Group was
formed in 2013 to improve the delivery of AC care.
JACC VOL. 68, NO. 2, 2016
JULY 12, 2016:217–26
Members of this work group developed a survey,
approved by the ACC, which was sent to physicians
who care for patients on AC who undergo a proce-
dure. Initially, the online survey was distributed to
9,165 members of the ACC who agreed to participate.
General cardiologists (n ¼ 158, response rate 6.5%),
interventional cardiologists (n ¼ 161, response rate
3.3%), and electrophysiologists (n ¼ 163, response
rate 8.8%) completed the survey.
Internal medicine primary care physicians, gas-
troenterologists, and orthopedic surgeons were
identified through the Medical Panel of Research
Now, Inc. The proprietary Research Now Medical
panel is actively managed and updated with weekly
verification. The Research Now Medical panel uses a
“by invitation only” methodology, including online
recruitment, as well as a direct mail enrollment
campaign. The Research Now Medical panel is
American Medical Association verified to ensure that
all members enrolled in the panel are physicians, and
therefore provides accurate targeting across all med-
ical specialties. The survey was distributed to 3,054
physicians and was completed by internists (n ¼ 152,
response rate 13.9%), gastroenterologists (n ¼ 160,
response rate 13.0%), and orthopedic surgeons (n ¼
153, response rate 21.0%). For participation in this
survey, each panelist from the Research Now Medical
panel received $35.
The ACC provided financial support to Research
Now, which conducted the survey for the non-
cardiologists. The survey was performed between
July 22, 2015, and August 27, 2015. The complete
survey is available in the Online Appendix. The re-
spondents represented both private and academic
practices across the United States. Of the cardiologists
surveyed, 85% had primary board certification in in-
ternal medicine and 99% were board certified in car-
diovascular diseases. Detailed profile information
about the respondents is also available in the Online
Appendix.
RESULTS
WHO MANAGES PERIPROCEDURAL AC? When
asked who manages AC during and after surgical or
invasive procedures, the survey respondents said
that cardiologists are extensively involved in
decision-making processes, more commonly than the
physician performing the procedure (Figure 1). A
number of other health care professionals, including
primary care physicians, pharmacists, and nurses, are
involved in the periprocedural management of the
patient who receives oral AC.
JACC VOL. 68, NO. 2, 2016 Flaker et al. 219
JULY 12, 2016:217–26 Bridging Anticoagulation Practices

F I G U R E 1 Typical Managers of Periprocedural Anticoagulation

During After

Cardiologist 56% Cardiologist 47%

Physician performing procedure 36% Primary care physician 34%

Primary care physician 28% Anticoagulation service 28%

Anticoagulation service 27% Physician performing procedure 21%

Pharmacist 15% Pharmacist 14%

Nurse practitioner 11% Nurse practitioner 11%

Other 5% Other 4%

Not sure 2% Not sure 2%

IDENTIFICATION OF PATIENTS AT HIGH RISK respondents indicate that a CHADS2 score of 2 is

FOR TE. Patients at higher risk for TE may benefit
from parenteral AC. Most respondents (82%) consider
the presence of a mechanical heart valve to be of suf-
ficient risk to warrant parenteral AC. A prior stoke or
transient ischemic attack is sufficient to warrant
parenteral AC by 74% of the respondents. However,
there are significant variations in specialties as to
which patient is at high enough risk to warrant par-
enteral AC (Table 1). A greater percentage of general
cardiologists, electrophysiologists, and interventional
cardiologists identified the presence of a mechanical
heart valve, or a history of a stroke or transient
ischemic attack as a marker of increased risk for peri-
operative TE than survey respondents as a whole.
When asked what criteria are used for a patient
with atrial fibrillation without a mechanical heart
valve who undergoes surgery with a high risk of
bleeding, nearly one-half (44%) of the respondents
reported that they use the CHA2DS2 -VASc score to
decide who should receive parenteral AC. Fewer (11%)
report the use of the CHADS2 score. However, there
was no consensus as to what score is high enough to
warrant bridging (Figure 2). Of note is that 27% of
sufficient to warrant bridging, despite the fact that
the BRIDGE study, published in June 2015 (a month
before the survey), demonstrated that for patients at
moderately increased risk for stroke (mean CHADS 2
score 2.3) no bridging was noninferior to bridging for
preventing TE and was superior to bridging for pre-
venting major bleeding (4).
IDENTIFICATION OF SURGICAL PROCEDURES AT
LOW RISK FOR BLEEDING. A number of surgical
procedures with a low risk for bleeding can be
performed with brief or no interruption of
warfarin. These include pacemaker and implantable
cardioverter-defibrillator implantation, dental
extraction, and cataract surgery. Despite these data,
the perceived need to interrupt AC therapy and to use
parenteral AC among respondents was substantial.
Depending on the procedure, 17% to 37% of the
respondents would typically or sometimes inter-
rupt VKA therapy and administer parenteral AC
for procedures generally safe to be performed with-
out interruption (Figure 3). Differences between spe-
cialties are noted in Table 2. General cardiologists,

T A B L E 1 Most Common Parameters Used to Identify Patients at Increased Risk for TE During AC Interruption

General Interventional Orthopedic

Total Cardiologists Electrophysiologists Cardiologists Internists Gastroenterologists Surgeons
(n ¼ 947) (n ¼ 158) (n ¼ 163) (n ¼ 161) (n ¼ 152) (n ¼ 160) (n ¼ 153)

Mechanical heart valve 82% 94%* 95%* 91%* 65% 82%* 64%
Prior stroke or TIA 74% 78%* 83%* 79%* 64% 75% 62%
Risk of stroke/CHA2DS2-VASc score 70% 68% 71% 65% 70% 76% 69%

*Indicates significant differences at the 95% confidence level between physician groups compared with the group as a whole.
AC ¼ anticoagulation; TE ¼ thromboembolic event; TIA ¼ transient ischemic attack.
220 Flaker et al. JACC VOL. 68, NO. 2, 2016

Bridging Anticoagulation Practices JULY 12, 2016:217–26

F I G U R E 2 Risk Scores/Clinical Factors Used to Identify a Patient at High Risk for a TE

Risk Score / Clinical Factors Used to Determine Bridging

(n=947)

CHA2DS2-VASc 44%

Prior stroke or TIA alone 25%

CHADS2 11%
CHA2DS2-VASc Score for Bridging
(n=416) CHADS2 Score for Bridging
Other 3%
(n=106)
1 0%
None 5% 1 1%

2 20% Never bridge for Afib alone 12%

2 27%

3 37%
3 40%

4 23%
4 20%

>4 12%
>4 9%

Depends 7%
Depends 3%

Risk scoring systems (CHADS2 or CHA2DS2-VASc) used to determine if parenteral anticoagulation should be considered in the periprocedural
period and the risk at which respondents would provide bridging anticoagulation in a patient without a mechanical valve. Afib ¼ atrial
fibrillation; TE ¼ thromboembolic event; TIA ¼ transient ischemic attack.

interventional cardiologists, and electrophysiologists DEVICE IMPLANTATIONS IN PATIENTS RECEIVING

were more comfortable performing procedures at
lower risk of bleeding without AC interruption than
were noncardiologists.
USE OF PARENTERAL AC WITH DOACs. Given the
rapid onset and offset of the anticoagulant effect with
DOACs, parenteral AC might be unnecessary in pa-
tients undergoing temporary interruption of antico-
agulant therapy for surgical procedures. However,
when asked about periprocedural management of a
patient with atrial fibrillation at increased risk of TE
(CHA2DS 2-VASc $2), the number of respondents who
would bridge with a parenteral AC was similar, irre-
spective of whether the patient was taking warfarin or
a DOAC (Figure 4). As noted subsequently in the
Variation of Periprocedural Management section, it
appears that the pharmacokinetics of warfarin may
have been extrapolated by the respondents to the
DOACs, and DOACs are interrupted too early before a
procedure. The addition of a parental agent because
the interruption is so long then compounds this
“unforced error,” causing a second error because the
pharmacokinetics of LMWH are similar to those of
DOACs.
VKAS AND DOACs. When the survey results are
restricted only to electrophysiologists who are
implanting devices, nearly 9 of 10 (89%) are the sole
or major decision makers regarding perioperative AC.
Published studies have demonstrated the safety and
efficacy of performing device surgery without the
interruption of VKA antagonists (5–7). Most electro-
physiologists (69%) would not interrupt warfarin for
device implantation (Online Figure 1). Nearly 8 of 10
electrophysiologists (78%) would not stop warfarin
for device replacement.
However, there remains uncertainty on the part
of electrophysiologists concerning the safety and
efficacy of performing device implantation surgery
in the patient receiving a DOAC. A substantial minority
(34%) of electrophysiologists would perform an initial
implant without interruption of a DOAC, and 38%
would perform device replacement without interrup-
tion. When asked about a specific case scenario for
device implantation involving a hypothetical patient
with normal renal function who receives apixaban
5 mg twice daily (b.i.d.), the duration of interruption of
the DOAC varied. Nearly 1 in 5 of electrophysiologists
JACC VOL. 68, NO. 2, 2016 Flaker et al. 221
JULY 12, 2016:217–26 Bridging Anticoagulation Practices

F I G U R E 3 Procedures Performed With/Without Interruption of VKAs and With/Without Parenteral AC

Total Interrupt and

Administer PA

Dentel cleaning 4% 64% 15% 12% 5% 17%

Cataract removal 4% 45% 26% 14% 11% 25%

Upper endoscopy 3% 33% 35% 19% 10% 29%

Dental extraction 3% 26% 42% 19% 10% 29%

Pacemaker or defibrillator replacement 4% 35% 28% 17% 16% 33%

Colonoscopy 3% 27% 36% 22% 12% 34%

Coronary angiography 2% 20% 43% 20% 15% 35%

Pacemaker or defibrillator implantation 4% 30% 31% 17% 18% 35%

Catheter ablation 5% 33% 26% 20% 16% 36%

Typically interrupt and administer parenteral anticoagulant Interrupt without parenteral anticoagulation No answer

Sometimes interrupt and administer parenteral anticoagulant Proceed uninterrupted

Selected procedures performed with and without interruption of VKAs and procedures performed with and without parenteral AC.
AC ¼ anticoagulation; PA ¼ parenteral anticoagulation; VKAs ¼ vitamin K antagonists.

(21%) omit 1 dose, but others recommend stopping
DOAC therapy for 1 day (25%) or 2 days (25%) (Online
Figure 2). Electrophysiologists uncommonly (12%)
use parenteral AC in this situation.
VARIATION IN PERIPROCEDURAL
BETWEEN SPECIALTIES. There are differences of
opinion between the general cardiologist, the inter-
nist, and physicians performing procedures about the
management of periprocedural AC, as illustrated in the
following section. Two hypothetical patients were
presented including: 1) a 70-year-old with a mechani-
cal mitral valve prosthesis on warfarin; and 2) a patient
with a CHA 2DS2-VASc score of 5 and normal renal
function (creatinine clearance 90 ml/min) who is
treated with apixaban. Each patient undergoes an
elective procedure, including colonoscopy, hip
MANAGEMENT replacement, and coronary angiography (by a femoral
approach).
COLONOSCOPY. For colonoscopy, general cardiolo-
gists consider themselves to be either the sole or
major decision maker for the patient on warfarin
88% of the time. In contrast, the gastroenterologists
consider themselves to be the sole or major decision
maker for the patient on warfarin 67% of the time.

T A B L E 2 Percentage of Respondents Who Would Interrupt AC and Administer Parenteral AC for Various Procedures in a Patient on VKA
Who Is Not Low Risk for Stroke (CHA 2 DS 2 -VASc $2)

General Interventional Orthopedic

Cardiologists Electrophysiologists Cardiologists Internists Gastroenterologists Surgeons
(n ¼ 158) (n ¼ 163) (n ¼ 161) (n ¼ 152) (n ¼ 160) (n ¼ 153)

Dental cleaning 4% 9% 5% 29% 31% 29%

Cataract removal 9% 15% 9% 46% 40% 31%
Upper endoscopy 16% 21% 17% 37% 44% 39%
Dental extraction 16% 23% 14% 45% 39% 37%
Pacemaker or defibrillator replacement 15% 10% 15% 57% 56% 53%
Colonoscopy 23% 28% 19% 45% 46% 45%
Coronary angiography 18% 24% 16% 49% 53% 49%
Pacemaker or defibrillator implantation 17% 12% 17% 57% 59% 52%
Catheter ablation 27% 13% 24% 53% 53% 48%
Epidural injection for back pain relief 22% 36% 25% 53% 42% 41%

AC ¼ anticoagulation; VKA ¼ vitamin K antagonist.

222 Flaker et al. JACC VOL. 68, NO. 2, 2016

Bridging Anticoagulation Practices JULY 12, 2016:217–26

F I G U R E 4 Common Surgical Procedures in Which AC Is Interrupted

Warfarin Scenario DOAC Scenario

56% 23% 33% 28% 19% 47%

Hip replacement

56% 24% 32% Knee replacement 27% 19% 46%

51% 22% 29% Implantation of an EVAR 25% 16% 41%

Coronary artery bypass 41%

50% 19% 31% 26% 15%
surgery
Resection of abdominal 27% 15% 43%
49% 19% 30%
aortic aneurysm

45% 26% 19% Arthroscopic knee surgery 18% 20% 39%

44% Laparoscopic 18% 17% 35%

23% 21%
cholecystectomy
42% 21% 21% Transurethral transection 19% 17% 36%
prostate biopsy
40% 22% 18% Colonoscopy with biopsy 15% 16% 31%

38% 21% 17% Upper endoscopy with 15% 16% 31%

biopsy
38% 22% 16% Prostate biopsy 15% 18% 33%

Typically interrupt and administer parenteral anticoagulant Sometimes interrupt and administer parenteral anticoagulant

Common surgical procedures in which anticoagulation, either with vitamin K antagonists (VKAs) or with direct-acting oral anticoagulants (DOACs),
is interrupted and parenteral anticoagulation is provided. AC ¼ anticoagulation; EVAR ¼ endovascular aneurysm repair.

This underscores the importance of a team approach
in these patients and of assuring that someone is
actively directing management.
In the case of a patient with a mechanical mitral
valve, nearly 8 of 10 clinicians (79%) stop warfarin
prior to colonoscopy. A similar percentage of general
cardiologists (74%), internists (77%), and gastroen-
terologists (83%) stop warfarin 3 to 5 days prior to
colonoscopy.
Enoxaparin or another low molecular weight
heparin is the preferred parenteral anticoagulant
for cardiologists (78%), internists (70%), and gastro-
enterologists (70%). However, the duration of post-
procedure enoxaparin differed. More than 3 of 5
(63%) general cardiologists, internists, and gastroen-
terologists continue enoxaparin until the interna-
tional normalized ratio (INR) is above a threshold
value. The remaining respondents arbitrarily wait
several days before discontinuing enoxaparin, irre-
spective of the INR. This likely reflects the fact that
these patients are now outpatients not getting daily
INR monitoring, and the duration of enoxaparin is on
the basis of the pharmacokinetics of warfarin. Cardi-
ologists (81%) preferred to continue enoxaparin until
the INR was $2.0. An arbitrary 2 or more days was
preferred by internists (50%) and gastroenterologists
(48%) (Online Figure 3).
For the hypothetical patient with normal renal
function on apixaban who undergoes elective colo-
noscopy, the majority of cardiologists (67%) stop
apixaban 1 to 2 days prior to colonoscopy, compared
with 37% of internists and 42% of gastroenterologists.
Forty percent of gastroenterologists stop apixaban
3 to 5 days prior to the procedure. A comparison of
survey responses for periprocedural management
with warfarin and apixaban is illustrated in Table 3.
Interestingly, a substantial minority of respondents
would not interrupt warfarin or apixaban prior to
elective colonoscopy.
In this situation, most clinicians (89%) do not use
parenteral AC, but would simply start apixaban either
24 h (56%) or 48 h (33%) after the colonoscopy.
HIP REPLACEMENT. The responsibility for peri-
procedural management in a person with a mechan-
ical heart valve who is undergoing hip replacement
surgery is shared. The orthopedic surgeon considers
himself/herself to be either the sole or major decision
maker 53% of the time. The decision making is shared
31% of the time. The orthopedic surgeon has no in-
fluence or a minor influence only 15% of the time.
In this case scenario, almost 9 of 10 orthopedic
surgeons (87%) stop warfarin at least 3 days prior to
hip replacement. If parenteral AC is used, enoxaparin
JACC VOL. 68, NO. 2, 2016 Flaker et al. 223
JULY 12, 2016:217–26 Bridging Anticoagulation Practices

or some other low molecular weight heparin is

T A B L E 3 Duration of Interruption of AC Therapy for a 70-Year-Old Man With
preferred by 72% of orthopedic surgeons. Unfractio- a Mechanical Mitral Valve (Warfarin) or a Patient With a CHA 2 DS 2 -VASc Score
nated heparin is preferred by 19% of orthopedic sur- of 5 With Normal Renal Function (Apixaban) Undergoing Colonoscopy
geons. If enoxaparin is used for parenteral AC, 44% of
General Cardiology Internist Gastroenterologist
orthopedic surgeons favored 1 mg/kg daily. Another
(n ¼ 158) (n ¼ 158) (n ¼ 146) (n ¼ 141) (n ¼ 154) (n ¼ 157)
22% favored a dose of 1.5 mg/kg daily. Others
Warfarin Apixaban Warfarin Apixaban Warfarin Apixaban
preferred a non–weight-based dosing schedule.
5 days 37% 3% 42% 21% 40% 13%
Varying dosing schedules may reflect bleeding con-
4 days 16% 0% 5% 6% 12% 6%
cerns or a reversion to prophylactic dosing. Enox- 3 days 24% 12% 27% 18% 31% 21%
aparin is restarted on the day of the procedure by 24% 2 days 3% 45% 4% 25% 6% 27%
and 24 h after the procedure by 70% of orthopedic 1 day 0% 22% 3% 12% 3% 15%
surgeons. The majority of orthopedic surgeons (64%) Not stopped 17% 12% 9% 13% 9% 12%

continue the parenteral agent until the INR is above a

threshold value. The remaining orthopedic surgeons
arbitrarily wait several days before discontinuing
enoxaparin (Online Figure 4).
For the hypothetical patient with normal renal
function who receives apixaban and who undergoes
elective hip replacement, there is no consensus of
opinion as to the number of days apixaban should to
be discontinued. The largest number (31%) stop
apixaban 5 days prior to the procedure. Others stop
apixaban 1 day (9%), 2 days (23%), 3 days (22%), or 4
days (9%) prior to surgery (Online Figure 5).
In this situation, most orthopedic surgeons (84%)
do not use parenteral AC and simply start apixaban
within 24 h (53%) or 48 h (48%) after the procedure.
CORONARY ANGIOGRAPHY. Periprocedural man-
agement at the time of elective coronary angiography
via femoral approach in the patient with a mechanical
heart valve is primarily by the interventional cardi-
ologist. Interventional cardiologists consider them-
selves either the sole or major decision maker 92% of
the time.
For the patient receiving warfarin, more than 4 of 5
interventional cardiologists (83%) stop warfarin at
least 3 days prior to coronary angiography. Warfarin is
stopped 5 days, 4 days, and 3 days prior to the pro-
cedure by 27%, 15%, and 41% of interventional car-
diologists, respectively.
Nearly 9 of 10 interventional cardiologists (88%)
administer parenteral AC in the patient with a me-
chanical heart valve. Enoxaparin or another low mo-
lecular weight heparin is preferred by 73% of
interventional cardiologists and 14% prefer unfrac-
tionated heparin.
The dose of enoxaparin is 1 mg/kg b.i.d. for 86%
of interventional cardiologists. Enoxaparin is restar-
ted, either on the day of the procedure (51%) or 24 h
after the procedure (40%). Most interventional car-
diologists (71%) continue enoxaparin until the INR
is $2.0. Of the remaining interventional cardiologists,
21% prefer an arbitrary period of time before
AC ¼ anticoagulation.
enoxaparin is discontinued, irrespective of the INR
(Online Figure 6).
In the hypothetical patient with normal renal
function who is receiving apixaban, there is no
consensus of opinion as to the number of days apix-
aban needs to be discontinued. Almost all interven-
tional cardiologists (96%) stop apixaban prior to the
procedure. The duration of interruption of apixaban
is variable. Some stop apixaban for 1 dose (21%),
others for 1 day (25%), and others for 2 days (25%)
prior to coronary angiography. A comparison of when
warfarin and apixaban are stopped prior to elective
coronary angiography is shown in Online Figure 7.
If a patient receives apixaban, a parenteral agent is
not used by 84% of interventional cardiologists.
Instead, nearly 9 of 10 (89%) simply begin apixaban
within 24 h of coronary angiography.
INSTITUTIONAL GUIDELINES. More than 3 of 4 sur-
vey respondents believed that a standardized process
or protocol is either extremely or very important for
the periprocedural management of AC. However, 70%
of respondents said that a standardized protocol for
guidance of periprocedural management of AC is not
available in their practice, and an additional 13% were
not aware of whether a protocol was in place.
DISCUSSION
Bridging AC is a complicated undertaking (Central
Illustration). The periprocedural management of pa-
tients who receive AC and who undergo surgery or
invasive procedures involves a number of health care
providers, including physicians who perform pro-
cedures, the primary care physician, nurses, and
pharmacists. Although a decision maker is usually
identified, this role varies, depending on the proce-
dure. Given the number of health care providers
involved in this process, standardized protocols or
224 Flaker et al. JACC VOL. 68, NO. 2, 2016

Bridging Anticoagulation Practices JULY 12, 2016:217–26

one-half of the respondents were cardiologists). In

C E NT R AL IL L U ST R AT IO N The Need to Develop Consistent addition, the overall response rate was only 7.8%,
Clinical Guidelines for Perioperative/Post-Operative
which might bias the validity of the study. Regard-
Anticoagulation
less, a key message is that cardiologists are heavily
involved in the management of AC in patients un-
Cardiologist dergoing surgery or invasive procedures.
The survey highlighted that the CHA 2DS2-VASc
score is frequently used to identify patients for
or or Physician
performing parenteral AC. This score was originally designed to
Other
procedure identify risk factors for stroke in patients not
receiving AC. The score is currently used to identify
or Inconsistent or
decision maker patients who receive AC who might receive parenteral
in AC bridging AC when oral AC is interrupted. Although this makes
Primary
during/after
Pharmacist care clinical sense, the approach has never been validated.
surgery
physician
Furthermore, there is disagreement as to what score
or or
constitutes a high enough risk to justify bridging with
or parenteral AC. Given the excess risk of parenteral AC,
Nurse Anticoagulation
practitioner service clarification of risk factors in this population would
allow clinicians to provide parenteral AC only to pa-
tients at appropriately high risk. The double-blind,
Inconsistent randomized controlled trial of post-operative low
decisions made molecular weight heparin bridging versus placebo
in AC bridging
during/after bridging for patients who are at high risk for arterial
surgery thromboembolism (PERIOP 2 [Safety and Effective-
ness Study of LMWH Bridging Therapy for Patients on
Variance in which CHA2DS2-VASc Long Term Warfarin and Require Temporary Inter-
scores constitute a high ruption of Their Warfarin]) will help address the
thrombotic risk
efficacy of parenteral AC in patients at high risk of
TEs, including those with mechanical heart valves.
Variance in which procedures
constitute a high bleeding risk Previous studies have shown that the risk of TE is
low in selected patients who have AC interrupted for
No interruption surgical procedures (4,16). Surgical procedures at low
of VKA or DOAC Bridge with
or risk for bleeding that do not require interruption of
during surgery parenteral AC
AC have also been defined (5–9). Although many
Variance in lower-risk patients are undergoing procedures with a
parenteral AC lower risk of bleeding without interruption of VKA,
dosing strategies
there are still many physicians who prefer to inter-
Variance in rupt oral AC and provide parenteral AC. This puts
duration of patients at increased bleeding risk. The survey
parenteral AC
demonstrated that a considerable number of low-risk
patients receive parenteral AC and a considerable
Flaker, G.C. et al. J Am Coll Cardiol. 2016;68(2):217–26. number of low-risk procedures are being performed
with parenteral AC. More education for physicians
Systems of care in the perioperative management of anticoagulation. AC ¼ anti- caring for these patients is warranted.
coagulation; DOAC ¼ direct-acting oral anticoagulant; VKA ¼ vitamin K antagonist.
The survey confirmed the wide variability in dose
and duration of parenteral AC, consistent with what
has been reported in previous publications. Fixed,
pathways of care make sense. However, only a mi- weight-based doses (4), a weight-based initial dose
nority of respondents reported the existence of followed by reduced doses closer to surgery (17–19),
standardized periprocedural protocols at their in- and once a day or b.i.d. dosing (20) have been studied
stitutions. It has been shown that when AC protocols or advocated. Parenteral agents are administered
are devised and implemented, low TE and low rates only before the procedure, only after the procedure,
of major bleeding are observed (15). Cardiologists or both before and after the procedure (17). Although
were over-represented in this survey (approximately the most common parenteral agent was enoxaparin,
JACC VOL. 68, NO. 2, 2016
JULY 12, 2016:217–26
several dosing strategies were noted and tended to
vary depending on profession. These differences may
be explained by the philosophy of the clinician. For
example, less than one-half of orthopedic surgeons
use enoxaparin 1 mg/kg b.i.d. in the perioperative
period for hip replacement. In contrast, nearly 9 of
10 interventional cardiologists use a dose of 1 mg/kg
b.i.d. in the perioperative period for coronary angi-
ography. The orthopedic surgeon, concerned about
post-operative bleeding, favors a lower dose. The
interventional cardiologist, more concerned about
thrombus formation, favors the higher dose. After the
BRIDGE study, parenteral AC will likely be used less
often for patients at lower risk. Until additional data
are obtained, patients at the highest TE risk (most
mechanical heart valves, patients with CHADS scores
>4) might continue to receive b.i.d. parenteral AC;
others might receive once-daily parenteral AC, as
previously recommended in high-volume institutions
(19). In most patients, a reduction of the daily dose by
50% on the morning prior to the procedure can be
considered.
The survey also demonstrated that the duration of
parenteral AC was highly variable. For a patient
treated with warfarin, most survey respondents
favored continuation of parenteral AC until a thera-
peutic level of warfarin AC was reached. This
approach is labor intensive and requires repeated
post-operative measurements. This strategy is
optimal for ensuring continuous adequate AC
coverage for a patient at risk for TE. A number of
other survey respondents favored a more practical
approach, estimating the days required to reach a
therapeutic INR and continuing parenteral AC for
several days without repeated blood sampling. Given
the fact that 22% of AC errors involve parenteral AC
(21), additional studies to define the optimal dose and
duration of parenteral AC are needed. Guidance for
common AC-related management issues by expert
consensus in the United States was published in 2012
(20) and needs updating.
The survey also highlights confusion about peri-
procedural management of AC in the patient treated
with a DOAC. In the survey, management questions
related to a patient taking apixaban were posed. It
may not be valid to extrapolate these responses to
other DOACs. Although it would have been ideal to
include questions about dabigatran, rivaroxaban, and
edoxaban as well, the survey would have been un-
wieldy. Apixaban was selected due to its general fa-
miliarity and growing use in the cardiac community.
The purpose of the questions was to help define if
respondents thought that procedures could safely be
performed without DOAC interruption and, if not,
Flaker et al. 225
Bridging Anticoagulation Practices
how long should the DOAC be interrupted? Should
parenteral AC be used during DOAC interruption?
Guidance in this area is available (22) and derived
from post hoc analyses of clinical trials and from
registries. It appears that selected procedures with
low risk for bleeding can be performed without
interruption of the DOAC (apixaban) (23). Catheter
ablation is now safely performed without interrup-
tion of DOACs (24,25). If procedures have sufficient
bleeding risks, brief interruption of DOACs is associ-
ated with a low rate of TE, comparable to warfarin
(23,26). The duration of AC interruption is shorter
with a DOAC than with warfarin (24). In patients
receiving predominantly rivaroxaban or dabigatran,
the use of heparin bridging is associated with a higher
risk of major bleeding compared with those who did
not receive major bleeding, emphasizing concern
about the use of parenteral AC with DOAC interrup-
tion (27). In a prospective study with time of discon-
tinuation and resumption of dabigatran on the basis
of pharmacokinetic information and the type of sur-
gery or invasive procedure, a low risk of TE and major
bleeding has been reported (28).
The role of parenteral AC with DOACs was difficult
to understand in the survey. Some of the findings can
be explained by a deficit in knowledge of the phar-
macokinetic properties of DOACs.
Although the overall results suggested a similar
rate of use of parenteral AC in warfarin-treated and in
DOAC-treated patients who undergo surgery or an
invasive procedure, the case scenarios do suggest
that groups of individual clinicians, who perform
procedures at lower risk in patients at moderate risk,
infrequently use parental AC in DOAC-treated pa-
tients. The role of parenteral AC in higher-risk pa-
tients treated with DOACs undergoing surgeries or
invasive procedures with higher bleeding risks, and
who likely require interruption of AC for longer pe-
riods of time, remains uncertain.
CONCLUSIONS
Given this complex clinical scenario involving multi-
ple health care professionals, it makes sense to
develop consistently applied clinical pathways with
standardized institutional protocols. Most re-
spondents thought that this would be important.
This represents an important opportunity for pro-
fessional societies and guidelines committees to work
together to provide meaningful suggestions on the
basis of current data. There are areas in the peri-
procedural management of AC where clinical evi-
dence is clear-cut, and others where guidance needs
to be tempered by clinical judgment. Guidelines,
226 Flaker et al.
Bridging Anticoagulation Practices
unfortunately, do not chart a clear path in all cir-
cumstances. For the present, consensus documents,
clinical pathways, and point-of-care tools have enor-
mous potential to improve care in this area. Coordi-
nation among specialties, pharmacists, nursing, and
other health professionals has great potential for
enhancement of care.
There is also an opportunity for professional soci-
eties to use these data to support research to address
REFERENCES
1. Douketis JD, Spyropoulos AC, Spencer FA, et al.
Perioperative management of antithrombotic
therapy: Antithrombotic Therapy and Prevention
of Thrombosis, 9th ed: American College of Chest
Physicians Evidence-based Clinical Practice
Guidelines. Chest 2012;141:e326S–50S.
2. Steinberg BA, Peterson ED, Kim S, et al., Out-
comes Registry for Better Informed Treatment of
Atrial Fibrillation (ORBIT-AF) investigators and
patients. Use and outcomes associated with
bridging during anticoagulation interruptions in
patients with atrial fibrillation: findings from the
Outcomes Registry for Better Informed Treatment
of Atrial Fibrillation (ORBIT-AF). Circulation 2015;
131:488–94.
3. Siegal D, Yudin J, Kaatz S, et al. Periprocedural
heparin bridging in patients receiving vitamin K
antagonists: systematic review and meta-analysis
of bleeding and thromboembolic rates. Circula-
tion 2012;126:1630–9.
4. Douketis JD, Spyropoulos AC, Kaatz S, et al.,
BRIDGE Investigators. Perioperative bridging
anticoagulation in patients with atrial fibrillation.
N Engl J Med 2015;373:823–33.
5. Birnie DH, Healey JS, Wells GA, et al., BRUISE
CONTROL Investigators. Pacemaker or defibrillator
surgery without interruption of anticoagulation.
N Engl J Med 2013;268(22):2084–93.
6. Coyle D, Coyle K, Essebag V, et al. Cost
effectiveness of continued- warfarin versus
haparin-bringing therapy during pacemaker and
defibrillator surgery. J Am Coll Cardiol 2015;65:
957–9.
7. Cheng A, Nazarian S, Brinker JA, et al. Continu-
ation of warfarin during pacemaker or implantable
cardioverter-defibrillator implantation: a random-
ized clinical trial. Heart Rhythm 2011;8:536–40.
8. Bajkin BV, Popovic SL, Selakovic SD. Random-
ized, prospective trial comparing bridging therapy
using low-molecular-weight heparin with mainte-
nance of oral anticoagulation during extraction of
teeth. J Oral Maxillfac Surg 2009;67:990–5.
9. Katz J, Feldman MA, Bass EB, et al., Study of
Medical Testing for Cataract Surgery Team. Risks
and benefits of anticoagulant and antiplatelet
medication use before cataract surgery. Ophthal-
mology 2003;110:1784–8.
10. Patel MR, Mahaffey KW, Garg J, et al., ROCKET
AF Investigators. Rivaroxaban versus warfarin in
nonvalvular atrial fibrillation. N Engl J Med 2011;
365:883–91.
JACC VOL. 68, NO. 2, 2016
JULY 12, 2016:217–26
unanswered questions in the clinical space, as well as
to promote strong educational programs to improve
AC care.
REPRINT REQUESTS AND CORRESPONDENCE: Dr.
Greg C. Flaker, Department of Cardiovascular Medicine,
University of Missouri-Columbia, CE 351 University Hos-
pital, CE351, One Hospital Drive, Columbia, Missouri 65212.
E-mail: flakerg@health.missouri.edu.
11. Sherwood MW, Douketis JD, Patel MR, et al., Chest Physicians Evidence-Based Clinical Practice
ROCKET AF Investigators. Outcomes of temporary Guidelines. Chest 2012;141:e152S–84S.
interruption of rivaroxaban compared with
21. Amorosi SL, Tsilimingras K, Thompson D,
warfarin in patients with nonvalvular atrial fibril-
et al. Cost analysis of “bridging therapy” with
lation: results from the rivaroxaban once daily,
low-molecular-weight heparin versus unfractio-
oral, direct factor Xa inhibition compared with
nated heparin using temporary interruption of
vitamin k antagonism for prevention of stroke and
chronic anticoagulation. Am J Cardiol 2004;93:
embolism trial in atrial fibrillation (ROCKET AF).
509–11.
Circulation 2014;129:1850–9.
22. Heidbuchel H, Verhamme P, Alings M, et al.
12. Granger CB, Alexander JH, McMurray JJV,
European Heart Rhythm Association practical
et al., ARISTOTLE Committees and Investigators.
guide on the use of new oral anticoagulants in
Apixaban versus warfarin in patients with atrial
patients with non-valvular atrial fibrillation.
fibrillation. N Engl J Med 2011;365:981–92.
Europace 2013;15:625–51.
13. Granger CB, Lopes RD, Hanna M, et al. Clinical
23. Garcia D, Alexander JH, Wallentin L, et al.
events after transitioning from apixaban versus
Management and clinical outcomes in patients
warfarin to warfarin at the end of the Apixaban for
treated with apixaban vs warfarin undergoing
Reduction in Stroke and Other Thromboembolic
procedures. Blood 2014;124:3692–8.
Events in ATRIAL FIBRIllation (ARISTOTLE) trial.
Am Heart J 2015;169:25–30. 24. Maan A, Heist EK, Ruskin JN, et al. Practical
issues in the management of novel oral
14. Kovacs RJ, Flaker GC, Saxonhouse SJ, et al.
anticoagulants-cardioversion and ablation.
Practical management of anticoagulation in pa-
J Thorac Dis 2015;7:115–31.
tients with atrial fibrillation. J Am Coll Cardiol
2015;65:1340–60. 25. Cappato R, Marchlinski FE, Hohnloser SH,
et al., VENTURE-AF Investigators. Uninterrupted
15. Pengo V, Cucchini U, Denas G, et al., Italian
rivaroxaban vs. uninterrupted vitamin K antago-
Federation of Centers for the Diagnosis of
nists for catheter ablation in non-valvular atrial
Thrombosis and Management of Antithrombotic
fibrillation. Eur Heart J 2015;36:1805–11.
Therapies (FCSA). Standardized low-molecular-
26. Healy JS, Eikelboom J, Douketis J, et al., RE-
weight heparin bridging regimen in outpatients
LY Investigators. Periprocedural bleeding and
on oral anticoagulants undergoing invasive pro-
cedure or surgery: an inception cohort manage- thromboembolic events with dabigatran compared
ment study. Circulation 2009;119:2920–7. with warfarin: results from the Randomized Eval-
uation of Long-Term Anticoagulation Therapy
16. Garcia DA, Regan S, Henault LE, et al. Risk of
(RE-LY) randomized trial. Circulation 2012;126:
thromboembolism with short-term interruption of
343–8.
warfarin therapy. Arch Intern Med 2008;168:
63–9. 27. Beyer-Westendorf J, Gelbricht V, Förster K,
et al. Peri-interventional management of novel
17. Rechenmacher SJ, Fang JC. Bridging anti-
oral anticoagulants in daily care: results from the
coagulation: primum non nocere. J Am Coll Cardiol
prospective Dresden NOAC registry. Eur Heart J
2015;66:1392–403.
2014;35:1888–96.
18. Kovacs MJ, Kearon C, Rodger M, et al. Single-
28. Schulman S, Carrier M, Lee AYY, et al., Periop
arm study of bridging therapy with low-
Dabigitran Study Group. Perioperative manage-
molecular-weight heparin for patients at risk of
ment of dabigatran: a prospective cohort study.
arterial embolism who require temporary inter-
Circulation 2015;132:167–73.
ruption of warfarin. Circulation 2004;110:
1658–63.
19. Wysokinski W, McBane R II Periprocedural KEY WORDS atrial fibrillation, direct-
bridging management of anticoagulation. Circula- acting oral anticoagulant, parenteral
tion 2012;126:486–90. anticoagulation, vitamin K antagonists
20. Holbrook A, Schulman S, Witt DM, et al.,
American College of Chest Physicians. Evidence-
based management of anticoagulant therapy. A PPE NDI X For an expanded Methods section
Antithrombotic Therapy and Prevention of as well as supplemental tables and figures,
Thrombosis, 9th edition: American College of please see the online version of this article.